postgresql/src/backend/utils/adt/geo_ops.c

5558 lines
122 KiB
C
Raw Normal View History

/*-------------------------------------------------------------------------
*
* geo_ops.c
* 2D geometric operations
*
* This module implements the geometric functions and operators. The
* geometric types are (from simple to more complicated):
*
* - point
* - line
* - line segment
* - box
* - circle
* - polygon
*
* Portions Copyright (c) 1996-2020, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
*
* IDENTIFICATION
2010-09-20 22:08:53 +02:00
* src/backend/utils/adt/geo_ops.c
*
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include <math.h>
#include <limits.h>
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
#include <float.h>
#include <ctype.h>
#include "libpq/pqformat.h"
#include "miscadmin.h"
#include "utils/float.h"
#include "utils/fmgrprotos.h"
1997-03-15 00:21:12 +01:00
#include "utils/geo_decls.h"
/*
* * Type constructors have this form:
* void type_construct(Type *result, ...);
*
* * Operators commonly have signatures such as
* void type1_operator_type2(Type *result, Type1 *obj1, Type2 *obj2);
*
* Common operators are:
* * Intersection point:
* bool type1_interpt_type2(Point *result, Type1 *obj1, Type2 *obj2);
* Return whether the two objects intersect. If *result is not NULL,
* it is set to the intersection point.
*
* * Containment:
* bool type1_contain_type2(Type1 *obj1, Type2 *obj2);
* Return whether obj1 contains obj2.
* bool type1_contain_type2(Type1 *contains_obj, Type1 *contained_obj);
* Return whether obj1 contains obj2 (used when types are the same)
*
* * Distance of closest point in or on obj1 to obj2:
* float8 type1_closept_type2(Point *result, Type1 *obj1, Type2 *obj2);
* Returns the shortest distance between two objects. If *result is not
* NULL, it is set to the closest point in or on obj1 to obj2.
*
* These functions may be used to implement multiple SQL-level operators. For
* example, determining whether two lines are parallel is done by checking
* whether they don't intersect.
*/
/*
* Internal routines
*/
enum path_delim
{
PATH_NONE, PATH_OPEN, PATH_CLOSED
};
/* Routines for points */
static inline void point_construct(Point *result, float8 x, float8 y);
static inline void point_add_point(Point *result, Point *pt1, Point *pt2);
static inline void point_sub_point(Point *result, Point *pt1, Point *pt2);
static inline void point_mul_point(Point *result, Point *pt1, Point *pt2);
static inline void point_div_point(Point *result, Point *pt1, Point *pt2);
static inline bool point_eq_point(Point *pt1, Point *pt2);
static inline float8 point_dt(Point *pt1, Point *pt2);
static inline float8 point_sl(Point *pt1, Point *pt2);
1998-09-01 05:29:17 +02:00
static int point_inside(Point *p, int npts, Point *plist);
/* Routines for lines */
static inline void line_construct(LINE *result, Point *pt, float8 m);
static inline float8 line_sl(LINE *line);
static inline float8 line_invsl(LINE *line);
static bool line_interpt_line(Point *result, LINE *l1, LINE *l2);
static bool line_contain_point(LINE *line, Point *point);
static float8 line_closept_point(Point *result, LINE *line, Point *pt);
/* Routines for line segments */
static inline void statlseg_construct(LSEG *lseg, Point *pt1, Point *pt2);
static inline float8 lseg_sl(LSEG *lseg);
static inline float8 lseg_invsl(LSEG *lseg);
static bool lseg_interpt_line(Point *result, LSEG *lseg, LINE *line);
static bool lseg_interpt_lseg(Point *result, LSEG *l1, LSEG *l2);
static int lseg_crossing(float8 x, float8 y, float8 px, float8 py);
static bool lseg_contain_point(LSEG *lseg, Point *point);
static float8 lseg_closept_point(Point *result, LSEG *lseg, Point *pt);
static float8 lseg_closept_line(Point *result, LSEG *lseg, LINE *line);
static float8 lseg_closept_lseg(Point *result, LSEG *on_lseg, LSEG *to_lseg);
/* Routines for boxes */
static inline void box_construct(BOX *result, Point *pt1, Point *pt2);
static void box_cn(Point *center, BOX *box);
static bool box_ov(BOX *box1, BOX *box2);
static float8 box_ar(BOX *box);
static float8 box_ht(BOX *box);
static float8 box_wd(BOX *box);
static bool box_contain_point(BOX *box, Point *point);
static bool box_contain_box(BOX *contains_box, BOX *contained_box);
static bool box_contain_lseg(BOX *box, LSEG *lseg);
static bool box_interpt_lseg(Point *result, BOX *box, LSEG *lseg);
static float8 box_closept_point(Point *result, BOX *box, Point *point);
static float8 box_closept_lseg(Point *result, BOX *box, LSEG *lseg);
/* Routines for circles */
static float8 circle_ar(CIRCLE *circle);
/* Routines for polygons */
static void make_bound_box(POLYGON *poly);
static void poly_to_circle(CIRCLE *result, POLYGON *poly);
static bool lseg_inside_poly(Point *a, Point *b, POLYGON *poly, int start);
static bool poly_contain_poly(POLYGON *contains_poly, POLYGON *contained_poly);
1998-09-01 05:29:17 +02:00
static bool plist_same(int npts, Point *p1, Point *p2);
static float8 dist_ppoly_internal(Point *pt, POLYGON *poly);
/* Routines for encoding and decoding */
static float8 single_decode(char *num, char **endptr_p,
const char *type_name, const char *orig_string);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
static void single_encode(float8 x, StringInfo str);
static void pair_decode(char *str, float8 *x, float8 *y, char **endptr_p,
const char *type_name, const char *orig_string);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
static void pair_encode(float8 x, float8 y, StringInfo str);
static int pair_count(char *s, char delim);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
static void path_decode(char *str, bool opentype, int npts, Point *p,
bool *isopen, char **endptr_p,
const char *type_name, const char *orig_string);
static char *path_encode(enum path_delim path_delim, int npts, Point *pt);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*
* Delimiters for input and output strings.
* LDELIM, RDELIM, and DELIM are left, right, and separator delimiters, respectively.
* LDELIM_EP, RDELIM_EP are left and right delimiters for paths with endpoints.
*/
#define LDELIM '('
#define RDELIM ')'
#define DELIM ','
#define LDELIM_EP '['
#define RDELIM_EP ']'
#define LDELIM_C '<'
#define RDELIM_C '>'
#define LDELIM_L '{'
#define RDELIM_L '}'
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*
* Geometric data types are composed of points.
* This code tries to support a common format throughout the data types,
* to allow for more predictable usage and data type conversion.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
* The fundamental unit is the point. Other units are line segments,
* open paths, boxes, closed paths, and polygons (which should be considered
* non-intersecting closed paths).
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*
* Data representation is as follows:
* point: (x,y)
* line segment: [(x1,y1),(x2,y2)]
* box: (x1,y1),(x2,y2)
* open path: [(x1,y1),...,(xn,yn)]
* closed path: ((x1,y1),...,(xn,yn))
* polygon: ((x1,y1),...,(xn,yn))
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*
* For boxes, the points are opposite corners with the first point at the top right.
* For closed paths and polygons, the points should be reordered to allow
* fast and correct equality comparisons.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*
* XXX perhaps points in complex shapes should be reordered internally
* to allow faster internal operations, but should keep track of input order
* and restore that order for text output - tgl 97/01/16
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
static float8
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
single_decode(char *num, char **endptr_p,
const char *type_name, const char *orig_string)
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
return float8in_internal(num, endptr_p, type_name, orig_string);
2017-06-21 20:39:04 +02:00
} /* single_decode() */
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
static void
single_encode(float8 x, StringInfo str)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
char *xstr = float8out_internal(x);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
appendStringInfoString(str, xstr);
pfree(xstr);
2017-06-21 20:39:04 +02:00
} /* single_encode() */
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
static void
pair_decode(char *str, float8 *x, float8 *y, char **endptr_p,
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
const char *type_name, const char *orig_string)
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
bool has_delim;
while (isspace((unsigned char) *str))
str++;
if ((has_delim = (*str == LDELIM)))
str++;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
*x = float8in_internal(str, &str, type_name, orig_string);
if (*str++ != DELIM)
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
type_name, orig_string)));
*y = float8in_internal(str, &str, type_name, orig_string);
if (has_delim)
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
if (*str++ != RDELIM)
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
type_name, orig_string)));
while (isspace((unsigned char) *str))
str++;
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
/* report stopping point if wanted, else complain if not end of string */
if (endptr_p)
*endptr_p = str;
else if (*str != '\0')
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
type_name, orig_string)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
static void
pair_encode(float8 x, float8 y, StringInfo str)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
char *xstr = float8out_internal(x);
char *ystr = float8out_internal(y);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
appendStringInfo(str, "%s,%s", xstr, ystr);
pfree(xstr);
pfree(ystr);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
static void
path_decode(char *str, bool opentype, int npts, Point *p,
bool *isopen, char **endptr_p,
const char *type_name, const char *orig_string)
{
int depth = 0;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
char *cp;
int i;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
while (isspace((unsigned char) *str))
str++;
if ((*isopen = (*str == LDELIM_EP)))
{
/* no open delimiter allowed? */
if (!opentype)
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
type_name, orig_string)));
depth++;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
str++;
1997-09-05 20:13:45 +02:00
}
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
else if (*str == LDELIM)
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
cp = (str + 1);
while (isspace((unsigned char) *cp))
cp++;
if (*cp == LDELIM)
{
depth++;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
str = cp;
}
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
else if (strrchr(str, LDELIM) == str)
{
depth++;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
str = cp;
}
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < npts; i++)
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
pair_decode(str, &(p->x), &(p->y), &str, type_name, orig_string);
if (*str == DELIM)
str++;
p++;
1997-09-05 20:13:45 +02:00
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
while (depth > 0)
{
if (*str == RDELIM || (*str == RDELIM_EP && *isopen && depth == 1))
{
depth--;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
str++;
while (isspace((unsigned char) *str))
str++;
}
else
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
type_name, orig_string)));
}
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
/* report stopping point if wanted, else complain if not end of string */
if (endptr_p)
*endptr_p = str;
else if (*str != '\0')
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
type_name, orig_string)));
2017-06-21 20:39:04 +02:00
} /* path_decode() */
static char *
path_encode(enum path_delim path_delim, int npts, Point *pt)
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
StringInfoData str;
int i;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
initStringInfo(&str);
switch (path_delim)
{
case PATH_CLOSED:
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
appendStringInfoChar(&str, LDELIM);
break;
case PATH_OPEN:
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
appendStringInfoChar(&str, LDELIM_EP);
break;
case PATH_NONE:
break;
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < npts; i++)
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
if (i > 0)
appendStringInfoChar(&str, DELIM);
appendStringInfoChar(&str, LDELIM);
pair_encode(pt->x, pt->y, &str);
appendStringInfoChar(&str, RDELIM);
pt++;
}
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
switch (path_delim)
{
case PATH_CLOSED:
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
appendStringInfoChar(&str, RDELIM);
break;
case PATH_OPEN:
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
appendStringInfoChar(&str, RDELIM_EP);
break;
case PATH_NONE:
break;
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
return str.data;
2017-06-21 20:39:04 +02:00
} /* path_encode() */
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*-------------------------------------------------------------
* pair_count - count the number of points
* allow the following notation:
* '((1,2),(3,4))'
* '(1,3,2,4)'
* require an odd number of delim characters in the string
*-------------------------------------------------------------*/
static int
pair_count(char *s, char delim)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
int ndelim = 0;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
while ((s = strchr(s, delim)) != NULL)
{
ndelim++;
s++;
}
1998-09-01 05:29:17 +02:00
return (ndelim % 2) ? ((ndelim + 1) / 2) : -1;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/***********************************************************************
**
** Routines for two-dimensional boxes.
**
***********************************************************************/
/*----------------------------------------------------------
* Formatting and conversion routines.
*---------------------------------------------------------*/
/* box_in - convert a string to internal form.
*
* External format: (two corners of box)
* "(f8, f8), (f8, f8)"
* also supports the older style "(f8, f8, f8, f8)"
*/
Datum
box_in(PG_FUNCTION_ARGS)
{
char *str = PG_GETARG_CSTRING(0);
BOX *box = (BOX *) palloc(sizeof(BOX));
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
bool isopen;
float8 x,
y;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
path_decode(str, false, 2, &(box->high), &isopen, NULL, "box", str);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* reorder corners if necessary... */
if (float8_lt(box->high.x, box->low.x))
{
x = box->high.x;
box->high.x = box->low.x;
box->low.x = x;
}
if (float8_lt(box->high.y, box->low.y))
{
y = box->high.y;
box->high.y = box->low.y;
box->low.y = y;
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_BOX_P(box);
}
/* box_out - convert a box to external form.
*/
Datum
box_out(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CSTRING(path_encode(PATH_NONE, 2, &(box->high)));
}
2003-05-13 20:03:08 +02:00
/*
* box_recv - converts external binary format to box
*/
Datum
box_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
BOX *box;
float8 x,
2003-05-13 20:03:08 +02:00
y;
box = (BOX *) palloc(sizeof(BOX));
box->high.x = pq_getmsgfloat8(buf);
box->high.y = pq_getmsgfloat8(buf);
box->low.x = pq_getmsgfloat8(buf);
box->low.y = pq_getmsgfloat8(buf);
/* reorder corners if necessary... */
if (float8_lt(box->high.x, box->low.x))
2003-05-13 20:03:08 +02:00
{
x = box->high.x;
box->high.x = box->low.x;
box->low.x = x;
}
if (float8_lt(box->high.y, box->low.y))
2003-05-13 20:03:08 +02:00
{
y = box->high.y;
box->high.y = box->low.y;
box->low.y = y;
}
PG_RETURN_BOX_P(box);
}
/*
* box_send - converts box to binary format
*/
Datum
box_send(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
StringInfoData buf;
pq_begintypsend(&buf);
pq_sendfloat8(&buf, box->high.x);
pq_sendfloat8(&buf, box->high.y);
pq_sendfloat8(&buf, box->low.x);
pq_sendfloat8(&buf, box->low.y);
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}
/* box_construct - fill in a new box.
*/
static inline void
box_construct(BOX *result, Point *pt1, Point *pt2)
{
if (float8_gt(pt1->x, pt2->x))
{
result->high.x = pt1->x;
result->low.x = pt2->x;
}
else
{
result->high.x = pt2->x;
result->low.x = pt1->x;
}
if (float8_gt(pt1->y, pt2->y))
{
result->high.y = pt1->y;
result->low.y = pt2->y;
}
else
{
result->high.y = pt2->y;
result->low.y = pt1->y;
}
}
/*----------------------------------------------------------
* Relational operators for BOXes.
* <, >, <=, >=, and == are based on box area.
*---------------------------------------------------------*/
/* box_same - are two boxes identical?
*/
Datum
box_same(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(point_eq_point(&box1->high, &box2->high) &&
point_eq_point(&box1->low, &box2->low));
}
/* box_overlap - does box1 overlap box2?
*/
Datum
box_overlap(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(box_ov(box1, box2));
}
static bool
box_ov(BOX *box1, BOX *box2)
{
return (FPle(box1->low.x, box2->high.x) &&
FPle(box2->low.x, box1->high.x) &&
FPle(box1->low.y, box2->high.y) &&
FPle(box2->low.y, box1->high.y));
}
/* box_left - is box1 strictly left of box2?
*/
Datum
box_left(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPlt(box1->high.x, box2->low.x));
}
/* box_overleft - is the right edge of box1 at or left of
* the right edge of box2?
*
* This is "less than or equal" for the end of a time range,
* when time ranges are stored as rectangles.
*/
Datum
box_overleft(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPle(box1->high.x, box2->high.x));
}
/* box_right - is box1 strictly right of box2?
*/
Datum
box_right(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPgt(box1->low.x, box2->high.x));
}
/* box_overright - is the left edge of box1 at or right of
* the left edge of box2?
*
* This is "greater than or equal" for time ranges, when time ranges
* are stored as rectangles.
*/
Datum
box_overright(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPge(box1->low.x, box2->low.x));
}
/* box_below - is box1 strictly below box2?
*/
Datum
box_below(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPlt(box1->high.y, box2->low.y));
}
/* box_overbelow - is the upper edge of box1 at or below
* the upper edge of box2?
*/
Datum
box_overbelow(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPle(box1->high.y, box2->high.y));
}
/* box_above - is box1 strictly above box2?
*/
Datum
box_above(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPgt(box1->low.y, box2->high.y));
}
/* box_overabove - is the lower edge of box1 at or above
* the lower edge of box2?
*/
Datum
box_overabove(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPge(box1->low.y, box2->low.y));
}
/* box_contained - is box1 contained by box2?
*/
Datum
box_contained(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(box_contain_box(box2, box1));
}
/* box_contain - does box1 contain box2?
*/
Datum
box_contain(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(box_contain_box(box1, box2));
}
/*
* Check whether the second box is in the first box or on its border
*/
static bool
box_contain_box(BOX *contains_box, BOX *contained_box)
{
return FPge(contains_box->high.x, contained_box->high.x) &&
FPle(contains_box->low.x, contained_box->low.x) &&
FPge(contains_box->high.y, contained_box->high.y) &&
FPle(contains_box->low.y, contained_box->low.y);
}
/* box_positionop -
* is box1 entirely {above,below} box2?
*
* box_below_eq and box_above_eq are obsolete versions that (probably
* erroneously) accept the equal-boundaries case. Since these are not
* in sync with the box_left and box_right code, they are deprecated and
* not supported in the PG 8.1 rtree operator class extension.
*/
Datum
box_below_eq(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPle(box1->high.y, box2->low.y));
}
Datum
box_above_eq(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPge(box1->low.y, box2->high.y));
}
/* box_relop - is area(box1) relop area(box2), within
* our accuracy constraint?
*/
Datum
box_lt(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPlt(box_ar(box1), box_ar(box2)));
}
Datum
box_gt(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPgt(box_ar(box1), box_ar(box2)));
}
Datum
box_eq(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPeq(box_ar(box1), box_ar(box2)));
}
Datum
box_le(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPle(box_ar(box1), box_ar(box2)));
}
Datum
box_ge(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(FPge(box_ar(box1), box_ar(box2)));
}
/*----------------------------------------------------------
* "Arithmetic" operators on boxes.
*---------------------------------------------------------*/
/* box_area - returns the area of the box.
*/
Datum
box_area(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
PG_RETURN_FLOAT8(box_ar(box));
}
/* box_width - returns the width of the box
* (horizontal magnitude).
*/
Datum
box_width(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
PG_RETURN_FLOAT8(box_wd(box));
}
/* box_height - returns the height of the box
* (vertical magnitude).
*/
Datum
box_height(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
PG_RETURN_FLOAT8(box_ht(box));
}
/* box_distance - returns the distance between the
* center points of two boxes.
*/
Datum
box_distance(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
Point a,
b;
box_cn(&a, box1);
box_cn(&b, box2);
PG_RETURN_FLOAT8(point_dt(&a, &b));
}
/* box_center - returns the center point of the box.
*/
Datum
box_center(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
Point *result = (Point *) palloc(sizeof(Point));
box_cn(result, box);
PG_RETURN_POINT_P(result);
}
/* box_ar - returns the area of the box.
*/
static float8
1997-09-08 22:59:27 +02:00
box_ar(BOX *box)
{
return float8_mul(box_wd(box), box_ht(box));
}
/* box_cn - stores the centerpoint of the box into *center.
*/
static void
box_cn(Point *center, BOX *box)
{
center->x = float8_div(float8_pl(box->high.x, box->low.x), 2.0);
center->y = float8_div(float8_pl(box->high.y, box->low.y), 2.0);
}
/* box_wd - returns the width (length) of the box
* (horizontal magnitude).
*/
static float8
1997-09-08 22:59:27 +02:00
box_wd(BOX *box)
{
return float8_mi(box->high.x, box->low.x);
}
/* box_ht - returns the height of the box
* (vertical magnitude).
*/
static float8
1997-09-08 22:59:27 +02:00
box_ht(BOX *box)
{
return float8_mi(box->high.y, box->low.y);
}
/*----------------------------------------------------------
* Funky operations.
*---------------------------------------------------------*/
/* box_intersect -
* returns the overlapping portion of two boxes,
* or NULL if they do not intersect.
*/
Datum
box_intersect(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0);
BOX *box2 = PG_GETARG_BOX_P(1);
BOX *result;
if (!box_ov(box1, box2))
PG_RETURN_NULL();
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (BOX *) palloc(sizeof(BOX));
result->high.x = float8_min(box1->high.x, box2->high.x);
result->low.x = float8_max(box1->low.x, box2->low.x);
result->high.y = float8_min(box1->high.y, box2->high.y);
result->low.y = float8_max(box1->low.y, box2->low.y);
PG_RETURN_BOX_P(result);
}
/* box_diagonal -
* returns a line segment which happens to be the
* positive-slope diagonal of "box".
*/
Datum
box_diagonal(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
LSEG *result = (LSEG *) palloc(sizeof(LSEG));
statlseg_construct(result, &box->high, &box->low);
PG_RETURN_LSEG_P(result);
}
/***********************************************************************
**
** Routines for 2D lines.
**
***********************************************************************/
static bool
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
line_decode(char *s, const char *str, LINE *line)
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
/* s was already advanced over leading '{' */
line->A = single_decode(s, &s, "line", str);
if (*s++ != DELIM)
return false;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
line->B = single_decode(s, &s, "line", str);
if (*s++ != DELIM)
return false;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
line->C = single_decode(s, &s, "line", str);
if (*s++ != RDELIM_L)
return false;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
while (isspace((unsigned char) *s))
s++;
if (*s != '\0')
return false;
return true;
}
Datum
line_in(PG_FUNCTION_ARGS)
{
char *str = PG_GETARG_CSTRING(0);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
LINE *line = (LINE *) palloc(sizeof(LINE));
LSEG lseg;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
bool isopen;
char *s;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
s = str;
while (isspace((unsigned char) *s))
s++;
if (*s == LDELIM_L)
{
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
if (!line_decode(s + 1, str, line))
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
errmsg("invalid input syntax for type %s: \"%s\"",
"line", str)));
if (FPzero(line->A) && FPzero(line->B))
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid line specification: A and B cannot both be zero")));
}
else
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
{
path_decode(s, true, 2, &lseg.p[0], &isopen, NULL, "line", str);
if (point_eq_point(&lseg.p[0], &lseg.p[1]))
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid line specification: must be two distinct points")));
line_construct(line, &lseg.p[0], lseg_sl(&lseg));
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
}
PG_RETURN_LINE_P(line);
}
Datum
line_out(PG_FUNCTION_ARGS)
{
LINE *line = PG_GETARG_LINE_P(0);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
char *astr = float8out_internal(line->A);
char *bstr = float8out_internal(line->B);
char *cstr = float8out_internal(line->C);
PG_RETURN_CSTRING(psprintf("%c%s%c%s%c%s%c", LDELIM_L, astr, DELIM, bstr,
DELIM, cstr, RDELIM_L));
}
2003-05-13 20:03:08 +02:00
/*
* line_recv - converts external binary format to line
*/
Datum
line_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
LINE *line;
line = (LINE *) palloc(sizeof(LINE));
line->A = pq_getmsgfloat8(buf);
line->B = pq_getmsgfloat8(buf);
line->C = pq_getmsgfloat8(buf);
if (FPzero(line->A) && FPzero(line->B))
ereport(ERROR,
(errcode(ERRCODE_INVALID_BINARY_REPRESENTATION),
errmsg("invalid line specification: A and B cannot both be zero")));
PG_RETURN_LINE_P(line);
2003-05-13 20:03:08 +02:00
}
/*
* line_send - converts line to binary format
*/
Datum
line_send(PG_FUNCTION_ARGS)
{
LINE *line = PG_GETARG_LINE_P(0);
StringInfoData buf;
pq_begintypsend(&buf);
pq_sendfloat8(&buf, line->A);
pq_sendfloat8(&buf, line->B);
pq_sendfloat8(&buf, line->C);
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
2003-05-13 20:03:08 +02:00
}
/*----------------------------------------------------------
* Conversion routines from one line formula to internal.
* Internal form: Ax+By+C=0
*---------------------------------------------------------*/
/*
* Fill already-allocated LINE struct from the point and the slope
*/
static inline void
line_construct(LINE *result, Point *pt, float8 m)
{
if (m == DBL_MAX)
{
/* vertical - use "x = C" */
result->A = -1.0;
result->B = 0.0;
result->C = pt->x;
}
else
{
/* use "mx - y + yinter = 0" */
result->A = m;
result->B = -1.0;
result->C = float8_mi(pt->y, float8_mul(m, pt->x));
/* on some platforms, the preceding expression tends to produce -0 */
if (result->C == 0.0)
result->C = 0.0;
}
}
/* line_construct_pp()
* two points
*/
Datum
line_construct_pp(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
LINE *result = (LINE *) palloc(sizeof(LINE));
if (point_eq_point(pt1, pt2))
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("invalid line specification: must be two distinct points")));
line_construct(result, pt1, point_sl(pt1, pt2));
PG_RETURN_LINE_P(result);
}
/*----------------------------------------------------------
* Relative position routines.
*---------------------------------------------------------*/
Datum
line_intersect(PG_FUNCTION_ARGS)
{
LINE *l1 = PG_GETARG_LINE_P(0);
LINE *l2 = PG_GETARG_LINE_P(1);
PG_RETURN_BOOL(line_interpt_line(NULL, l1, l2));
}
Datum
line_parallel(PG_FUNCTION_ARGS)
{
LINE *l1 = PG_GETARG_LINE_P(0);
LINE *l2 = PG_GETARG_LINE_P(1);
PG_RETURN_BOOL(!line_interpt_line(NULL, l1, l2));
}
Datum
line_perp(PG_FUNCTION_ARGS)
{
LINE *l1 = PG_GETARG_LINE_P(0);
LINE *l2 = PG_GETARG_LINE_P(1);
if (FPzero(l1->A))
PG_RETURN_BOOL(FPzero(l2->B));
if (FPzero(l2->A))
PG_RETURN_BOOL(FPzero(l1->B));
if (FPzero(l1->B))
PG_RETURN_BOOL(FPzero(l2->A));
if (FPzero(l2->B))
PG_RETURN_BOOL(FPzero(l1->A));
PG_RETURN_BOOL(FPeq(float8_div(float8_mul(l1->A, l2->A),
float8_mul(l1->B, l2->B)), -1.0));
}
Datum
line_vertical(PG_FUNCTION_ARGS)
{
LINE *line = PG_GETARG_LINE_P(0);
PG_RETURN_BOOL(FPzero(line->B));
}
Datum
line_horizontal(PG_FUNCTION_ARGS)
{
LINE *line = PG_GETARG_LINE_P(0);
PG_RETURN_BOOL(FPzero(line->A));
}
/*
* Check whether the two lines are the same
*
* We consider NaNs values to be equal to each other to let those lines
* to be found.
*/
Datum
line_eq(PG_FUNCTION_ARGS)
{
LINE *l1 = PG_GETARG_LINE_P(0);
LINE *l2 = PG_GETARG_LINE_P(1);
float8 ratio;
if (!FPzero(l2->A) && !isnan(l2->A))
ratio = float8_div(l1->A, l2->A);
else if (!FPzero(l2->B) && !isnan(l2->B))
ratio = float8_div(l1->B, l2->B);
else if (!FPzero(l2->C) && !isnan(l2->C))
ratio = float8_div(l1->C, l2->C);
else
ratio = 1.0;
PG_RETURN_BOOL((FPeq(l1->A, float8_mul(ratio, l2->A)) &&
FPeq(l1->B, float8_mul(ratio, l2->B)) &&
FPeq(l1->C, float8_mul(ratio, l2->C))) ||
(float8_eq(l1->A, l2->A) &&
float8_eq(l1->B, l2->B) &&
float8_eq(l1->C, l2->C)));
}
/*----------------------------------------------------------
* Line arithmetic routines.
*---------------------------------------------------------*/
/*
* Return slope of the line
*/
static inline float8
line_sl(LINE *line)
{
if (FPzero(line->A))
return 0.0;
if (FPzero(line->B))
return DBL_MAX;
return float8_div(line->A, -line->B);
}
/*
* Return inverse slope of the line
*/
static inline float8
line_invsl(LINE *line)
{
if (FPzero(line->A))
return DBL_MAX;
if (FPzero(line->B))
return 0.0;
return float8_div(line->B, line->A);
}
/* line_distance()
* Distance between two lines.
*/
Datum
line_distance(PG_FUNCTION_ARGS)
{
LINE *l1 = PG_GETARG_LINE_P(0);
LINE *l2 = PG_GETARG_LINE_P(1);
float8 ratio;
if (line_interpt_line(NULL, l1, l2)) /* intersecting? */
PG_RETURN_FLOAT8(0.0);
if (!FPzero(l1->A) && !isnan(l1->A) && !FPzero(l2->A) && !isnan(l2->A))
ratio = float8_div(l1->A, l2->A);
else if (!FPzero(l1->B) && !isnan(l1->B) && !FPzero(l2->B) && !isnan(l2->B))
ratio = float8_div(l1->B, l2->B);
else
ratio = 1.0;
PG_RETURN_FLOAT8(float8_div(fabs(float8_mi(l1->C,
float8_mul(ratio, l2->C))),
HYPOT(l1->A, l1->B)));
}
/* line_interpt()
* Point where two lines l1, l2 intersect (if any)
*/
Datum
line_interpt(PG_FUNCTION_ARGS)
{
LINE *l1 = PG_GETARG_LINE_P(0);
LINE *l2 = PG_GETARG_LINE_P(1);
Point *result;
result = (Point *) palloc(sizeof(Point));
if (!line_interpt_line(result, l1, l2))
PG_RETURN_NULL();
PG_RETURN_POINT_P(result);
}
/*
* Internal version of line_interpt
*
* Return whether two lines intersect. If *result is not NULL, it is set to
* the intersection point.
*
* NOTE: If the lines are identical then we will find they are parallel
* and report "no intersection". This is a little weird, but since
* there's no *unique* intersection, maybe it's appropriate behavior.
*
* If the lines have NaN constants, we will return true, and the intersection
* point would have NaN coordinates. We shouldn't return false in this case
* because that would mean the lines are parallel.
*/
static bool
line_interpt_line(Point *result, LINE *l1, LINE *l2)
{
float8 x,
y;
if (!FPzero(l1->B))
{
if (FPeq(l2->A, float8_mul(l1->A, float8_div(l2->B, l1->B))))
return false;
x = float8_div(float8_mi(float8_mul(l1->B, l2->C),
float8_mul(l2->B, l1->C)),
float8_mi(float8_mul(l1->A, l2->B),
float8_mul(l2->A, l1->B)));
y = float8_div(-float8_pl(float8_mul(l1->A, x), l1->C), l1->B);
}
else if (!FPzero(l2->B))
{
if (FPeq(l1->A, float8_mul(l2->A, float8_div(l1->B, l2->B))))
return false;
x = float8_div(float8_mi(float8_mul(l2->B, l1->C),
float8_mul(l1->B, l2->C)),
float8_mi(float8_mul(l2->A, l1->B),
float8_mul(l1->A, l2->B)));
y = float8_div(-float8_pl(float8_mul(l2->A, x), l2->C), l2->B);
}
else
return false;
/* On some platforms, the preceding expressions tend to produce -0. */
if (x == 0.0)
x = 0.0;
if (y == 0.0)
y = 0.0;
if (result != NULL)
point_construct(result, x, y);
return true;
}
/***********************************************************************
**
** Routines for 2D paths (sequences of line segments, also
** called `polylines').
**
** This is not a general package for geometric paths,
** which of course include polygons; the emphasis here
** is on (for example) usefulness in wire layout.
**
***********************************************************************/
/*----------------------------------------------------------
* String to path / path to string conversion.
* External format:
* "((xcoord, ycoord),... )"
* "[(xcoord, ycoord),... ]"
* "(xcoord, ycoord),... "
* "[xcoord, ycoord,... ]"
* Also support older format:
* "(closed, npts, xcoord, ycoord,... )"
*---------------------------------------------------------*/
*) inet_(client|server)_(addr|port)() and necessary documentation for the four functions. > Also, please justify the temp-related changes. I was not aware that we > had any breakage there. patch-tmp-schema.txt contains the following bits: *) Changes pg_namespace_aclmask() so that the superuser is always able to create objects in the temp namespace. *) Changes pg_namespace_aclmask() so that if this is a temp namespace, objects are only allowed to be created in the temp namespace if the user has TEMP privs on the database. This encompasses all object creation, not just TEMP tables. *) InitTempTableNamespace() checks to see if the current user, not the session user, has access to create a temp namespace. The first two changes are necessary to support the third change. Now it's possible to revoke all temp table privs from non-super users and limiting all creation of temp tables/schemas via a function that's executed with elevated privs (security definer). Before this change, it was not possible to have a setuid function to create a temp table/schema if the session user had no TEMP privs. patch-area-path.txt contains: *) Can now determine the area of a closed path. patch-dfmgr.txt contains: *) Small tweak to add the library path that's being expanded. I was using $lib/foo.so and couldn't easily figure out what the error message, "invalid macro name in dynamic library path" meant without looking through the source code. With the path in there, at least I know where to start looking in my config file. Sean Chittenden
2004-05-26 20:35:51 +02:00
Datum
path_area(PG_FUNCTION_ARGS)
{
2004-08-29 07:07:03 +02:00
PATH *path = PG_GETARG_PATH_P(0);
float8 area = 0.0;
2004-08-29 07:07:03 +02:00
int i,
j;
*) inet_(client|server)_(addr|port)() and necessary documentation for the four functions. > Also, please justify the temp-related changes. I was not aware that we > had any breakage there. patch-tmp-schema.txt contains the following bits: *) Changes pg_namespace_aclmask() so that the superuser is always able to create objects in the temp namespace. *) Changes pg_namespace_aclmask() so that if this is a temp namespace, objects are only allowed to be created in the temp namespace if the user has TEMP privs on the database. This encompasses all object creation, not just TEMP tables. *) InitTempTableNamespace() checks to see if the current user, not the session user, has access to create a temp namespace. The first two changes are necessary to support the third change. Now it's possible to revoke all temp table privs from non-super users and limiting all creation of temp tables/schemas via a function that's executed with elevated privs (security definer). Before this change, it was not possible to have a setuid function to create a temp table/schema if the session user had no TEMP privs. patch-area-path.txt contains: *) Can now determine the area of a closed path. patch-dfmgr.txt contains: *) Small tweak to add the library path that's being expanded. I was using $lib/foo.so and couldn't easily figure out what the error message, "invalid macro name in dynamic library path" meant without looking through the source code. With the path in there, at least I know where to start looking in my config file. Sean Chittenden
2004-05-26 20:35:51 +02:00
if (!path->closed)
PG_RETURN_NULL();
2004-08-29 07:07:03 +02:00
for (i = 0; i < path->npts; i++)
{
*) inet_(client|server)_(addr|port)() and necessary documentation for the four functions. > Also, please justify the temp-related changes. I was not aware that we > had any breakage there. patch-tmp-schema.txt contains the following bits: *) Changes pg_namespace_aclmask() so that the superuser is always able to create objects in the temp namespace. *) Changes pg_namespace_aclmask() so that if this is a temp namespace, objects are only allowed to be created in the temp namespace if the user has TEMP privs on the database. This encompasses all object creation, not just TEMP tables. *) InitTempTableNamespace() checks to see if the current user, not the session user, has access to create a temp namespace. The first two changes are necessary to support the third change. Now it's possible to revoke all temp table privs from non-super users and limiting all creation of temp tables/schemas via a function that's executed with elevated privs (security definer). Before this change, it was not possible to have a setuid function to create a temp table/schema if the session user had no TEMP privs. patch-area-path.txt contains: *) Can now determine the area of a closed path. patch-dfmgr.txt contains: *) Small tweak to add the library path that's being expanded. I was using $lib/foo.so and couldn't easily figure out what the error message, "invalid macro name in dynamic library path" meant without looking through the source code. With the path in there, at least I know where to start looking in my config file. Sean Chittenden
2004-05-26 20:35:51 +02:00
j = (i + 1) % path->npts;
area = float8_pl(area, float8_mul(path->p[i].x, path->p[j].y));
area = float8_mi(area, float8_mul(path->p[i].y, path->p[j].x));
*) inet_(client|server)_(addr|port)() and necessary documentation for the four functions. > Also, please justify the temp-related changes. I was not aware that we > had any breakage there. patch-tmp-schema.txt contains the following bits: *) Changes pg_namespace_aclmask() so that the superuser is always able to create objects in the temp namespace. *) Changes pg_namespace_aclmask() so that if this is a temp namespace, objects are only allowed to be created in the temp namespace if the user has TEMP privs on the database. This encompasses all object creation, not just TEMP tables. *) InitTempTableNamespace() checks to see if the current user, not the session user, has access to create a temp namespace. The first two changes are necessary to support the third change. Now it's possible to revoke all temp table privs from non-super users and limiting all creation of temp tables/schemas via a function that's executed with elevated privs (security definer). Before this change, it was not possible to have a setuid function to create a temp table/schema if the session user had no TEMP privs. patch-area-path.txt contains: *) Can now determine the area of a closed path. patch-dfmgr.txt contains: *) Small tweak to add the library path that's being expanded. I was using $lib/foo.so and couldn't easily figure out what the error message, "invalid macro name in dynamic library path" meant without looking through the source code. With the path in there, at least I know where to start looking in my config file. Sean Chittenden
2004-05-26 20:35:51 +02:00
}
PG_RETURN_FLOAT8(float8_div(fabs(area), 2.0));
*) inet_(client|server)_(addr|port)() and necessary documentation for the four functions. > Also, please justify the temp-related changes. I was not aware that we > had any breakage there. patch-tmp-schema.txt contains the following bits: *) Changes pg_namespace_aclmask() so that the superuser is always able to create objects in the temp namespace. *) Changes pg_namespace_aclmask() so that if this is a temp namespace, objects are only allowed to be created in the temp namespace if the user has TEMP privs on the database. This encompasses all object creation, not just TEMP tables. *) InitTempTableNamespace() checks to see if the current user, not the session user, has access to create a temp namespace. The first two changes are necessary to support the third change. Now it's possible to revoke all temp table privs from non-super users and limiting all creation of temp tables/schemas via a function that's executed with elevated privs (security definer). Before this change, it was not possible to have a setuid function to create a temp table/schema if the session user had no TEMP privs. patch-area-path.txt contains: *) Can now determine the area of a closed path. patch-dfmgr.txt contains: *) Small tweak to add the library path that's being expanded. I was using $lib/foo.so and couldn't easily figure out what the error message, "invalid macro name in dynamic library path" meant without looking through the source code. With the path in there, at least I know where to start looking in my config file. Sean Chittenden
2004-05-26 20:35:51 +02:00
}
Datum
path_in(PG_FUNCTION_ARGS)
{
char *str = PG_GETARG_CSTRING(0);
PATH *path;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
bool isopen;
char *s;
int npts;
int size;
int base_size;
int depth = 0;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
if ((npts = pair_count(str, ',')) <= 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
errmsg("invalid input syntax for type %s: \"%s\"",
"path", str)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
s = str;
while (isspace((unsigned char) *s))
s++;
/* skip single leading paren */
if ((*s == LDELIM) && (strrchr(s, LDELIM) == s))
{
s++;
depth++;
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
base_size = sizeof(path->p[0]) * npts;
2017-06-21 20:39:04 +02:00
size = offsetof(PATH, p) + base_size;
/* Check for integer overflow */
if (base_size / npts != sizeof(path->p[0]) || size <= base_size)
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("too many points requested")));
path = (PATH *) palloc(size);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
SET_VARSIZE(path, size);
path->npts = npts;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
path_decode(s, true, npts, &(path->p[0]), &isopen, &s, "path", str);
if (depth >= 1)
{
if (*s++ != RDELIM)
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for type %s: \"%s\"",
"path", str)));
while (isspace((unsigned char) *s))
s++;
}
if (*s != '\0')
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
errmsg("invalid input syntax for type %s: \"%s\"",
"path", str)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
path->closed = (!isopen);
/* prevent instability in unused pad bytes */
path->dummy = 0;
PG_RETURN_PATH_P(path);
}
Datum
path_out(PG_FUNCTION_ARGS)
{
PATH *path = PG_GETARG_PATH_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CSTRING(path_encode(path->closed ? PATH_CLOSED : PATH_OPEN, path->npts, path->p));
}
2003-05-13 20:03:08 +02:00
/*
* path_recv - converts external binary format to path
*
* External representation is closed flag (a boolean byte), int32 number
* of points, and the points.
*/
Datum
path_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
PATH *path;
int closed;
int32 npts;
int32 i;
int size;
closed = pq_getmsgbyte(buf);
npts = pq_getmsgint(buf, sizeof(int32));
if (npts <= 0 || npts >= (int32) ((INT_MAX - offsetof(PATH, p)) / sizeof(Point)))
ereport(ERROR,
(errcode(ERRCODE_INVALID_BINARY_REPRESENTATION),
errmsg("invalid number of points in external \"path\" value")));
2003-05-13 20:03:08 +02:00
2017-06-21 20:39:04 +02:00
size = offsetof(PATH, p) + sizeof(path->p[0]) * npts;
2003-05-13 20:03:08 +02:00
path = (PATH *) palloc(size);
SET_VARSIZE(path, size);
2003-05-13 20:03:08 +02:00
path->npts = npts;
path->closed = (closed ? 1 : 0);
/* prevent instability in unused pad bytes */
path->dummy = 0;
2003-05-13 20:03:08 +02:00
for (i = 0; i < npts; i++)
{
path->p[i].x = pq_getmsgfloat8(buf);
path->p[i].y = pq_getmsgfloat8(buf);
}
PG_RETURN_PATH_P(path);
}
/*
* path_send - converts path to binary format
*/
Datum
path_send(PG_FUNCTION_ARGS)
{
PATH *path = PG_GETARG_PATH_P(0);
StringInfoData buf;
int32 i;
pq_begintypsend(&buf);
pq_sendbyte(&buf, path->closed ? 1 : 0);
pq_sendint32(&buf, path->npts);
2003-05-13 20:03:08 +02:00
for (i = 0; i < path->npts; i++)
{
pq_sendfloat8(&buf, path->p[i].x);
pq_sendfloat8(&buf, path->p[i].y);
}
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}
/*----------------------------------------------------------
* Relational operators.
* These are based on the path cardinality,
* as stupid as that sounds.
*
* Better relops and access methods coming soon.
*---------------------------------------------------------*/
Datum
path_n_lt(PG_FUNCTION_ARGS)
{
PATH *p1 = PG_GETARG_PATH_P(0);
PATH *p2 = PG_GETARG_PATH_P(1);
PG_RETURN_BOOL(p1->npts < p2->npts);
}
Datum
path_n_gt(PG_FUNCTION_ARGS)
{
PATH *p1 = PG_GETARG_PATH_P(0);
PATH *p2 = PG_GETARG_PATH_P(1);
PG_RETURN_BOOL(p1->npts > p2->npts);
}
Datum
path_n_eq(PG_FUNCTION_ARGS)
{
PATH *p1 = PG_GETARG_PATH_P(0);
PATH *p2 = PG_GETARG_PATH_P(1);
PG_RETURN_BOOL(p1->npts == p2->npts);
}
Datum
path_n_le(PG_FUNCTION_ARGS)
{
PATH *p1 = PG_GETARG_PATH_P(0);
PATH *p2 = PG_GETARG_PATH_P(1);
PG_RETURN_BOOL(p1->npts <= p2->npts);
}
Datum
path_n_ge(PG_FUNCTION_ARGS)
{
PATH *p1 = PG_GETARG_PATH_P(0);
PATH *p2 = PG_GETARG_PATH_P(1);
PG_RETURN_BOOL(p1->npts >= p2->npts);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*----------------------------------------------------------
* Conversion operators.
*---------------------------------------------------------*/
Datum
path_isclosed(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_BOOL(path->closed);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
path_isopen(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
2001-03-22 05:01:46 +01:00
PG_RETURN_BOOL(!path->closed);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
path_npoints(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P(0);
PG_RETURN_INT32(path->npts);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
path_close(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P_COPY(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
path->closed = true;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_PATH_P(path);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
path_open(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P_COPY(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
path->closed = false;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_PATH_P(path);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* path_inter -
* Does p1 intersect p2 at any point?
* Use bounding boxes for a quick (O(n)) check, then do a
* O(n^2) iterative edge check.
*/
Datum
path_inter(PG_FUNCTION_ARGS)
{
PATH *p1 = PG_GETARG_PATH_P(0);
PATH *p2 = PG_GETARG_PATH_P(1);
BOX b1,
b2;
int i,
j;
LSEG seg1,
seg2;
Assert(p1->npts > 0 && p2->npts > 0);
b1.high.x = b1.low.x = p1->p[0].x;
b1.high.y = b1.low.y = p1->p[0].y;
for (i = 1; i < p1->npts; i++)
{
b1.high.x = float8_max(p1->p[i].x, b1.high.x);
b1.high.y = float8_max(p1->p[i].y, b1.high.y);
b1.low.x = float8_min(p1->p[i].x, b1.low.x);
b1.low.y = float8_min(p1->p[i].y, b1.low.y);
}
b2.high.x = b2.low.x = p2->p[0].x;
b2.high.y = b2.low.y = p2->p[0].y;
for (i = 1; i < p2->npts; i++)
{
b2.high.x = float8_max(p2->p[i].x, b2.high.x);
b2.high.y = float8_max(p2->p[i].y, b2.high.y);
b2.low.x = float8_min(p2->p[i].x, b2.low.x);
b2.low.y = float8_min(p2->p[i].y, b2.low.y);
}
if (!box_ov(&b1, &b2))
PG_RETURN_BOOL(false);
/* pairwise check lseg intersections */
for (i = 0; i < p1->npts; i++)
{
int iprev;
if (i > 0)
iprev = i - 1;
else
{
if (!p1->closed)
continue;
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
iprev = p1->npts - 1; /* include the closure segment */
}
for (j = 0; j < p2->npts; j++)
{
int jprev;
if (j > 0)
jprev = j - 1;
else
{
if (!p2->closed)
continue;
jprev = p2->npts - 1; /* include the closure segment */
}
statlseg_construct(&seg1, &p1->p[iprev], &p1->p[i]);
statlseg_construct(&seg2, &p2->p[jprev], &p2->p[j]);
if (lseg_interpt_lseg(NULL, &seg1, &seg2))
PG_RETURN_BOOL(true);
}
}
/* if we dropped through, no two segs intersected */
PG_RETURN_BOOL(false);
}
/* path_distance()
* This essentially does a cartesian product of the lsegs in the
* two paths, and finds the min distance between any two lsegs
*/
Datum
path_distance(PG_FUNCTION_ARGS)
{
PATH *p1 = PG_GETARG_PATH_P(0);
PATH *p2 = PG_GETARG_PATH_P(1);
float8 min = 0.0; /* initialize to keep compiler quiet */
bool have_min = false;
float8 tmp;
int i,
j;
LSEG seg1,
seg2;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < p1->npts; i++)
{
int iprev;
if (i > 0)
iprev = i - 1;
else
{
if (!p1->closed)
continue;
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
iprev = p1->npts - 1; /* include the closure segment */
}
for (j = 0; j < p2->npts; j++)
{
int jprev;
if (j > 0)
jprev = j - 1;
else
{
if (!p2->closed)
continue;
jprev = p2->npts - 1; /* include the closure segment */
}
statlseg_construct(&seg1, &p1->p[iprev], &p1->p[i]);
statlseg_construct(&seg2, &p2->p[jprev], &p2->p[j]);
tmp = lseg_closept_lseg(NULL, &seg1, &seg2);
if (!have_min || float8_lt(tmp, min))
{
min = tmp;
have_min = true;
}
1997-09-05 20:13:45 +02:00
}
}
2001-03-22 05:01:46 +01:00
if (!have_min)
PG_RETURN_NULL();
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_FLOAT8(min);
}
/*----------------------------------------------------------
* "Arithmetic" operations.
*---------------------------------------------------------*/
Datum
path_length(PG_FUNCTION_ARGS)
{
PATH *path = PG_GETARG_PATH_P(0);
float8 result = 0.0;
int i;
for (i = 0; i < path->npts; i++)
{
int iprev;
if (i > 0)
iprev = i - 1;
else
{
if (!path->closed)
continue;
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
iprev = path->npts - 1; /* include the closure segment */
}
result = float8_pl(result, point_dt(&path->p[iprev], &path->p[i]));
}
PG_RETURN_FLOAT8(result);
}
/***********************************************************************
**
** Routines for 2D points.
**
***********************************************************************/
/*----------------------------------------------------------
* String to point, point to string conversion.
* External format:
* "(x,y)"
* "x,y"
*---------------------------------------------------------*/
Datum
point_in(PG_FUNCTION_ARGS)
{
char *str = PG_GETARG_CSTRING(0);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
Point *point = (Point *) palloc(sizeof(Point));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
pair_decode(str, &point->x, &point->y, NULL, "point", str);
PG_RETURN_POINT_P(point);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
point_out(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CSTRING(path_encode(PATH_NONE, 1, pt));
}
/*
* point_recv - converts external binary format to point
*/
Datum
point_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
Point *point;
point = (Point *) palloc(sizeof(Point));
point->x = pq_getmsgfloat8(buf);
point->y = pq_getmsgfloat8(buf);
PG_RETURN_POINT_P(point);
}
/*
* point_send - converts point to binary format
*/
Datum
point_send(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
StringInfoData buf;
pq_begintypsend(&buf);
pq_sendfloat8(&buf, pt->x);
pq_sendfloat8(&buf, pt->y);
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}
/*
* Initialize a point
*/
static inline void
point_construct(Point *result, float8 x, float8 y)
{
result->x = x;
result->y = y;
}
/*----------------------------------------------------------
* Relational operators for Points.
* Since we do have a sense of coordinates being
* "equal" to a given accuracy (point_vert, point_horiz),
* the other ops must preserve that sense. This means
* that results may, strictly speaking, be a lie (unless
* EPSILON = 0.0).
*---------------------------------------------------------*/
Datum
point_left(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(FPlt(pt1->x, pt2->x));
}
Datum
point_right(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(FPgt(pt1->x, pt2->x));
}
Datum
point_above(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(FPgt(pt1->y, pt2->y));
}
Datum
point_below(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(FPlt(pt1->y, pt2->y));
}
Datum
point_vert(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(FPeq(pt1->x, pt2->x));
}
Datum
point_horiz(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(FPeq(pt1->y, pt2->y));
}
Datum
point_eq(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(point_eq_point(pt1, pt2));
}
Datum
point_ne(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(!point_eq_point(pt1, pt2));
}
/*
* Check whether the two points are the same
*
* We consider NaNs coordinates to be equal to each other to let those points
* to be found.
*/
static inline bool
point_eq_point(Point *pt1, Point *pt2)
{
return ((FPeq(pt1->x, pt2->x) && FPeq(pt1->y, pt2->y)) ||
(float8_eq(pt1->x, pt2->x) && float8_eq(pt1->y, pt2->y)));
}
/*----------------------------------------------------------
* "Arithmetic" operators on points.
*---------------------------------------------------------*/
Datum
point_distance(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_FLOAT8(point_dt(pt1, pt2));
}
static inline float8
point_dt(Point *pt1, Point *pt2)
{
return HYPOT(float8_mi(pt1->x, pt2->x), float8_mi(pt1->y, pt2->y));
}
Datum
point_slope(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
PG_RETURN_FLOAT8(point_sl(pt1, pt2));
}
/*
* Return slope of two points
*
* Note that this function returns DBL_MAX when the points are the same.
*/
static inline float8
point_sl(Point *pt1, Point *pt2)
{
if (FPeq(pt1->x, pt2->x))
return DBL_MAX;
if (FPeq(pt1->y, pt2->y))
return 0.0;
return float8_div(float8_mi(pt1->y, pt2->y), float8_mi(pt1->x, pt2->x));
}
/*
* Return inverse slope of two points
*
* Note that this function returns 0.0 when the points are the same.
*/
static inline float8
point_invsl(Point *pt1, Point *pt2)
{
if (FPeq(pt1->x, pt2->x))
return 0.0;
if (FPeq(pt1->y, pt2->y))
return DBL_MAX;
return float8_div(float8_mi(pt1->x, pt2->x), float8_mi(pt2->y, pt1->y));
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/***********************************************************************
**
** Routines for 2D line segments.
**
***********************************************************************/
/*----------------------------------------------------------
* String to lseg, lseg to string conversion.
* External forms: "[(x1, y1), (x2, y2)]"
* "(x1, y1), (x2, y2)"
* "x1, y1, x2, y2"
* closed form ok "((x1, y1), (x2, y2))"
* (old form) "(x1, y1, x2, y2)"
*---------------------------------------------------------*/
Datum
lseg_in(PG_FUNCTION_ARGS)
{
char *str = PG_GETARG_CSTRING(0);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
LSEG *lseg = (LSEG *) palloc(sizeof(LSEG));
bool isopen;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
path_decode(str, true, 2, &lseg->p[0], &isopen, NULL, "lseg", str);
PG_RETURN_LSEG_P(lseg);
}
Datum
lseg_out(PG_FUNCTION_ARGS)
{
LSEG *ls = PG_GETARG_LSEG_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CSTRING(path_encode(PATH_OPEN, 2, &ls->p[0]));
}
2003-05-13 20:03:08 +02:00
/*
* lseg_recv - converts external binary format to lseg
*/
Datum
lseg_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
LSEG *lseg;
lseg = (LSEG *) palloc(sizeof(LSEG));
lseg->p[0].x = pq_getmsgfloat8(buf);
lseg->p[0].y = pq_getmsgfloat8(buf);
lseg->p[1].x = pq_getmsgfloat8(buf);
lseg->p[1].y = pq_getmsgfloat8(buf);
PG_RETURN_LSEG_P(lseg);
}
/*
* lseg_send - converts lseg to binary format
*/
Datum
lseg_send(PG_FUNCTION_ARGS)
{
LSEG *ls = PG_GETARG_LSEG_P(0);
StringInfoData buf;
pq_begintypsend(&buf);
pq_sendfloat8(&buf, ls->p[0].x);
pq_sendfloat8(&buf, ls->p[0].y);
pq_sendfloat8(&buf, ls->p[1].x);
pq_sendfloat8(&buf, ls->p[1].y);
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}
/* lseg_construct -
* form a LSEG from two Points.
*/
Datum
lseg_construct(PG_FUNCTION_ARGS)
{
Point *pt1 = PG_GETARG_POINT_P(0);
Point *pt2 = PG_GETARG_POINT_P(1);
LSEG *result = (LSEG *) palloc(sizeof(LSEG));
statlseg_construct(result, pt1, pt2);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_LSEG_P(result);
}
/* like lseg_construct, but assume space already allocated */
static inline void
statlseg_construct(LSEG *lseg, Point *pt1, Point *pt2)
{
lseg->p[0].x = pt1->x;
lseg->p[0].y = pt1->y;
lseg->p[1].x = pt2->x;
lseg->p[1].y = pt2->y;
}
/*
* Return slope of the line segment
*/
static inline float8
lseg_sl(LSEG *lseg)
{
return point_sl(&lseg->p[0], &lseg->p[1]);
}
/*
* Return inverse slope of the line segment
*/
static inline float8
lseg_invsl(LSEG *lseg)
{
return point_invsl(&lseg->p[0], &lseg->p[1]);
}
Datum
lseg_length(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
PG_RETURN_FLOAT8(point_dt(&lseg->p[0], &lseg->p[1]));
}
/*----------------------------------------------------------
* Relative position routines.
*---------------------------------------------------------*/
/*
** find intersection of the two lines, and see if it falls on
** both segments.
*/
Datum
lseg_intersect(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(lseg_interpt_lseg(NULL, l1, l2));
}
Datum
lseg_parallel(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(FPeq(lseg_sl(l1), lseg_sl(l2)));
}
/*
* Determine if two line segments are perpendicular.
*/
Datum
lseg_perp(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(FPeq(lseg_sl(l1), lseg_invsl(l2)));
}
Datum
lseg_vertical(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
PG_RETURN_BOOL(FPeq(lseg->p[0].x, lseg->p[1].x));
}
Datum
lseg_horizontal(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
PG_RETURN_BOOL(FPeq(lseg->p[0].y, lseg->p[1].y));
}
Datum
lseg_eq(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(point_eq_point(&l1->p[0], &l2->p[0]) &&
point_eq_point(&l1->p[1], &l2->p[1]));
}
Datum
lseg_ne(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(!point_eq_point(&l1->p[0], &l2->p[0]) ||
!point_eq_point(&l1->p[1], &l2->p[1]));
}
Datum
lseg_lt(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(FPlt(point_dt(&l1->p[0], &l1->p[1]),
point_dt(&l2->p[0], &l2->p[1])));
}
Datum
lseg_le(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(FPle(point_dt(&l1->p[0], &l1->p[1]),
point_dt(&l2->p[0], &l2->p[1])));
}
Datum
lseg_gt(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(FPgt(point_dt(&l1->p[0], &l1->p[1]),
point_dt(&l2->p[0], &l2->p[1])));
}
Datum
lseg_ge(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(FPge(point_dt(&l1->p[0], &l1->p[1]),
point_dt(&l2->p[0], &l2->p[1])));
}
/*----------------------------------------------------------
* Line arithmetic routines.
*---------------------------------------------------------*/
/* lseg_distance -
* If two segments don't intersect, then the closest
* point will be from one of the endpoints to the other
* segment.
*/
Datum
lseg_distance(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_FLOAT8(lseg_closept_lseg(NULL, l1, l2));
}
Datum
lseg_center(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
Point *result;
result = (Point *) palloc(sizeof(Point));
result->x = float8_div(float8_pl(lseg->p[0].x, lseg->p[1].x), 2.0);
result->y = float8_div(float8_pl(lseg->p[0].y, lseg->p[1].y), 2.0);
PG_RETURN_POINT_P(result);
}
/*
* Return whether the two segments intersect. If *result is not NULL,
* it is set to the intersection point.
*
* This function is almost perfectly symmetric, even though it doesn't look
* like it. See lseg_interpt_line() for the other half of it.
*/
static bool
lseg_interpt_lseg(Point *result, LSEG *l1, LSEG *l2)
{
Point interpt;
LINE tmp;
line_construct(&tmp, &l2->p[0], lseg_sl(l2));
if (!lseg_interpt_line(&interpt, l1, &tmp))
return false;
2001-03-22 05:01:46 +01:00
/*
* If the line intersection point isn't within l2, there is no valid
* segment intersection point at all.
*/
if (!lseg_contain_point(l2, &interpt))
return false;
2001-03-22 05:01:46 +01:00
if (result != NULL)
*result = interpt;
return true;
}
Datum
lseg_interpt(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
Point *result;
2010-02-26 03:01:40 +01:00
result = (Point *) palloc(sizeof(Point));
if (!lseg_interpt_lseg(result, l1, l2))
PG_RETURN_NULL();
PG_RETURN_POINT_P(result);
}
/***********************************************************************
**
** Routines for position comparisons of differently-typed
** 2D objects.
**
***********************************************************************/
/*---------------------------------------------------------------------
* dist_
* Minimum distance from one object to another.
*-------------------------------------------------------------------*/
/*
* Distance from a point to a line
*/
Datum
dist_pl(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
LINE *line = PG_GETARG_LINE_P(1);
PG_RETURN_FLOAT8(line_closept_point(NULL, line, pt));
}
/*
* Distance from a line to a point
*/
Datum
dist_lp(PG_FUNCTION_ARGS)
{
LINE *line = PG_GETARG_LINE_P(0);
Point *pt = PG_GETARG_POINT_P(1);
PG_RETURN_FLOAT8(line_closept_point(NULL, line, pt));
}
/*
* Distance from a point to a lseg
*/
Datum
dist_ps(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
LSEG *lseg = PG_GETARG_LSEG_P(1);
PG_RETURN_FLOAT8(lseg_closept_point(NULL, lseg, pt));
}
/*
* Distance from a lseg to a point
*/
Datum
dist_sp(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
Point *pt = PG_GETARG_POINT_P(1);
PG_RETURN_FLOAT8(lseg_closept_point(NULL, lseg, pt));
}
static float8
dist_ppath_internal(Point *pt, PATH *path)
{
float8 result = 0.0; /* keep compiler quiet */
bool have_min = false;
float8 tmp;
int i;
LSEG lseg;
Assert(path->npts > 0);
/*
* The distance from a point to a path is the smallest distance from the
* point to any of its constituent segments.
*/
for (i = 0; i < path->npts; i++)
{
int iprev;
if (i > 0)
iprev = i - 1;
else
{
if (!path->closed)
continue;
iprev = path->npts - 1; /* Include the closure segment */
}
statlseg_construct(&lseg, &path->p[iprev], &path->p[i]);
tmp = lseg_closept_point(NULL, &lseg, pt);
if (!have_min || float8_lt(tmp, result))
{
result = tmp;
have_min = true;
}
}
return result;
}
/*
* Distance from a point to a path
*/
Datum
dist_ppath(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
PATH *path = PG_GETARG_PATH_P(1);
PG_RETURN_FLOAT8(dist_ppath_internal(pt, path));
}
/*
* Distance from a path to a point
*/
Datum
dist_pathp(PG_FUNCTION_ARGS)
{
PATH *path = PG_GETARG_PATH_P(0);
Point *pt = PG_GETARG_POINT_P(1);
PG_RETURN_FLOAT8(dist_ppath_internal(pt, path));
}
/*
* Distance from a point to a box
*/
Datum
dist_pb(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
BOX *box = PG_GETARG_BOX_P(1);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_FLOAT8(box_closept_point(NULL, box, pt));
}
/*
* Distance from a box to a point
*/
Datum
dist_bp(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
Point *pt = PG_GETARG_POINT_P(1);
PG_RETURN_FLOAT8(box_closept_point(NULL, box, pt));
}
/*
* Distance from a lseg to a line
*/
Datum
dist_sl(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
LINE *line = PG_GETARG_LINE_P(1);
PG_RETURN_FLOAT8(lseg_closept_line(NULL, lseg, line));
}
/*
* Distance from a line to a lseg
*/
Datum
dist_ls(PG_FUNCTION_ARGS)
{
LINE *line = PG_GETARG_LINE_P(0);
LSEG *lseg = PG_GETARG_LSEG_P(1);
PG_RETURN_FLOAT8(lseg_closept_line(NULL, lseg, line));
}
/*
* Distance from a lseg to a box
*/
Datum
dist_sb(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
BOX *box = PG_GETARG_BOX_P(1);
PG_RETURN_FLOAT8(box_closept_lseg(NULL, box, lseg));
}
/*
* Distance from a box to a lseg
*/
Datum
dist_bs(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
LSEG *lseg = PG_GETARG_LSEG_P(1);
PG_RETURN_FLOAT8(box_closept_lseg(NULL, box, lseg));
}
/*
* Distance from a line to a box
*/
Datum
dist_lb(PG_FUNCTION_ARGS)
{
#ifdef NOT_USED
LINE *line = PG_GETARG_LINE_P(0);
BOX *box = PG_GETARG_BOX_P(1);
#endif
/* need to think about this one for a while */
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("function \"dist_lb\" not implemented")));
PG_RETURN_NULL();
}
/*
* Distance from a box to a line
*/
Datum
dist_bl(PG_FUNCTION_ARGS)
{
#ifdef NOT_USED
BOX *box = PG_GETARG_BOX_P(0);
LINE *line = PG_GETARG_LINE_P(1);
#endif
/* need to think about this one for a while */
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("function \"dist_bl\" not implemented")));
PG_RETURN_NULL();
}
static float8
dist_cpoly_internal(CIRCLE *circle, POLYGON *poly)
{
float8 result;
/* calculate distance to center, and subtract radius */
result = float8_mi(dist_ppoly_internal(&circle->center, poly),
circle->radius);
if (result < 0.0)
result = 0.0;
return result;
}
/*
* Distance from a circle to a polygon
*/
Datum
dist_cpoly(PG_FUNCTION_ARGS)
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
POLYGON *poly = PG_GETARG_POLYGON_P(1);
PG_RETURN_FLOAT8(dist_cpoly_internal(circle, poly));
}
/*
* Distance from a polygon to a circle
*/
Datum
dist_polyc(PG_FUNCTION_ARGS)
{
POLYGON *poly = PG_GETARG_POLYGON_P(0);
CIRCLE *circle = PG_GETARG_CIRCLE_P(1);
PG_RETURN_FLOAT8(dist_cpoly_internal(circle, poly));
}
/*
* Distance from a point to a polygon
*/
Datum
dist_ppoly(PG_FUNCTION_ARGS)
{
Point *point = PG_GETARG_POINT_P(0);
POLYGON *poly = PG_GETARG_POLYGON_P(1);
PG_RETURN_FLOAT8(dist_ppoly_internal(point, poly));
}
Datum
dist_polyp(PG_FUNCTION_ARGS)
{
POLYGON *poly = PG_GETARG_POLYGON_P(0);
Point *point = PG_GETARG_POINT_P(1);
PG_RETURN_FLOAT8(dist_ppoly_internal(point, poly));
}
static float8
dist_ppoly_internal(Point *pt, POLYGON *poly)
{
float8 result;
float8 d;
int i;
LSEG seg;
if (point_inside(pt, poly->npts, poly->p) != 0)
return 0.0;
/* initialize distance with segment between first and last points */
seg.p[0].x = poly->p[0].x;
seg.p[0].y = poly->p[0].y;
seg.p[1].x = poly->p[poly->npts - 1].x;
seg.p[1].y = poly->p[poly->npts - 1].y;
result = lseg_closept_point(NULL, &seg, pt);
/* check distances for other segments */
for (i = 0; i < poly->npts - 1; i++)
{
seg.p[0].x = poly->p[i].x;
seg.p[0].y = poly->p[i].y;
seg.p[1].x = poly->p[i + 1].x;
seg.p[1].y = poly->p[i + 1].y;
d = lseg_closept_point(NULL, &seg, pt);
if (float8_lt(d, result))
result = d;
}
return result;
}
/*---------------------------------------------------------------------
* interpt_
* Intersection point of objects.
* We choose to ignore the "point" of intersection between
* lines and boxes, since there are typically two.
*-------------------------------------------------------------------*/
/*
* Return whether the line segment intersect with the line. If *result is not
* NULL, it is set to the intersection point.
*/
static bool
lseg_interpt_line(Point *result, LSEG *lseg, LINE *line)
{
Point interpt;
LINE tmp;
/*
* First, we promote the line segment to a line, because we know how to
* find the intersection point of two lines. If they don't have an
* intersection point, we are done.
*/
line_construct(&tmp, &lseg->p[0], lseg_sl(lseg));
if (!line_interpt_line(&interpt, &tmp, line))
return false;
/*
* Then, we check whether the intersection point is actually on the line
* segment.
*/
if (!lseg_contain_point(lseg, &interpt))
return false;
if (result != NULL)
{
/*
* If there is an intersection, then check explicitly for matching
* endpoints since there may be rounding effects with annoying LSB
* residue.
*/
if (point_eq_point(&lseg->p[0], &interpt))
*result = lseg->p[0];
else if (point_eq_point(&lseg->p[1], &interpt))
*result = lseg->p[1];
else
*result = interpt;
}
return true;
}
/*---------------------------------------------------------------------
* close_
* Point of closest proximity between objects.
*-------------------------------------------------------------------*/
/*
* If *result is not NULL, it is set to the intersection point of a
* perpendicular of the line through the point. Returns the distance
* of those two points.
*/
static float8
line_closept_point(Point *result, LINE *line, Point *point)
{
Point closept;
LINE tmp;
/*
* We drop a perpendicular to find the intersection point. Ordinarily we
* should always find it, but that can fail in the presence of NaN
* coordinates, and perhaps even from simple roundoff issues.
*/
line_construct(&tmp, point, line_invsl(line));
if (!line_interpt_line(&closept, &tmp, line))
{
if (result != NULL)
*result = *point;
return get_float8_nan();
}
if (result != NULL)
*result = closept;
return point_dt(&closept, point);
}
Datum
close_pl(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
LINE *line = PG_GETARG_LINE_P(1);
Point *result;
result = (Point *) palloc(sizeof(Point));
if (isnan(line_closept_point(result, line, pt)))
PG_RETURN_NULL();
PG_RETURN_POINT_P(result);
}
/*
* Closest point on line segment to specified point.
*
* If *result is not NULL, set it to the closest point on the line segment
* to the point. Returns the distance of the two points.
*/
static float8
lseg_closept_point(Point *result, LSEG *lseg, Point *pt)
{
Point closept;
LINE tmp;
/*
* To find the closest point, we draw a perpendicular line from the point
* to the line segment.
*/
line_construct(&tmp, pt, point_invsl(&lseg->p[0], &lseg->p[1]));
lseg_closept_line(&closept, lseg, &tmp);
if (result != NULL)
*result = closept;
return point_dt(&closept, pt);
}
Datum
close_ps(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
LSEG *lseg = PG_GETARG_LSEG_P(1);
Point *result;
result = (Point *) palloc(sizeof(Point));
if (isnan(lseg_closept_point(result, lseg, pt)))
PG_RETURN_NULL();
PG_RETURN_POINT_P(result);
}
/*
* Closest point on line segment to line segment
*/
static float8
lseg_closept_lseg(Point *result, LSEG *on_lseg, LSEG *to_lseg)
{
Point point;
float8 dist,
d;
/* First, we handle the case when the line segments are intersecting. */
if (lseg_interpt_lseg(result, on_lseg, to_lseg))
return 0.0;
/*
* Then, we find the closest points from the endpoints of the second line
* segment, and keep the closest one.
*/
dist = lseg_closept_point(result, on_lseg, &to_lseg->p[0]);
d = lseg_closept_point(&point, on_lseg, &to_lseg->p[1]);
if (float8_lt(d, dist))
{
dist = d;
if (result != NULL)
*result = point;
}
/* The closest point can still be one of the endpoints, so we test them. */
d = lseg_closept_point(NULL, to_lseg, &on_lseg->p[0]);
if (float8_lt(d, dist))
{
dist = d;
if (result != NULL)
*result = on_lseg->p[0];
}
d = lseg_closept_point(NULL, to_lseg, &on_lseg->p[1]);
if (float8_lt(d, dist))
{
dist = d;
if (result != NULL)
*result = on_lseg->p[1];
}
return dist;
}
Datum
close_lseg(PG_FUNCTION_ARGS)
{
LSEG *l1 = PG_GETARG_LSEG_P(0);
LSEG *l2 = PG_GETARG_LSEG_P(1);
Point *result;
if (lseg_sl(l1) == lseg_sl(l2))
PG_RETURN_NULL();
result = (Point *) palloc(sizeof(Point));
if (isnan(lseg_closept_lseg(result, l2, l1)))
PG_RETURN_NULL();
PG_RETURN_POINT_P(result);
}
/*
* Closest point on or in box to specified point.
*
* If *result is not NULL, set it to the closest point on the box to the
* given point, and return the distance of the two points.
*/
static float8
box_closept_point(Point *result, BOX *box, Point *pt)
{
float8 dist,
d;
Point point,
closept;
LSEG lseg;
if (box_contain_point(box, pt))
{
if (result != NULL)
*result = *pt;
return 0.0;
}
/* pairwise check lseg distances */
point.x = box->low.x;
point.y = box->high.y;
statlseg_construct(&lseg, &box->low, &point);
dist = lseg_closept_point(result, &lseg, pt);
statlseg_construct(&lseg, &box->high, &point);
d = lseg_closept_point(&closept, &lseg, pt);
if (float8_lt(d, dist))
{
dist = d;
if (result != NULL)
*result = closept;
}
point.x = box->high.x;
point.y = box->low.y;
statlseg_construct(&lseg, &box->low, &point);
d = lseg_closept_point(&closept, &lseg, pt);
if (float8_lt(d, dist))
{
dist = d;
if (result != NULL)
*result = closept;
}
statlseg_construct(&lseg, &box->high, &point);
d = lseg_closept_point(&closept, &lseg, pt);
if (float8_lt(d, dist))
{
dist = d;
if (result != NULL)
*result = closept;
}
return dist;
}
Datum
close_pb(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
BOX *box = PG_GETARG_BOX_P(1);
Point *result;
result = (Point *) palloc(sizeof(Point));
if (isnan(box_closept_point(result, box, pt)))
PG_RETURN_NULL();
PG_RETURN_POINT_P(result);
}
/* close_sl()
* Closest point on line to line segment.
*
* XXX THIS CODE IS WRONG
* The code is actually calculating the point on the line segment
* which is backwards from the routine naming convention.
* Copied code to new routine close_ls() but haven't fixed this one yet.
* - thomas 1998-01-31
*/
Datum
close_sl(PG_FUNCTION_ARGS)
{
#ifdef NOT_USED
LSEG *lseg = PG_GETARG_LSEG_P(0);
LINE *line = PG_GETARG_LINE_P(1);
Point *result;
float8 d1,
d2;
result = (Point *) palloc(sizeof(Point));
if (lseg_interpt_line(result, lseg, line))
PG_RETURN_POINT_P(result);
d1 = line_closept_point(NULL, line, &lseg->p[0]);
d2 = line_closept_point(NULL, line, &lseg->p[1]);
if (float8_lt(d1, d2))
*result = lseg->p[0];
else
*result = lseg->p[1];
PG_RETURN_POINT_P(result);
#endif
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("function \"close_sl\" not implemented")));
PG_RETURN_NULL();
}
/*
* Closest point on line segment to line.
*
* Return the distance between the line and the closest point of the line
* segment to the line. If *result is not NULL, set it to that point.
*
* NOTE: When the lines are parallel, endpoints of one of the line segment
* are FPeq(), in presence of NaN or Infinite coordinates, or perhaps =
* even because of simple roundoff issues, there may not be a single closest
* point. We are likely to set the result to the second endpoint in these
* cases.
*/
static float8
lseg_closept_line(Point *result, LSEG *lseg, LINE *line)
{
float8 dist1,
dist2;
if (lseg_interpt_line(result, lseg, line))
return 0.0;
dist1 = line_closept_point(NULL, line, &lseg->p[0]);
dist2 = line_closept_point(NULL, line, &lseg->p[1]);
if (dist1 < dist2)
{
if (result != NULL)
*result = lseg->p[0];
return dist1;
}
else
{
if (result != NULL)
*result = lseg->p[1];
return dist2;
}
}
Datum
close_ls(PG_FUNCTION_ARGS)
{
LINE *line = PG_GETARG_LINE_P(0);
LSEG *lseg = PG_GETARG_LSEG_P(1);
Point *result;
if (lseg_sl(lseg) == line_sl(line))
PG_RETURN_NULL();
result = (Point *) palloc(sizeof(Point));
if (isnan(lseg_closept_line(result, lseg, line)))
PG_RETURN_NULL();
PG_RETURN_POINT_P(result);
}
/*
* Closest point on or in box to line segment.
*
* Returns the distance between the closest point on or in the box to
* the line segment. If *result is not NULL, it is set to that point.
*/
static float8
box_closept_lseg(Point *result, BOX *box, LSEG *lseg)
{
float8 dist,
d;
Point point,
closept;
LSEG bseg;
if (box_interpt_lseg(result, box, lseg))
return 0.0;
/* pairwise check lseg distances */
point.x = box->low.x;
point.y = box->high.y;
statlseg_construct(&bseg, &box->low, &point);
dist = lseg_closept_lseg(result, &bseg, lseg);
statlseg_construct(&bseg, &box->high, &point);
d = lseg_closept_lseg(&closept, &bseg, lseg);
if (float8_lt(d, dist))
{
dist = d;
if (result != NULL)
*result = closept;
}
point.x = box->high.x;
point.y = box->low.y;
statlseg_construct(&bseg, &box->low, &point);
d = lseg_closept_lseg(&closept, &bseg, lseg);
if (float8_lt(d, dist))
{
dist = d;
if (result != NULL)
*result = closept;
}
statlseg_construct(&bseg, &box->high, &point);
d = lseg_closept_lseg(&closept, &bseg, lseg);
if (float8_lt(d, dist))
{
dist = d;
if (result != NULL)
*result = closept;
}
return dist;
}
Datum
close_sb(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
BOX *box = PG_GETARG_BOX_P(1);
Point *result;
result = (Point *) palloc(sizeof(Point));
if (isnan(box_closept_lseg(result, box, lseg)))
PG_RETURN_NULL();
PG_RETURN_POINT_P(result);
}
Datum
close_lb(PG_FUNCTION_ARGS)
{
#ifdef NOT_USED
LINE *line = PG_GETARG_LINE_P(0);
BOX *box = PG_GETARG_BOX_P(1);
#endif
/* think about this one for a while */
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("function \"close_lb\" not implemented")));
PG_RETURN_NULL();
}
/*---------------------------------------------------------------------
* on_
* Whether one object lies completely within another.
*-------------------------------------------------------------------*/
/*
* Does the point satisfy the equation?
*/
static bool
line_contain_point(LINE *line, Point *point)
{
return FPzero(float8_pl(float8_pl(float8_mul(line->A, point->x),
float8_mul(line->B, point->y)),
line->C));
}
Datum
on_pl(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
LINE *line = PG_GETARG_LINE_P(1);
PG_RETURN_BOOL(line_contain_point(line, pt));
}
/*
* Determine colinearity by detecting a triangle inequality.
* This algorithm seems to behave nicely even with lsb residues - tgl 1997-07-09
*/
static bool
lseg_contain_point(LSEG *lseg, Point *pt)
{
return FPeq(point_dt(pt, &lseg->p[0]) +
point_dt(pt, &lseg->p[1]),
point_dt(&lseg->p[0], &lseg->p[1]));
}
Datum
on_ps(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
LSEG *lseg = PG_GETARG_LSEG_P(1);
PG_RETURN_BOOL(lseg_contain_point(lseg, pt));
}
/*
* Check whether the point is in the box or on its border
*/
static bool
box_contain_point(BOX *box, Point *point)
{
return box->high.x >= point->x && box->low.x <= point->x &&
box->high.y >= point->y && box->low.y <= point->y;
}
Datum
on_pb(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
BOX *box = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(box_contain_point(box, pt));
}
2010-01-14 17:31:09 +01:00
Datum
2010-02-26 03:01:40 +01:00
box_contain_pt(PG_FUNCTION_ARGS)
2010-01-14 17:31:09 +01:00
{
BOX *box = PG_GETARG_BOX_P(0);
Point *pt = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(box_contain_point(box, pt));
2010-01-14 17:31:09 +01:00
}
/* on_ppath -
* Whether a point lies within (on) a polyline.
* If open, we have to (groan) check each segment.
* (uses same algorithm as for point intersecting segment - tgl 1997-07-09)
* If closed, we use the old O(n) ray method for point-in-polygon.
* The ray is horizontal, from pt out to the right.
* Each segment that crosses the ray counts as an
* intersection; note that an endpoint or edge may touch
* but not cross.
* (we can do p-in-p in lg(n), but it takes preprocessing)
*/
Datum
on_ppath(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
PATH *path = PG_GETARG_PATH_P(1);
int i,
n;
float8 a,
b;
1998-10-29 19:11:28 +01:00
/*-- OPEN --*/
if (!path->closed)
1998-10-29 19:11:28 +01:00
{
n = path->npts - 1;
a = point_dt(pt, &path->p[0]);
for (i = 0; i < n; i++)
{
b = point_dt(pt, &path->p[i + 1]);
if (FPeq(float8_pl(a, b), point_dt(&path->p[i], &path->p[i + 1])))
PG_RETURN_BOOL(true);
a = b;
}
PG_RETURN_BOOL(false);
}
1998-10-29 19:11:28 +01:00
/*-- CLOSED --*/
PG_RETURN_BOOL(point_inside(pt, path->npts, path->p) != 0);
}
/*
* Check whether the line segment is on the line or close enough
*
* It is, if both of its points are on the line or close enough.
*/
Datum
on_sl(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
LINE *line = PG_GETARG_LINE_P(1);
PG_RETURN_BOOL(line_contain_point(line, &lseg->p[0]) &&
line_contain_point(line, &lseg->p[1]));
}
/*
* Check whether the line segment is in the box or on its border
*
* It is, if both of its points are in the box or on its border.
*/
static bool
box_contain_lseg(BOX *box, LSEG *lseg)
{
return box_contain_point(box, &lseg->p[0]) &&
box_contain_point(box, &lseg->p[1]);
}
Datum
on_sb(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
BOX *box = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(box_contain_lseg(box, lseg));
}
/*---------------------------------------------------------------------
* inter_
* Whether one object intersects another.
*-------------------------------------------------------------------*/
Datum
inter_sl(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
LINE *line = PG_GETARG_LINE_P(1);
PG_RETURN_BOOL(lseg_interpt_line(NULL, lseg, line));
}
/*
* Do line segment and box intersect?
*
* Segment completely inside box counts as intersection.
* If you want only segments crossing box boundaries,
* try converting box to path first.
*
* This function also sets the *result to the closest point on the line
* segment to the center of the box when they overlap and the result is
* not NULL. It is somewhat arbitrary, but maybe the best we can do as
* there are typically two points they intersect.
*
* Optimize for non-intersection by checking for box intersection first.
* - thomas 1998-01-30
*/
static bool
box_interpt_lseg(Point *result, BOX *box, LSEG *lseg)
{
BOX lbox;
LSEG bseg;
Point point;
lbox.low.x = float8_min(lseg->p[0].x, lseg->p[1].x);
lbox.low.y = float8_min(lseg->p[0].y, lseg->p[1].y);
lbox.high.x = float8_max(lseg->p[0].x, lseg->p[1].x);
lbox.high.y = float8_max(lseg->p[0].y, lseg->p[1].y);
/* nothing close to overlap? then not going to intersect */
if (!box_ov(&lbox, box))
return false;
if (result != NULL)
{
box_cn(&point, box);
lseg_closept_point(result, lseg, &point);
}
/* an endpoint of segment is inside box? then clearly intersects */
if (box_contain_point(box, &lseg->p[0]) ||
box_contain_point(box, &lseg->p[1]))
return true;
/* pairwise check lseg intersections */
point.x = box->low.x;
point.y = box->high.y;
statlseg_construct(&bseg, &box->low, &point);
if (lseg_interpt_lseg(NULL, &bseg, lseg))
return true;
statlseg_construct(&bseg, &box->high, &point);
if (lseg_interpt_lseg(NULL, &bseg, lseg))
return true;
point.x = box->high.x;
point.y = box->low.y;
statlseg_construct(&bseg, &box->low, &point);
if (lseg_interpt_lseg(NULL, &bseg, lseg))
return true;
statlseg_construct(&bseg, &box->high, &point);
if (lseg_interpt_lseg(NULL, &bseg, lseg))
return true;
/* if we dropped through, no two segs intersected */
return false;
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
inter_sb(PG_FUNCTION_ARGS)
{
LSEG *lseg = PG_GETARG_LSEG_P(0);
BOX *box = PG_GETARG_BOX_P(1);
PG_RETURN_BOOL(box_interpt_lseg(NULL, box, lseg));
}
/* inter_lb()
* Do line and box intersect?
*/
Datum
inter_lb(PG_FUNCTION_ARGS)
{
LINE *line = PG_GETARG_LINE_P(0);
BOX *box = PG_GETARG_BOX_P(1);
LSEG bseg;
Point p1,
p2;
/* pairwise check lseg intersections */
p1.x = box->low.x;
p1.y = box->low.y;
p2.x = box->low.x;
p2.y = box->high.y;
statlseg_construct(&bseg, &p1, &p2);
if (lseg_interpt_line(NULL, &bseg, line))
PG_RETURN_BOOL(true);
p1.x = box->high.x;
p1.y = box->high.y;
statlseg_construct(&bseg, &p1, &p2);
if (lseg_interpt_line(NULL, &bseg, line))
PG_RETURN_BOOL(true);
p2.x = box->high.x;
p2.y = box->low.y;
statlseg_construct(&bseg, &p1, &p2);
if (lseg_interpt_line(NULL, &bseg, line))
PG_RETURN_BOOL(true);
p1.x = box->low.x;
p1.y = box->low.y;
statlseg_construct(&bseg, &p1, &p2);
if (lseg_interpt_line(NULL, &bseg, line))
PG_RETURN_BOOL(true);
/* if we dropped through, no intersection */
PG_RETURN_BOOL(false);
}
/*------------------------------------------------------------------
* The following routines define a data type and operator class for
* POLYGONS .... Part of which (the polygon's bounding box) is built on
* top of the BOX data type.
*
* make_bound_box - create the bounding box for the input polygon
*------------------------------------------------------------------*/
/*---------------------------------------------------------------------
* Make the smallest bounding box for the given polygon.
*---------------------------------------------------------------------*/
static void
make_bound_box(POLYGON *poly)
{
int i;
float8 x1,
y1,
x2,
y2;
Assert(poly->npts > 0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
x1 = x2 = poly->p[0].x;
y2 = y1 = poly->p[0].y;
for (i = 1; i < poly->npts; i++)
{
if (float8_lt(poly->p[i].x, x1))
x1 = poly->p[i].x;
if (float8_gt(poly->p[i].x, x2))
x2 = poly->p[i].x;
if (float8_lt(poly->p[i].y, y1))
y1 = poly->p[i].y;
if (float8_gt(poly->p[i].y, y2))
y2 = poly->p[i].y;
}
poly->boundbox.low.x = x1;
poly->boundbox.high.x = x2;
poly->boundbox.low.y = y1;
poly->boundbox.high.y = y2;
}
/*------------------------------------------------------------------
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
* poly_in - read in the polygon from a string specification
*
* External format:
* "((x0,y0),...,(xn,yn))"
* "x0,y0,...,xn,yn"
* also supports the older style "(x1,...,xn,y1,...yn)"
*------------------------------------------------------------------*/
Datum
poly_in(PG_FUNCTION_ARGS)
{
char *str = PG_GETARG_CSTRING(0);
POLYGON *poly;
int npts;
int size;
int base_size;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
bool isopen;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
if ((npts = pair_count(str, ',')) <= 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
errmsg("invalid input syntax for type %s: \"%s\"",
"polygon", str)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
base_size = sizeof(poly->p[0]) * npts;
2017-06-21 20:39:04 +02:00
size = offsetof(POLYGON, p) + base_size;
/* Check for integer overflow */
if (base_size / npts != sizeof(poly->p[0]) || size <= base_size)
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("too many points requested")));
poly = (POLYGON *) palloc0(size); /* zero any holes */
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
SET_VARSIZE(poly, size);
poly->npts = npts;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
path_decode(str, false, npts, &(poly->p[0]), &isopen, NULL, "polygon", str);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
make_bound_box(poly);
PG_RETURN_POLYGON_P(poly);
}
/*---------------------------------------------------------------
* poly_out - convert internal POLYGON representation to the
* character string format "((f8,f8),...,(f8,f8))"
*---------------------------------------------------------------*/
Datum
poly_out(PG_FUNCTION_ARGS)
{
2001-03-22 05:01:46 +01:00
POLYGON *poly = PG_GETARG_POLYGON_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CSTRING(path_encode(PATH_CLOSED, poly->npts, poly->p));
}
2003-05-13 20:03:08 +02:00
/*
* poly_recv - converts external binary format to polygon
*
* External representation is int32 number of points, and the points.
* We recompute the bounding box on read, instead of trusting it to
* be valid. (Checking it would take just as long, so may as well
* omit it from external representation.)
*/
Datum
poly_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
2003-08-04 02:43:34 +02:00
POLYGON *poly;
2003-05-13 20:03:08 +02:00
int32 npts;
int32 i;
int size;
npts = pq_getmsgint(buf, sizeof(int32));
if (npts <= 0 || npts >= (int32) ((INT_MAX - offsetof(POLYGON, p)) / sizeof(Point)))
ereport(ERROR,
(errcode(ERRCODE_INVALID_BINARY_REPRESENTATION),
errmsg("invalid number of points in external \"polygon\" value")));
2003-05-13 20:03:08 +02:00
2017-06-21 20:39:04 +02:00
size = offsetof(POLYGON, p) + sizeof(poly->p[0]) * npts;
2003-05-13 20:03:08 +02:00
poly = (POLYGON *) palloc0(size); /* zero any holes */
SET_VARSIZE(poly, size);
2003-05-13 20:03:08 +02:00
poly->npts = npts;
for (i = 0; i < npts; i++)
{
poly->p[i].x = pq_getmsgfloat8(buf);
poly->p[i].y = pq_getmsgfloat8(buf);
}
make_bound_box(poly);
PG_RETURN_POLYGON_P(poly);
}
/*
* poly_send - converts polygon to binary format
*/
Datum
poly_send(PG_FUNCTION_ARGS)
{
2003-08-04 02:43:34 +02:00
POLYGON *poly = PG_GETARG_POLYGON_P(0);
2003-05-13 20:03:08 +02:00
StringInfoData buf;
int32 i;
pq_begintypsend(&buf);
pq_sendint32(&buf, poly->npts);
2003-05-13 20:03:08 +02:00
for (i = 0; i < poly->npts; i++)
{
pq_sendfloat8(&buf, poly->p[i].x);
pq_sendfloat8(&buf, poly->p[i].y);
}
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}
/*-------------------------------------------------------
* Is polygon A strictly left of polygon B? i.e. is
* the right most point of A left of the left most point
* of B?
*-------------------------------------------------------*/
Datum
poly_left(PG_FUNCTION_ARGS)
{
2001-03-22 05:01:46 +01:00
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
result = polya->boundbox.high.x < polyb->boundbox.low.x;
2001-03-22 05:01:46 +01:00
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
2001-03-22 05:01:46 +01:00
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*-------------------------------------------------------
* Is polygon A overlapping or left of polygon B? i.e. is
* the right most point of A at or left of the right most point
* of B?
*-------------------------------------------------------*/
Datum
poly_overleft(PG_FUNCTION_ARGS)
{
2001-03-22 05:01:46 +01:00
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
result = polya->boundbox.high.x <= polyb->boundbox.high.x;
2001-03-22 05:01:46 +01:00
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
2001-03-22 05:01:46 +01:00
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*-------------------------------------------------------
* Is polygon A strictly right of polygon B? i.e. is
* the left most point of A right of the right most point
* of B?
*-------------------------------------------------------*/
Datum
poly_right(PG_FUNCTION_ARGS)
{
2001-03-22 05:01:46 +01:00
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
result = polya->boundbox.low.x > polyb->boundbox.high.x;
2001-03-22 05:01:46 +01:00
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
2001-03-22 05:01:46 +01:00
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*-------------------------------------------------------
* Is polygon A overlapping or right of polygon B? i.e. is
* the left most point of A at or right of the left most point
* of B?
*-------------------------------------------------------*/
Datum
poly_overright(PG_FUNCTION_ARGS)
{
2001-03-22 05:01:46 +01:00
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
result = polya->boundbox.low.x >= polyb->boundbox.low.x;
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*-------------------------------------------------------
* Is polygon A strictly below polygon B? i.e. is
* the upper most point of A below the lower most point
* of B?
*-------------------------------------------------------*/
Datum
poly_below(PG_FUNCTION_ARGS)
{
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
result = polya->boundbox.high.y < polyb->boundbox.low.y;
2001-03-22 05:01:46 +01:00
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
2001-03-22 05:01:46 +01:00
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*-------------------------------------------------------
* Is polygon A overlapping or below polygon B? i.e. is
* the upper most point of A at or below the upper most point
* of B?
*-------------------------------------------------------*/
Datum
poly_overbelow(PG_FUNCTION_ARGS)
{
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
result = polya->boundbox.high.y <= polyb->boundbox.high.y;
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*-------------------------------------------------------
* Is polygon A strictly above polygon B? i.e. is
* the lower most point of A above the upper most point
* of B?
*-------------------------------------------------------*/
Datum
poly_above(PG_FUNCTION_ARGS)
{
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
result = polya->boundbox.low.y > polyb->boundbox.high.y;
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*-------------------------------------------------------
* Is polygon A overlapping or above polygon B? i.e. is
* the lower most point of A at or above the lower most point
* of B?
*-------------------------------------------------------*/
Datum
poly_overabove(PG_FUNCTION_ARGS)
{
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
result = polya->boundbox.low.y >= polyb->boundbox.low.y;
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*-------------------------------------------------------
* Is polygon A the same as polygon B? i.e. are all the
* points the same?
* Check all points for matches in both forward and reverse
* direction since polygons are non-directional and are
* closed shapes.
*-------------------------------------------------------*/
Datum
poly_same(PG_FUNCTION_ARGS)
{
2001-03-22 05:01:46 +01:00
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
if (polya->npts != polyb->npts)
result = false;
else
result = plist_same(polya->npts, polya->p, polyb->p);
2001-03-22 05:01:46 +01:00
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
2001-03-22 05:01:46 +01:00
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_BOOL(result);
}
/*-----------------------------------------------------------------
2010-02-26 03:01:40 +01:00
* Determine if polygon A overlaps polygon B
*-----------------------------------------------------------------*/
Datum
poly_overlap(PG_FUNCTION_ARGS)
{
2001-03-22 05:01:46 +01:00
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
Assert(polya->npts > 0 && polyb->npts > 0);
/* Quick check by bounding box */
result = box_ov(&polya->boundbox, &polyb->boundbox);
/*
2010-02-26 03:01:40 +01:00
* Brute-force algorithm - try to find intersected edges, if so then
* polygons are overlapped else check is one polygon inside other or not
* by testing single point of them.
*/
if (result)
{
2010-02-26 03:01:40 +01:00
int ia,
ib;
LSEG sa,
sb;
/* Init first of polya's edge with last point */
sa.p[0] = polya->p[polya->npts - 1];
result = false;
for (ia = 0; ia < polya->npts && !result; ia++)
{
2010-02-26 03:01:40 +01:00
/* Second point of polya's edge is a current one */
sa.p[1] = polya->p[ia];
/* Init first of polyb's edge with last point */
sb.p[0] = polyb->p[polyb->npts - 1];
for (ib = 0; ib < polyb->npts && !result; ib++)
{
sb.p[1] = polyb->p[ib];
result = lseg_interpt_lseg(NULL, &sa, &sb);
sb.p[0] = sb.p[1];
}
2010-02-26 03:01:40 +01:00
/*
* move current endpoint to the first point of next edge
*/
sa.p[0] = sa.p[1];
}
if (!result)
{
result = (point_inside(polya->p, polyb->npts, polyb->p) ||
2010-02-26 03:01:40 +01:00
point_inside(polyb->p, polya->npts, polya->p));
}
}
2001-03-22 05:01:46 +01:00
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
2001-03-22 05:01:46 +01:00
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*
* Tests special kind of segment for in/out of polygon.
* Special kind means:
2010-02-26 03:01:40 +01:00
* - point a should be on segment s
* - segment (a,b) should not be contained by s
* Returns true if:
2010-02-26 03:01:40 +01:00
* - segment (a,b) is collinear to s and (a,b) is in polygon
* - segment (a,b) s not collinear to s. Note: that doesn't
* mean that segment is in polygon!
*/
static bool
touched_lseg_inside_poly(Point *a, Point *b, LSEG *s, POLYGON *poly, int start)
{
/* point a is on s, b is not */
2010-02-26 03:01:40 +01:00
LSEG t;
t.p[0] = *a;
t.p[1] = *b;
2010-02-26 03:01:40 +01:00
if (point_eq_point(a, s->p))
{
if (lseg_contain_point(&t, s->p + 1))
2010-02-26 03:01:40 +01:00
return lseg_inside_poly(b, s->p + 1, poly, start);
}
else if (point_eq_point(a, s->p + 1))
{
if (lseg_contain_point(&t, s->p))
return lseg_inside_poly(b, s->p, poly, start);
}
else if (lseg_contain_point(&t, s->p))
{
return lseg_inside_poly(b, s->p, poly, start);
}
else if (lseg_contain_point(&t, s->p + 1))
{
2010-02-26 03:01:40 +01:00
return lseg_inside_poly(b, s->p + 1, poly, start);
}
2010-02-26 03:01:40 +01:00
return true; /* may be not true, but that will check later */
}
/*
* Returns true if segment (a,b) is in polygon, option
2010-02-26 03:01:40 +01:00
* start is used for optimization - function checks
* polygon's edges starting from start
*/
static bool
lseg_inside_poly(Point *a, Point *b, POLYGON *poly, int start)
{
2010-02-26 03:01:40 +01:00
LSEG s,
t;
int i;
bool res = true,
intersection = false;
t.p[0] = *a;
t.p[1] = *b;
2010-02-26 03:01:40 +01:00
s.p[0] = poly->p[(start == 0) ? (poly->npts - 1) : (start - 1)];
for (i = start; i < poly->npts && res; i++)
{
Point interpt;
CHECK_FOR_INTERRUPTS();
s.p[1] = poly->p[i];
if (lseg_contain_point(&s, t.p))
{
if (lseg_contain_point(&s, t.p + 1))
2010-02-26 03:01:40 +01:00
return true; /* t is contained by s */
/* Y-cross */
2010-02-26 03:01:40 +01:00
res = touched_lseg_inside_poly(t.p, t.p + 1, &s, poly, i + 1);
}
else if (lseg_contain_point(&s, t.p + 1))
{
/* Y-cross */
2010-02-26 03:01:40 +01:00
res = touched_lseg_inside_poly(t.p + 1, t.p, &s, poly, i + 1);
}
else if (lseg_interpt_lseg(&interpt, &t, &s))
{
/*
* segments are X-crossing, go to check each subsegment
*/
intersection = true;
res = lseg_inside_poly(t.p, &interpt, poly, i + 1);
if (res)
res = lseg_inside_poly(t.p + 1, &interpt, poly, i + 1);
}
s.p[0] = s.p[1];
}
if (res && !intersection)
{
2010-02-26 03:01:40 +01:00
Point p;
/*
2010-02-26 03:01:40 +01:00
* if X-intersection wasn't found then check central point of tested
* segment. In opposite case we already check all subsegments
*/
p.x = float8_div(float8_pl(t.p[0].x, t.p[1].x), 2.0);
p.y = float8_div(float8_pl(t.p[0].y, t.p[1].y), 2.0);
2010-02-26 03:01:40 +01:00
res = point_inside(&p, poly->npts, poly->p);
}
return res;
}
/*
* Check whether the first polygon contains the second
*/
static bool
poly_contain_poly(POLYGON *contains_poly, POLYGON *contained_poly)
{
int i;
LSEG s;
Assert(contains_poly->npts > 0 && contained_poly->npts > 0);
/*
* Quick check to see if contained's bounding box is contained in
* contains' bb.
*/
if (!box_contain_box(&contains_poly->boundbox, &contained_poly->boundbox))
return false;
s.p[0] = contained_poly->p[contained_poly->npts - 1];
for (i = 0; i < contained_poly->npts; i++)
{
s.p[1] = contained_poly->p[i];
if (!lseg_inside_poly(s.p, s.p + 1, contains_poly, 0))
return false;
s.p[0] = s.p[1];
}
return true;
}
Datum
poly_contain(PG_FUNCTION_ARGS)
{
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
result = poly_contain_poly(polya, polyb);
2001-03-22 05:01:46 +01:00
/*
2005-10-15 04:49:52 +02:00
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
2001-03-22 05:01:46 +01:00
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
/*-----------------------------------------------------------------
* Determine if polygon A is contained by polygon B
*-----------------------------------------------------------------*/
Datum
poly_contained(PG_FUNCTION_ARGS)
{
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
bool result;
/* Just switch the arguments and pass it off to poly_contain */
result = poly_contain_poly(polyb, polya);
/*
* Avoid leaking memory for toasted inputs ... needed for rtree indexes
*/
PG_FREE_IF_COPY(polya, 0);
PG_FREE_IF_COPY(polyb, 1);
PG_RETURN_BOOL(result);
}
Datum
poly_contain_pt(PG_FUNCTION_ARGS)
{
2001-03-22 05:01:46 +01:00
POLYGON *poly = PG_GETARG_POLYGON_P(0);
Point *p = PG_GETARG_POINT_P(1);
PG_RETURN_BOOL(point_inside(p, poly->npts, poly->p) != 0);
}
Datum
pt_contained_poly(PG_FUNCTION_ARGS)
{
Point *p = PG_GETARG_POINT_P(0);
2001-03-22 05:01:46 +01:00
POLYGON *poly = PG_GETARG_POLYGON_P(1);
PG_RETURN_BOOL(point_inside(p, poly->npts, poly->p) != 0);
}
Datum
poly_distance(PG_FUNCTION_ARGS)
{
#ifdef NOT_USED
2001-03-22 05:01:46 +01:00
POLYGON *polya = PG_GETARG_POLYGON_P(0);
POLYGON *polyb = PG_GETARG_POLYGON_P(1);
#endif
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("function \"poly_distance\" not implemented")));
PG_RETURN_NULL();
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/***********************************************************************
**
** Routines for 2D points.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
**
***********************************************************************/
Datum
construct_point(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
float8 x = PG_GETARG_FLOAT8(0);
float8 y = PG_GETARG_FLOAT8(1);
Point *result;
result = (Point *) palloc(sizeof(Point));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_construct(result, x, y);
PG_RETURN_POINT_P(result);
}
static inline void
point_add_point(Point *result, Point *pt1, Point *pt2)
{
point_construct(result,
float8_pl(pt1->x, pt2->x),
float8_pl(pt1->y, pt2->y));
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
point_add(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
Point *p1 = PG_GETARG_POINT_P(0);
Point *p2 = PG_GETARG_POINT_P(1);
Point *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (Point *) palloc(sizeof(Point));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_add_point(result, p1, p2);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_POINT_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
static inline void
point_sub_point(Point *result, Point *pt1, Point *pt2)
{
point_construct(result,
float8_mi(pt1->x, pt2->x),
float8_mi(pt1->y, pt2->y));
}
Datum
point_sub(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
Point *p1 = PG_GETARG_POINT_P(0);
Point *p2 = PG_GETARG_POINT_P(1);
Point *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (Point *) palloc(sizeof(Point));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_sub_point(result, p1, p2);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_POINT_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
static inline void
point_mul_point(Point *result, Point *pt1, Point *pt2)
{
point_construct(result,
float8_mi(float8_mul(pt1->x, pt2->x),
float8_mul(pt1->y, pt2->y)),
float8_pl(float8_mul(pt1->x, pt2->y),
float8_mul(pt1->y, pt2->x)));
}
Datum
point_mul(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
Point *p1 = PG_GETARG_POINT_P(0);
Point *p2 = PG_GETARG_POINT_P(1);
Point *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (Point *) palloc(sizeof(Point));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_mul_point(result, p1, p2);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_POINT_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
static inline void
point_div_point(Point *result, Point *pt1, Point *pt2)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
float8 div;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
div = float8_pl(float8_mul(pt2->x, pt2->x), float8_mul(pt2->y, pt2->y));
point_construct(result,
float8_div(float8_pl(float8_mul(pt1->x, pt2->x),
float8_mul(pt1->y, pt2->y)), div),
float8_div(float8_mi(float8_mul(pt1->y, pt2->x),
float8_mul(pt1->x, pt2->y)), div));
}
Datum
point_div(PG_FUNCTION_ARGS)
{
Point *p1 = PG_GETARG_POINT_P(0);
Point *p2 = PG_GETARG_POINT_P(1);
Point *result;
result = (Point *) palloc(sizeof(Point));
point_div_point(result, p1, p2);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_POINT_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/***********************************************************************
**
** Routines for 2D boxes.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
**
***********************************************************************/
Datum
points_box(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
Point *p1 = PG_GETARG_POINT_P(0);
Point *p2 = PG_GETARG_POINT_P(1);
BOX *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (BOX *) palloc(sizeof(BOX));
box_construct(result, p1, p2);
PG_RETURN_BOX_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
box_add(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
BOX *box = PG_GETARG_BOX_P(0);
Point *p = PG_GETARG_POINT_P(1);
BOX *result;
result = (BOX *) palloc(sizeof(BOX));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_add_point(&result->high, &box->high, p);
point_add_point(&result->low, &box->low, p);
PG_RETURN_BOX_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
box_sub(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
BOX *box = PG_GETARG_BOX_P(0);
Point *p = PG_GETARG_POINT_P(1);
BOX *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (BOX *) palloc(sizeof(BOX));
point_sub_point(&result->high, &box->high, p);
point_sub_point(&result->low, &box->low, p);
PG_RETURN_BOX_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
box_mul(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
BOX *box = PG_GETARG_BOX_P(0);
Point *p = PG_GETARG_POINT_P(1);
BOX *result;
Point high,
low;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (BOX *) palloc(sizeof(BOX));
point_mul_point(&high, &box->high, p);
point_mul_point(&low, &box->low, p);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
box_construct(result, &high, &low);
PG_RETURN_BOX_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
box_div(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
BOX *box = PG_GETARG_BOX_P(0);
Point *p = PG_GETARG_POINT_P(1);
BOX *result;
Point high,
low;
result = (BOX *) palloc(sizeof(BOX));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_div_point(&high, &box->high, p);
point_div_point(&low, &box->low, p);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
box_construct(result, &high, &low);
PG_RETURN_BOX_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*
* Convert point to empty box
*/
Datum
point_box(PG_FUNCTION_ARGS)
{
Point *pt = PG_GETARG_POINT_P(0);
BOX *box;
box = (BOX *) palloc(sizeof(BOX));
box->high.x = pt->x;
box->low.x = pt->x;
box->high.y = pt->y;
box->low.y = pt->y;
PG_RETURN_BOX_P(box);
}
/*
* Smallest bounding box that includes both of the given boxes
*/
Datum
boxes_bound_box(PG_FUNCTION_ARGS)
{
BOX *box1 = PG_GETARG_BOX_P(0),
*box2 = PG_GETARG_BOX_P(1),
*container;
container = (BOX *) palloc(sizeof(BOX));
container->high.x = float8_max(box1->high.x, box2->high.x);
container->low.x = float8_min(box1->low.x, box2->low.x);
container->high.y = float8_max(box1->high.y, box2->high.y);
container->low.y = float8_min(box1->low.y, box2->low.y);
PG_RETURN_BOX_P(container);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/***********************************************************************
**
** Routines for 2D paths.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
**
***********************************************************************/
/* path_add()
* Concatenate two paths (only if they are both open).
*/
Datum
path_add(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *p1 = PG_GETARG_PATH_P(0);
PATH *p2 = PG_GETARG_PATH_P(1);
PATH *result;
int size,
base_size;
int i;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
if (p1->closed || p2->closed)
PG_RETURN_NULL();
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
base_size = sizeof(p1->p[0]) * (p1->npts + p2->npts);
2017-06-21 20:39:04 +02:00
size = offsetof(PATH, p) + base_size;
/* Check for integer overflow */
if (base_size / sizeof(p1->p[0]) != (p1->npts + p2->npts) ||
size <= base_size)
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("too many points requested")));
result = (PATH *) palloc(size);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
SET_VARSIZE(result, size);
result->npts = (p1->npts + p2->npts);
result->closed = p1->closed;
/* prevent instability in unused pad bytes */
result->dummy = 0;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < p1->npts; i++)
{
result->p[i].x = p1->p[i].x;
result->p[i].y = p1->p[i].y;
}
for (i = 0; i < p2->npts; i++)
{
result->p[i + p1->npts].x = p2->p[i].x;
result->p[i + p1->npts].y = p2->p[i].y;
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_PATH_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* path_add_pt()
* Translation operators.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
path_add_pt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P_COPY(0);
Point *point = PG_GETARG_POINT_P(1);
int i;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < path->npts; i++)
point_add_point(&path->p[i], &path->p[i], point);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_PATH_P(path);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
path_sub_pt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P_COPY(0);
Point *point = PG_GETARG_POINT_P(1);
int i;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < path->npts; i++)
point_sub_point(&path->p[i], &path->p[i], point);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_PATH_P(path);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* path_mul_pt()
* Rotation and scaling operators.
*/
Datum
path_mul_pt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P_COPY(0);
Point *point = PG_GETARG_POINT_P(1);
int i;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < path->npts; i++)
point_mul_point(&path->p[i], &path->p[i], point);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_PATH_P(path);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
path_div_pt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P_COPY(0);
Point *point = PG_GETARG_POINT_P(1);
int i;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < path->npts; i++)
point_div_point(&path->p[i], &path->p[i], point);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_PATH_P(path);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
path_center(PG_FUNCTION_ARGS)
{
#ifdef NOT_USED
PATH *path = PG_GETARG_PATH_P(0);
#endif
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("function \"path_center\" not implemented")));
PG_RETURN_NULL();
}
Datum
path_poly(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
PATH *path = PG_GETARG_PATH_P(0);
POLYGON *poly;
int size;
int i;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* This is not very consistent --- other similar cases return NULL ... */
if (!path->closed)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("open path cannot be converted to polygon")));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*
* Never overflows: the old size fit in MaxAllocSize, and the new size is
* just a small constant larger.
*/
2017-06-21 20:39:04 +02:00
size = offsetof(POLYGON, p) + sizeof(poly->p[0]) * path->npts;
poly = (POLYGON *) palloc(size);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
SET_VARSIZE(poly, size);
poly->npts = path->npts;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < path->npts; i++)
{
poly->p[i].x = path->p[i].x;
poly->p[i].y = path->p[i].y;
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
make_bound_box(poly);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_POLYGON_P(poly);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/***********************************************************************
**
** Routines for 2D polygons.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
**
***********************************************************************/
Datum
poly_npoints(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
2001-03-22 05:01:46 +01:00
POLYGON *poly = PG_GETARG_POLYGON_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_INT32(poly->npts);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
poly_center(PG_FUNCTION_ARGS)
{
2001-03-22 05:01:46 +01:00
POLYGON *poly = PG_GETARG_POLYGON_P(0);
Point *result;
CIRCLE circle;
result = (Point *) palloc(sizeof(Point));
poly_to_circle(&circle, poly);
*result = circle.center;
PG_RETURN_POINT_P(result);
}
Datum
poly_box(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
2001-03-22 05:01:46 +01:00
POLYGON *poly = PG_GETARG_POLYGON_P(0);
BOX *box;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
box = (BOX *) palloc(sizeof(BOX));
*box = poly->boundbox;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_BOX_P(box);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* box_poly()
* Convert a box to a polygon.
*/
Datum
box_poly(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
BOX *box = PG_GETARG_BOX_P(0);
POLYGON *poly;
int size;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* map four corners of the box to a polygon */
2017-06-21 20:39:04 +02:00
size = offsetof(POLYGON, p) + sizeof(poly->p[0]) * 4;
poly = (POLYGON *) palloc(size);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
SET_VARSIZE(poly, size);
poly->npts = 4;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
poly->p[0].x = box->low.x;
poly->p[0].y = box->low.y;
poly->p[1].x = box->low.x;
poly->p[1].y = box->high.y;
poly->p[2].x = box->high.x;
poly->p[2].y = box->high.y;
poly->p[3].x = box->high.x;
poly->p[3].y = box->low.y;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
box_construct(&poly->boundbox, &box->high, &box->low);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_POLYGON_P(poly);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
poly_path(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
2001-03-22 05:01:46 +01:00
POLYGON *poly = PG_GETARG_POLYGON_P(0);
PATH *path;
int size;
int i;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*
* Never overflows: the old size fit in MaxAllocSize, and the new size is
* smaller by a small constant.
*/
2017-06-21 20:39:04 +02:00
size = offsetof(PATH, p) + sizeof(path->p[0]) * poly->npts;
path = (PATH *) palloc(size);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
SET_VARSIZE(path, size);
path->npts = poly->npts;
path->closed = true;
/* prevent instability in unused pad bytes */
path->dummy = 0;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < poly->npts; i++)
{
path->p[i].x = poly->p[i].x;
path->p[i].y = poly->p[i].y;
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_PATH_P(path);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/***********************************************************************
**
** Routines for circles.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
**
***********************************************************************/
/*----------------------------------------------------------
* Formatting and conversion routines.
*---------------------------------------------------------*/
/* circle_in - convert a string to internal form.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*
* External format: (center and radius of circle)
* "<(f8,f8),f8>"
* also supports quick entry style "f8,f8,f8"
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_in(PG_FUNCTION_ARGS)
{
char *str = PG_GETARG_CSTRING(0);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
CIRCLE *circle = (CIRCLE *) palloc(sizeof(CIRCLE));
char *s,
*cp;
int depth = 0;
s = str;
while (isspace((unsigned char) *s))
s++;
if (*s == LDELIM_C)
depth++, s++;
else if (*s == LDELIM)
{
/* If there are two left parens, consume the first one */
cp = (s + 1);
while (isspace((unsigned char) *cp))
cp++;
if (*cp == LDELIM)
depth++, s = cp;
1997-09-05 20:13:45 +02:00
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* pair_decode will consume parens around the pair, if any */
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
pair_decode(s, &circle->center.x, &circle->center.y, &s, "circle", str);
if (*s == DELIM)
s++;
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
circle->radius = single_decode(s, &s, "circle", str);
/* We have to accept NaN. */
if (circle->radius < 0.0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
errmsg("invalid input syntax for type %s: \"%s\"",
"circle", str)));
while (depth > 0)
{
if ((*s == RDELIM) || ((*s == RDELIM_C) && (depth == 1)))
{
depth--;
s++;
while (isspace((unsigned char) *s))
s++;
}
else
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
errmsg("invalid input syntax for type %s: \"%s\"",
"circle", str)));
1997-09-05 20:13:45 +02:00
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
if (*s != '\0')
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
errmsg("invalid input syntax for type %s: \"%s\"",
"circle", str)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CIRCLE_P(circle);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* circle_out - convert a circle to external form.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_out(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
StringInfoData str;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
initStringInfo(&str);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
appendStringInfoChar(&str, LDELIM_C);
appendStringInfoChar(&str, LDELIM);
pair_encode(circle->center.x, circle->center.y, &str);
appendStringInfoChar(&str, RDELIM);
appendStringInfoChar(&str, DELIM);
single_encode(circle->radius, &str);
appendStringInfoChar(&str, RDELIM_C);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Improve portability of I/O behavior for the geometric types. Formerly, the geometric I/O routines such as box_in and point_out relied directly on strtod() and sprintf() for conversion of the float8 component values of their data types. However, the behavior of those functions is pretty platform-dependent, especially for edge-case values such as infinities and NaNs. This was exposed by commit acdf2a8b372aec1d, which added test cases involving boxes with infinity endpoints, and immediately failed on Windows and AIX buildfarm members. We solved these problems years ago in the main float8in and float8out functions, so let's fix it by making the geometric types use that code instead of depending directly on the platform-supplied functions. To do this, refactor the float8in code so that it can be used to parse just part of a string, and as a convenience make the guts of float8out usable without going through DirectFunctionCall. While at it, get rid of geo_ops.c's fairly shaky assumptions about the maximum output string length for a double, by having it build results in StringInfo buffers instead of fixed-length strings. In passing, convert all the "invalid input syntax for type foo" messages in this area of the code into "invalid input syntax for type %s" to reduce the number of distinct translatable strings, per recent discussion. We would have needed a fair number of the latter anyway for code-sharing reasons, so we might as well just go whole hog. Note: this patch is by no means intended to guarantee that the geometric types uniformly behave sanely for infinity or NaN component values. But any bugs we have in that line were there all along, they were just harder to reach in a platform-independent way.
2016-03-30 23:25:03 +02:00
PG_RETURN_CSTRING(str.data);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
2003-05-13 20:03:08 +02:00
/*
* circle_recv - converts external binary format to circle
*/
Datum
circle_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
CIRCLE *circle;
circle = (CIRCLE *) palloc(sizeof(CIRCLE));
circle->center.x = pq_getmsgfloat8(buf);
circle->center.y = pq_getmsgfloat8(buf);
circle->radius = pq_getmsgfloat8(buf);
/* We have to accept NaN. */
if (circle->radius < 0.0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_BINARY_REPRESENTATION),
errmsg("invalid radius in external \"circle\" value")));
2003-05-13 20:03:08 +02:00
PG_RETURN_CIRCLE_P(circle);
}
/*
* circle_send - converts circle to binary format
*/
Datum
circle_send(PG_FUNCTION_ARGS)
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
StringInfoData buf;
pq_begintypsend(&buf);
pq_sendfloat8(&buf, circle->center.x);
pq_sendfloat8(&buf, circle->center.y);
pq_sendfloat8(&buf, circle->radius);
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*----------------------------------------------------------
* Relational operators for CIRCLEs.
* <, >, <=, >=, and == are based on circle area.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*---------------------------------------------------------*/
/* circles identical?
*
* We consider NaNs values to be equal to each other to let those circles
* to be found.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_same(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(((isnan(circle1->radius) && isnan(circle1->radius)) ||
FPeq(circle1->radius, circle2->radius)) &&
point_eq_point(&circle1->center, &circle2->center));
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* circle_overlap - does circle1 overlap circle2?
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_overlap(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPle(point_dt(&circle1->center, &circle2->center),
float8_pl(circle1->radius, circle2->radius)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_overleft - is the right edge of circle1 at or left of
* the right edge of circle2?
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_overleft(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPle(float8_pl(circle1->center.x, circle1->radius),
float8_pl(circle2->center.x, circle2->radius)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_left - is circle1 strictly left of circle2?
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_left(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPlt(float8_pl(circle1->center.x, circle1->radius),
float8_mi(circle2->center.x, circle2->radius)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_right - is circle1 strictly right of circle2?
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_right(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPgt(float8_mi(circle1->center.x, circle1->radius),
float8_pl(circle2->center.x, circle2->radius)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_overright - is the left edge of circle1 at or right of
* the left edge of circle2?
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_overright(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPge(float8_mi(circle1->center.x, circle1->radius),
float8_mi(circle2->center.x, circle2->radius)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_contained - is circle1 contained by circle2?
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_contained(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPle(point_dt(&circle1->center, &circle2->center),
float8_mi(circle2->radius, circle1->radius)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_contain - does circle1 contain circle2?
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_contain(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPle(point_dt(&circle1->center, &circle2->center),
float8_mi(circle1->radius, circle2->radius)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_below - is circle1 strictly below circle2?
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_below(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPlt(float8_pl(circle1->center.y, circle1->radius),
float8_mi(circle2->center.y, circle2->radius)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_above - is circle1 strictly above circle2?
*/
Datum
circle_above(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPgt(float8_mi(circle1->center.y, circle1->radius),
float8_pl(circle2->center.y, circle2->radius)));
}
/* circle_overbelow - is the upper edge of circle1 at or below
* the upper edge of circle2?
*/
Datum
circle_overbelow(PG_FUNCTION_ARGS)
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPle(float8_pl(circle1->center.y, circle1->radius),
float8_pl(circle2->center.y, circle2->radius)));
}
/* circle_overabove - is the lower edge of circle1 at or above
* the lower edge of circle2?
*/
Datum
circle_overabove(PG_FUNCTION_ARGS)
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPge(float8_mi(circle1->center.y, circle1->radius),
float8_mi(circle2->center.y, circle2->radius)));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_relop - is area(circle1) relop area(circle2), within
* our accuracy constraint?
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_eq(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_BOOL(FPeq(circle_ar(circle1), circle_ar(circle2)));
}
Datum
circle_ne(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPne(circle_ar(circle1), circle_ar(circle2)));
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
circle_lt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPlt(circle_ar(circle1), circle_ar(circle2)));
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
circle_gt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPgt(circle_ar(circle1), circle_ar(circle2)));
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
circle_le(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_BOOL(FPle(circle_ar(circle1), circle_ar(circle2)));
}
Datum
circle_ge(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
PG_RETURN_BOOL(FPge(circle_ar(circle1), circle_ar(circle2)));
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*----------------------------------------------------------
* "Arithmetic" operators on circles.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*---------------------------------------------------------*/
/* circle_add_pt()
* Translation operator.
*/
Datum
circle_add_pt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Point *point = PG_GETARG_POINT_P(1);
CIRCLE *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (CIRCLE *) palloc(sizeof(CIRCLE));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_add_point(&result->center, &circle->center, point);
result->radius = circle->radius;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CIRCLE_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
circle_sub_pt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Point *point = PG_GETARG_POINT_P(1);
CIRCLE *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (CIRCLE *) palloc(sizeof(CIRCLE));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_sub_point(&result->center, &circle->center, point);
result->radius = circle->radius;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CIRCLE_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* circle_mul_pt()
* Rotation and scaling operators.
*/
Datum
circle_mul_pt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Point *point = PG_GETARG_POINT_P(1);
CIRCLE *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (CIRCLE *) palloc(sizeof(CIRCLE));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_mul_point(&result->center, &circle->center, point);
result->radius = float8_mul(circle->radius, HYPOT(point->x, point->y));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CIRCLE_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
circle_div_pt(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Point *point = PG_GETARG_POINT_P(1);
CIRCLE *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (CIRCLE *) palloc(sizeof(CIRCLE));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
point_div_point(&result->center, &circle->center, point);
result->radius = float8_div(circle->radius, HYPOT(point->x, point->y));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CIRCLE_P(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* circle_area - returns the area of the circle.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_area(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_FLOAT8(circle_ar(circle));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_diameter - returns the diameter of the circle.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_diameter(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_FLOAT8(float8_mul(circle->radius, 2.0));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_radius - returns the radius of the circle.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_radius(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_FLOAT8(circle->radius);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_distance - returns the distance between
* two circles.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_distance(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle1 = PG_GETARG_CIRCLE_P(0);
CIRCLE *circle2 = PG_GETARG_CIRCLE_P(1);
float8 result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = float8_mi(point_dt(&circle1->center, &circle2->center),
float8_pl(circle1->radius, circle2->radius));
if (result < 0.0)
result = 0.0;
PG_RETURN_FLOAT8(result);
}
Datum
circle_contain_pt(PG_FUNCTION_ARGS)
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Point *point = PG_GETARG_POINT_P(1);
float8 d;
d = point_dt(&circle->center, point);
PG_RETURN_BOOL(d <= circle->radius);
}
Datum
pt_contained_circle(PG_FUNCTION_ARGS)
{
Point *point = PG_GETARG_POINT_P(0);
CIRCLE *circle = PG_GETARG_CIRCLE_P(1);
float8 d;
d = point_dt(&circle->center, point);
PG_RETURN_BOOL(d <= circle->radius);
}
/* dist_pc - returns the distance between
* a point and a circle.
*/
Datum
dist_pc(PG_FUNCTION_ARGS)
{
Point *point = PG_GETARG_POINT_P(0);
CIRCLE *circle = PG_GETARG_CIRCLE_P(1);
float8 result;
result = float8_mi(point_dt(point, &circle->center),
circle->radius);
if (result < 0.0)
result = 0.0;
PG_RETURN_FLOAT8(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/*
* Distance from a circle to a point
*/
Datum
dist_cpoint(PG_FUNCTION_ARGS)
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Point *point = PG_GETARG_POINT_P(1);
float8 result;
result = float8_mi(point_dt(point, &circle->center), circle->radius);
if (result < 0.0)
result = 0.0;
PG_RETURN_FLOAT8(result);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
/* circle_center - returns the center point of the circle.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
Datum
circle_center(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
Point *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (Point *) palloc(sizeof(Point));
result->x = circle->center.x;
result->y = circle->center.y;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_POINT_P(result);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/* circle_ar - returns the area of the circle.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
static float8
circle_ar(CIRCLE *circle)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
return float8_mul(float8_mul(circle->radius, circle->radius), M_PI);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/*----------------------------------------------------------
* Conversion operators.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*---------------------------------------------------------*/
Datum
cr_circle(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
Point *center = PG_GETARG_POINT_P(0);
float8 radius = PG_GETARG_FLOAT8(1);
CIRCLE *result;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result = (CIRCLE *) palloc(sizeof(CIRCLE));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result->center.x = center->x;
result->center.y = center->y;
result->radius = radius;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_CIRCLE_P(result);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
Datum
circle_box(PG_FUNCTION_ARGS)
{
CIRCLE *circle = PG_GETARG_CIRCLE_P(0);
BOX *box;
float8 delta;
box = (BOX *) palloc(sizeof(BOX));
delta = float8_div(circle->radius, sqrt(2.0));
box->high.x = float8_pl(circle->center.x, delta);
box->low.x = float8_mi(circle->center.x, delta);
box->high.y = float8_pl(circle->center.y, delta);
box->low.y = float8_mi(circle->center.y, delta);
PG_RETURN_BOX_P(box);
}
/* box_circle()
* Convert a box to a circle.
*/
Datum
box_circle(PG_FUNCTION_ARGS)
{
BOX *box = PG_GETARG_BOX_P(0);
CIRCLE *circle;
circle = (CIRCLE *) palloc(sizeof(CIRCLE));
circle->center.x = float8_div(float8_pl(box->high.x, box->low.x), 2.0);
circle->center.y = float8_div(float8_pl(box->high.y, box->low.y), 2.0);
circle->radius = point_dt(&circle->center, &box->high);
PG_RETURN_CIRCLE_P(circle);
}
Datum
circle_poly(PG_FUNCTION_ARGS)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
int32 npts = PG_GETARG_INT32(0);
CIRCLE *circle = PG_GETARG_CIRCLE_P(1);
POLYGON *poly;
int base_size,
size;
int i;
float8 angle;
float8 anglestep;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
if (FPzero(circle->radius))
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("cannot convert circle with radius zero to polygon")));
if (npts < 2)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("must request at least 2 points")));
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
base_size = sizeof(poly->p[0]) * npts;
2017-06-21 20:39:04 +02:00
size = offsetof(POLYGON, p) + base_size;
/* Check for integer overflow */
if (base_size / npts != sizeof(poly->p[0]) || size <= base_size)
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("too many points requested")));
poly = (POLYGON *) palloc0(size); /* zero any holes */
SET_VARSIZE(poly, size);
poly->npts = npts;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
anglestep = float8_div(2.0 * M_PI, npts);
for (i = 0; i < npts; i++)
{
angle = float8_mul(anglestep, i);
poly->p[i].x = float8_mi(circle->center.x,
float8_mul(circle->radius, cos(angle)));
poly->p[i].y = float8_pl(circle->center.y,
float8_mul(circle->radius, sin(angle)));
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
make_bound_box(poly);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
PG_RETURN_POLYGON_P(poly);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
}
/*
* Convert polygon to circle
*
* The result must be preallocated.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*
* XXX This algorithm should use weighted means of line segments
* rather than straight average values of points - tgl 97/01/21.
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
*/
static void
poly_to_circle(CIRCLE *result, POLYGON *poly)
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
{
int i;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Assert(poly->npts > 0);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
result->center.x = 0;
result->center.y = 0;
result->radius = 0;
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < poly->npts; i++)
point_add_point(&result->center, &result->center, &poly->p[i]);
result->center.x = float8_div(result->center.x, poly->npts);
result->center.y = float8_div(result->center.y, poly->npts);
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
for (i = 0; i < poly->npts; i++)
result->radius = float8_pl(result->radius,
point_dt(&poly->p[i], &result->center));
result->radius = float8_div(result->radius, poly->npts);
}
Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> OK, here are a passel of patches for the geometric data types. These add a "circle" data type, new operators and functions for the existing data types, and change the default formats for some of the existing types to make them consistant with each other. Current formatting conventions (e.g. compatible with v6.0 to allow dump/reload) are supported, but the new conventions should be an improvement and we can eventually drop the old conventions entirely. For example, there are two kinds of paths (connected line segments), open and closed, and the old format was '(1,2,1,2,3,4)' for a closed path with two points (1,2) and (3,4) '(0,2,1,2,3,4)' for an open path with two points (1,2) and (3,4) Pretty arcane, huh? The new format for paths is '((1,2),(3,4))' for a closed path with two points (1,2) and (3,4) '[(1,2),(3,4)]' for an open path with two points (1,2) and (3,4) For polygons, the old convention is '(0,4,2,0,4,3)' for a triangle with points at (0,0),(4,4), and (2,3) and the new convention is '((0,0),(4,4),(2,3))' for a triangle with points at (0,0),(4,4), and (2,3) Other data types which are also represented as lists of points (e.g. boxes, line segments, and polygons) have similar representations (they surround each point with parens). For v6.1, any format which can be interpreted as the old style format is decoded as such; we can remove that backwards compatibility but ugly convention for v7.0. This will allow dump/reloads from v6.0. These include some updates to the regression test files to change the test for creating a data type from "circle" to "widget" to keep the test from trashing the new builtin circle type.
1997-04-22 19:35:09 +02:00
Datum
poly_circle(PG_FUNCTION_ARGS)
{
POLYGON *poly = PG_GETARG_POLYGON_P(0);
CIRCLE *result;
result = (CIRCLE *) palloc(sizeof(CIRCLE));
poly_to_circle(result, poly);
PG_RETURN_CIRCLE_P(result);
}
/***********************************************************************
**
** Private routines for multiple types.
**
***********************************************************************/
/*
* Test to see if the point is inside the polygon, returns 1/0, or 2 if
* the point is on the polygon.
* Code adapted but not copied from integer-based routines in WN: A
* Server for the HTTP
* version 1.15.1, file wn/image.c
* http://hopf.math.northwestern.edu/index.html
* Description of algorithm: http://www.linuxjournal.com/article/2197
* http://www.linuxjournal.com/article/2029
*/
#define POINT_ON_POLYGON INT_MAX
static int
1998-09-01 05:29:17 +02:00
point_inside(Point *p, int npts, Point *plist)
{
float8 x0,
y0;
float8 prev_x,
prev_y;
int i = 0;
float8 x,
y;
2007-11-15 22:14:46 +01:00
int cross,
total_cross = 0;
Assert(npts > 0);
/* compute first polygon point relative to single point */
x0 = float8_mi(plist[0].x, p->x);
y0 = float8_mi(plist[0].y, p->y);
prev_x = x0;
prev_y = y0;
/* loop over polygon points and aggregate total_cross */
for (i = 1; i < npts; i++)
{
/* compute next polygon point relative to single point */
x = float8_mi(plist[i].x, p->x);
y = float8_mi(plist[i].y, p->y);
/* compute previous to current point crossing */
if ((cross = lseg_crossing(x, y, prev_x, prev_y)) == POINT_ON_POLYGON)
return 2;
total_cross += cross;
2007-11-15 22:14:46 +01:00
prev_x = x;
prev_y = y;
}
/* now do the first point */
if ((cross = lseg_crossing(x0, y0, prev_x, prev_y)) == POINT_ON_POLYGON)
return 2;
total_cross += cross;
if (total_cross != 0)
return 1;
return 0;
}
/* lseg_crossing()
* Returns +/-2 if line segment crosses the positive X-axis in a +/- direction.
* Returns +/-1 if one point is on the positive X-axis.
* Returns 0 if both points are on the positive X-axis, or there is no crossing.
* Returns POINT_ON_POLYGON if the segment contains (0,0).
* Wow, that is one confusing API, but it is used above, and when summed,
* can tell is if a point is in a polygon.
*/
static int
lseg_crossing(float8 x, float8 y, float8 prev_x, float8 prev_y)
{
float8 z;
int y_sign;
if (FPzero(y))
2007-11-15 22:14:46 +01:00
{ /* y == 0, on X axis */
if (FPzero(x)) /* (x,y) is (0,0)? */
return POINT_ON_POLYGON;
else if (FPgt(x, 0))
2007-11-15 22:14:46 +01:00
{ /* x > 0 */
if (FPzero(prev_y)) /* y and prev_y are zero */
/* prev_x > 0? */
return FPgt(prev_x, 0.0) ? 0 : POINT_ON_POLYGON;
return FPlt(prev_y, 0.0) ? 1 : -1;
}
else
2007-11-15 22:14:46 +01:00
{ /* x < 0, x not on positive X axis */
if (FPzero(prev_y))
/* prev_x < 0? */
return FPlt(prev_x, 0.0) ? 0 : POINT_ON_POLYGON;
1998-09-01 05:29:17 +02:00
return 0;
}
1997-09-05 20:13:45 +02:00
}
else
2007-11-15 22:14:46 +01:00
{ /* y != 0 */
/* compute y crossing direction from previous point */
y_sign = FPgt(y, 0.0) ? 1 : -1;
if (FPzero(prev_y))
/* previous point was on X axis, so new point is either off or on */
return FPlt(prev_x, 0.0) ? 0 : y_sign;
else if ((y_sign < 0 && FPlt(prev_y, 0.0)) ||
(y_sign > 0 && FPgt(prev_y, 0.0)))
/* both above or below X axis */
2007-11-15 22:14:46 +01:00
return 0; /* same sign */
else
2007-11-15 22:14:46 +01:00
{ /* y and prev_y cross X-axis */
if (FPge(x, 0.0) && FPgt(prev_x, 0.0))
/* both non-negative so cross positive X-axis */
return 2 * y_sign;
if (FPlt(x, 0.0) && FPle(prev_x, 0.0))
/* both non-positive so do not cross positive X-axis */
return 0;
/* x and y cross axes, see URL above point_inside() */
z = float8_mi(float8_mul(float8_mi(x, prev_x), y),
float8_mul(float8_mi(y, prev_y), x));
if (FPzero(z))
return POINT_ON_POLYGON;
if ((y_sign < 0 && FPlt(z, 0.0)) ||
(y_sign > 0 && FPgt(z, 0.0)))
return 0;
return 2 * y_sign;
}
}
}
static bool
1998-09-01 05:29:17 +02:00
plist_same(int npts, Point *p1, Point *p2)
{
int i,
ii,
j;
/* find match for first point */
for (i = 0; i < npts; i++)
{
if (point_eq_point(&p2[i], &p1[0]))
{
/* match found? then look forward through remaining points */
for (ii = 1, j = i + 1; ii < npts; ii++, j++)
{
if (j >= npts)
j = 0;
if (!point_eq_point(&p2[j], &p1[ii]))
break;
}
if (ii == npts)
return true;
/* match not found forwards? then look backwards */
for (ii = 1, j = i - 1; ii < npts; ii++, j--)
{
if (j < 0)
j = (npts - 1);
if (!point_eq_point(&p2[j], &p1[ii]))
break;
}
if (ii == npts)
return true;
}
1997-09-05 20:13:45 +02:00
}
return false;
}
/*-------------------------------------------------------------------------
* Determine the hypotenuse.
*
* If required, x and y are swapped to make x the larger number. The
* traditional formula of x^2+y^2 is rearranged to factor x outside the
* sqrt. This allows computation of the hypotenuse for significantly
* larger values, and with a higher precision than when using the naive
* formula. In particular, this cannot overflow unless the final result
* would be out-of-range.
*
* sqrt( x^2 + y^2 ) = sqrt( x^2( 1 + y^2/x^2) )
* = x * sqrt( 1 + y^2/x^2 )
* = x * sqrt( 1 + y/x * y/x )
*
* It is expected that this routine will eventually be replaced with the
* C99 hypot() function.
*
* This implementation conforms to IEEE Std 1003.1 and GLIBC, in that the
* case of hypot(inf,nan) results in INF, and not NAN.
*-----------------------------------------------------------------------
*/
float8
pg_hypot(float8 x, float8 y)
{
float8 yx,
result;
/* Handle INF and NaN properly */
if (isinf(x) || isinf(y))
return get_float8_infinity();
if (isnan(x) || isnan(y))
return get_float8_nan();
/* Else, drop any minus signs */
x = fabs(x);
y = fabs(y);
/* Swap x and y if needed to make x the larger one */
if (x < y)
{
float8 temp = x;
x = y;
y = temp;
}
/*
* If y is zero, the hypotenuse is x. This test saves a few cycles in
* such cases, but more importantly it also protects against
* divide-by-zero errors, since now x >= y.
*/
if (y == 0.0)
return x;
/* Determine the hypotenuse */
yx = y / x;
result = x * sqrt(1.0 + (yx * yx));
Avoid a performance regression in float overflow/underflow detection. Commit 6bf0bc842 replaced float.c's CHECKFLOATVAL() macro with static inline subroutines, but that wasn't too well thought out. In the original coding, the unlikely condition (isinf(result) or result == 0) was checked first, and the inf_is_valid or zero_is_valid condition only afterwards. The inline-subroutine coding caused that to be swapped around, which is pretty horrid for performance because (a) in common cases the is_valid condition is twice as expensive to evaluate (e.g., requiring two isinf() calls not one) and (b) in common cases the is_valid condition is false, requiring us to perform the unlikely-condition check anyway. Net result is that one isinf() call becomes two or three, resulting in visible performance loss as reported by Keisuke Kuroda. The original fix proposal was to revert the replacement of the macro, but on second thought, that macro was just a bad idea from the beginning: if anything it's a net negative for readability of the code. So instead, let's just open-code all the overflow/underflow tests, being careful to test the unlikely condition first (and mark it unlikely() to help the compiler get the point). Also, rather than having N copies of the actual ereport() calls, collapse those into out-of-line error subroutines to save some code space. This does mean that the error file/line numbers won't be very helpful for figuring out where the issue really is --- but we'd already burned that bridge by putting the ereports into static inlines. In HEAD, check_float[48]_val() are gone altogether. In v12, leave them present in float.h but unused in the core code, just in case some extension is depending on them. Emre Hasegeli, with some kibitzing from me and Andres Freund Discussion: https://postgr.es/m/CANDwggLe1Gc1OrRqvPfGE=kM9K0FSfia0hbeFCEmwabhLz95AA@mail.gmail.com
2020-02-13 19:37:43 +01:00
if (unlikely(isinf(result)))
float_overflow_error();
if (unlikely(result == 0.0))
float_underflow_error();
return result;
}