rdelaage
/
postgresql
mirror of https://git.postgresql.org/git/postgresql.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
postgresql/src/backend/commands/opclasscmds.c

970 lines
26 KiB
C
Raw Normal View History

Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/*-------------------------------------------------------------------------
*
* opclasscmds.c
*
* Routines for opclass manipulation commands
*
Tag appropriate files for rc3 Also performed an initial run through of upgrading our Copyright date to extend to 2005 ... first run here was very simple ... change everything where: grep 1996-2004 && the word 'Copyright' ... scanned through the generated list with 'less' first, and after, to make sure that I only picked up the right entries ...
2004-12-31 23:04:05 +01:00
* Portions Copyright (c) 1996-2005, PostgreSQL Global Development Group
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
* Portions Copyright (c) 1994, Regents of the University of California
*
*
* IDENTIFICATION
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
* $PostgreSQL: pgsql/src/backend/commands/opclasscmds.c,v 1.30 2005/03/29 00:16:57 tgl Exp $
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
*
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include "access/genam.h"
#include "access/heapam.h"
#include "catalog/catname.h"
#include "catalog/dependency.h"
#include "catalog/indexing.h"
#include "catalog/namespace.h"
#include "catalog/pg_am.h"
#include "catalog/pg_amop.h"
#include "catalog/pg_amproc.h"
#include "catalog/pg_opclass.h"
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
#include "catalog/pg_operator.h"
#include "catalog/pg_proc.h"
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
#include "catalog/pg_type.h"
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
#include "commands/defrem.h"
#include "miscadmin.h"
#include "parser/parse_func.h"
#include "parser/parse_oper.h"
#include "parser/parse_type.h"
#include "utils/acl.h"
#include "utils/builtins.h"
#include "utils/fmgroids.h"
#include "utils/lsyscache.h"
#include "utils/syscache.h"
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
/*
* We use lists of this struct type to keep track of both operators and
* procedures during DefineOpClass.
*/
typedef struct
{
Oid object; /* operator or support proc's OID */
int number; /* strategy or support proc number */
Oid subtype; /* subtype */
bool recheck; /* oper recheck flag (unused for proc) */
} OpClassMember;
static Oid assignOperSubtype(Oid amoid, Oid typeoid, Oid operOid);
static Oid assignProcSubtype(Oid amoid, Oid typeoid, Oid procOid);
static void addClassMember(List **list, OpClassMember *member, bool isProc);
static void storeOperators(Oid opclassoid, List *operators);
static void storeProcedures(Oid opclassoid, List *procedures);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/*
* DefineOpClass
* Define a new index operator class.
*/
void
DefineOpClass(CreateOpClassStmt *stmt)
{
char *opcname; /* name of opclass we're creating */
Oid amoid, /* our AM's oid */
typeoid, /* indexable datatype oid */
storageoid, /* storage datatype oid, if any */
namespaceoid, /* namespace to create opclass in */
opclassoid; /* oid of opclass we create */
int numOperators, /* amstrategies value */
numProcs; /* amsupport value */
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
List *operators; /* OpClassMember list for operators */
List *procedures; /* OpClassMember list for support procs */
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
ListCell *l;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
Relation rel;
HeapTuple tup;
Datum values[Natts_pg_opclass];
char nulls[Natts_pg_opclass];
AclResult aclresult;
NameData opcName;
int i;
pgindent run.
2002-09-04 22:31:48 +02:00
ObjectAddress myself,
referenced;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/* Convert list of names to a name and namespace */
namespaceoid = QualifiedNameGetCreationNamespace(stmt->opclassname,
&opcname);
/* Check we have creation rights in target namespace */
aclresult = pg_namespace_aclcheck(namespaceoid, GetUserId(), ACL_CREATE);
if (aclresult != ACLCHECK_OK)
Adjust 'permission denied' messages to be more useful and consistent.
2003-08-01 02:15:26 +02:00
aclcheck_error(aclresult, ACL_KIND_NAMESPACE,
get_namespace_name(namespaceoid));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/* Get necessary info about access method */
tup = SearchSysCache(AMNAME,
CStringGetDatum(stmt->amname),
0, 0, 0);
if (!HeapTupleIsValid(tup))
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("access method \"%s\" does not exist",
stmt->amname)));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
amoid = HeapTupleGetOid(tup);
numOperators = ((Form_pg_am) GETSTRUCT(tup))->amstrategies;
numProcs = ((Form_pg_am) GETSTRUCT(tup))->amsupport;
/* XXX Should we make any privilege check against the AM? */
ReleaseSysCache(tup);
Restrict CREATE OPERATOR CLASS to superusers, per discussion some weeks ago.
2002-10-05 00:19:29 +02:00
/*
* Currently, we require superuser privileges to create an opclass.
* This seems necessary because we have no way to validate that the
* offered set of operators and functions are consistent with the AM's
pgindent run.
2003-08-04 02:43:34 +02:00
* expectations. It would be nice to provide such a check someday, if
* it can be done without solving the halting problem :-(
Restrict CREATE OPERATOR CLASS to superusers, per discussion some weeks ago.
2002-10-05 00:19:29 +02:00
*/
if (!superuser())
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
pgindent run.
2003-08-04 02:43:34 +02:00
errmsg("must be superuser to create an operator class")));
Restrict CREATE OPERATOR CLASS to superusers, per discussion some weeks ago.
2002-10-05 00:19:29 +02:00
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/* Look up the datatype */
typeoid = typenameTypeId(stmt->datatype);
Restrict CREATE OPERATOR CLASS to superusers, per discussion some weeks ago.
2002-10-05 00:19:29 +02:00
#ifdef NOT_USED
/* XXX this is unnecessary given the superuser check above */
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/* Check we have ownership of the datatype */
if (!pg_type_ownercheck(typeoid, GetUserId()))
Adjust 'permission denied' messages to be more useful and consistent.
2003-08-01 02:15:26 +02:00
aclcheck_error(ACLCHECK_NOT_OWNER, ACL_KIND_TYPE,
format_type_be(typeoid));
Restrict CREATE OPERATOR CLASS to superusers, per discussion some weeks ago.
2002-10-05 00:19:29 +02:00
#endif
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
operators = NIL;
procedures = NIL;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/* Storage datatype is optional */
storageoid = InvalidOid;
/*
* Scan the "items" list to obtain additional info.
*/
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
foreach(l, stmt->items)
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
{
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
CreateOpClassItem *item = lfirst(l);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
Oid operOid;
Oid funcOid;
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
OpClassMember *member;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
AclResult aclresult;
Assert(IsA(item, CreateOpClassItem));
switch (item->itemtype)
{
case OPCLASS_ITEM_OPERATOR:
if (item->number <= 0 || item->number > numOperators)
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
errmsg("invalid operator number %d,"
" must be between 1 and %d",
item->number, numOperators)));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
if (item->args != NIL)
{
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
TypeName *typeName1 = (TypeName *) linitial(item->args);
pgindent run.
2002-09-04 22:31:48 +02:00
TypeName *typeName2 = (TypeName *) lsecond(item->args);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
operOid = LookupOperNameTypeNames(item->name,
Some early work on error message editing. Operator-not-found and function-not-found messages now distinguish the cases no-match and ambiguous-match, and they follow the style guidelines too.
2003-07-04 04:51:34 +02:00
typeName1,
typeName2,
false);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
}
else
{
/* Default to binary op on input datatype */
Some early work on error message editing. Operator-not-found and function-not-found messages now distinguish the cases no-match and ambiguous-match, and they follow the style guidelines too.
2003-07-04 04:51:34 +02:00
operOid = LookupOperName(item->name, typeoid, typeoid,
false);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
}
/* Caller must have execute permission on operators */
funcOid = get_opcode(operOid);
aclresult = pg_proc_aclcheck(funcOid, GetUserId(),
ACL_EXECUTE);
if (aclresult != ACLCHECK_OK)
Adjust 'permission denied' messages to be more useful and consistent.
2003-08-01 02:15:26 +02:00
aclcheck_error(aclresult, ACL_KIND_PROC,
get_func_name(funcOid));
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
/* Save the info */
member = (OpClassMember *) palloc0(sizeof(OpClassMember));
member->object = operOid;
member->number = item->number;
member->subtype = assignOperSubtype(amoid, typeoid, operOid);
member->recheck = item->recheck;
addClassMember(&operators, member, false);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
break;
case OPCLASS_ITEM_FUNCTION:
if (item->number <= 0 || item->number > numProcs)
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
errmsg("invalid procedure number %d,"
" must be between 1 and %d",
item->number, numProcs)));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
funcOid = LookupFuncNameTypeNames(item->name, item->args,
Some early work on error message editing. Operator-not-found and function-not-found messages now distinguish the cases no-match and ambiguous-match, and they follow the style guidelines too.
2003-07-04 04:51:34 +02:00
false);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/* Caller must have execute permission on functions */
aclresult = pg_proc_aclcheck(funcOid, GetUserId(),
ACL_EXECUTE);
if (aclresult != ACLCHECK_OK)
Adjust 'permission denied' messages to be more useful and consistent.
2003-08-01 02:15:26 +02:00
aclcheck_error(aclresult, ACL_KIND_PROC,
get_func_name(funcOid));
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
/* Save the info */
member = (OpClassMember *) palloc0(sizeof(OpClassMember));
member->object = funcOid;
member->number = item->number;
member->subtype = assignProcSubtype(amoid, typeoid, funcOid);
addClassMember(&procedures, member, true);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
break;
case OPCLASS_ITEM_STORAGETYPE:
if (OidIsValid(storageoid))
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
pgindent run.
2003-08-04 02:43:34 +02:00
errmsg("storage type specified more than once")));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
storageoid = typenameTypeId(item->storedtype);
break;
default:
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
elog(ERROR, "unrecognized item type: %d", item->itemtype);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
break;
}
}
/*
* If storagetype is specified, make sure it's legal.
*/
if (OidIsValid(storageoid))
{
/* Just drop the spec if same as column datatype */
if (storageoid == typeoid)
storageoid = InvalidOid;
else
{
/*
* Currently, only GiST allows storagetype different from
* datatype. This hardcoded test should be eliminated in
* favor of adding another boolean column to pg_am ...
*/
if (amoid != GIST_AM_OID)
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
Message editing: remove gratuitous variations in message wording, standardize terms, add some clarifications, fix some untranslatable attempts at dynamic message building.
2003-09-25 08:58:07 +02:00
errmsg("storage type may not be different from data type for access method \"%s\"",
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
stmt->amname)));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
}
}
rel = heap_openr(OperatorClassRelationName, RowExclusiveLock);
/*
pgindent run.
2002-09-04 22:31:48 +02:00
* Make sure there is no existing opclass of this name (this is just
* to give a more friendly error message than "duplicate key").
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
*/
if (SearchSysCacheExists(CLAAMNAMENSP,
ObjectIdGetDatum(amoid),
CStringGetDatum(opcname),
ObjectIdGetDatum(namespaceoid),
0))
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_DUPLICATE_OBJECT),
Another round of error message editing, covering backend/commands/.
2003-07-20 23:56:35 +02:00
errmsg("operator class \"%s\" for access method \"%s\" already exists",
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
opcname, stmt->amname)));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/*
pgindent run.
2002-09-04 22:31:48 +02:00
* If we are creating a default opclass, check there isn't one
Create a 'type cache' that keeps track of the data needed for any particular datatype by array_eq and array_cmp; use this to solve problems with memory leaks in array indexing support. The parser's equality_oper and ordering_oper routines also use the cache. Change the operator search algorithms to look for appropriate btree or hash index opclasses, instead of assuming operators named '<' or '=' have the right semantics. (ORDER BY ASC/DESC now also look at opclasses, instead of assuming '<' and '>' are the right things.) Add several more index opclasses so that there is no regression in functionality for base datatypes. initdb forced due to catalog additions.
2003-08-17 21:58:06 +02:00
* already. (Note we do not restrict this test to visible opclasses;
* this ensures that typcache.c can find unique solutions to its
* questions.)
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
*/
if (stmt->isDefault)
{
pgindent run.
2002-09-04 22:31:48 +02:00
ScanKeyData skey[1];
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
SysScanDesc scan;
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey[0],
Anum_pg_opclass_opcamid,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(amoid));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
scan = systable_beginscan(rel, OpclassAmNameNspIndex, true,
SnapshotNow, 1, skey);
while (HeapTupleIsValid(tup = systable_getnext(scan)))
{
Form_pg_opclass opclass = (Form_pg_opclass) GETSTRUCT(tup);
if (opclass->opcintype == typeoid && opclass->opcdefault)
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_DUPLICATE_OBJECT),
Various message fixes, among those fixes for the previous round of fixes
2003-09-26 17:27:37 +02:00
errmsg("could not make operator class \"%s\" be default for type %s",
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
opcname,
TypeNameToString(stmt->datatype)),
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
errdetail("Operator class \"%s\" already is the default.",
NameStr(opclass->opcname))));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
}
systable_endscan(scan);
}
/*
* Okay, let's create the pg_opclass entry.
*/
for (i = 0; i < Natts_pg_opclass; ++i)
{
nulls[i] = ' ';
values[i] = (Datum) NULL; /* redundant, but safe */
}
i = 0;
values[i++] = ObjectIdGetDatum(amoid); /* opcamid */
namestrcpy(&opcName, opcname);
values[i++] = NameGetDatum(&opcName); /* opcname */
pgindent run.
2002-09-04 22:31:48 +02:00
values[i++] = ObjectIdGetDatum(namespaceoid); /* opcnamespace */
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
values[i++] = Int32GetDatum(GetUserId()); /* opcowner */
values[i++] = ObjectIdGetDatum(typeoid); /* opcintype */
pgindent run.
2002-09-04 22:31:48 +02:00
values[i++] = BoolGetDatum(stmt->isDefault); /* opcdefault */
values[i++] = ObjectIdGetDatum(storageoid); /* opckeytype */
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
tup = heap_formtuple(rel->rd_att, values, nulls);
opclassoid = simple_heap_insert(rel, tup);
Restructure system-catalog index updating logic. Instead of having hardwired lists of index names for each catalog, use the relcache's mechanism for caching lists of OIDs of indexes of any table. This reduces the common case of updating system catalog indexes to a single line, makes it much easier to add a new system index (in fact, you can now do so on-the-fly if you want to), and as a nice side benefit improves performance a little. Per recent pghackers discussion.
2002-08-05 05:29:17 +02:00
CatalogUpdateIndexes(rel, tup);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
heap_freetuple(tup);
/*
pgindent run.
2002-09-04 22:31:48 +02:00
* Now add tuples to pg_amop and pg_amproc tying in the operators and
* functions.
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
*/
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
storeOperators(opclassoid, operators);
storeProcedures(opclassoid, procedures);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/*
* Create dependencies. Note: we do not create a dependency link to
* the AM, because we don't currently support DROP ACCESS METHOD.
*/
myself.classId = RelationGetRelid(rel);
myself.objectId = opclassoid;
myself.objectSubId = 0;
/* dependency on namespace */
referenced.classId = get_system_catalog_relid(NamespaceRelationName);
referenced.objectId = namespaceoid;
referenced.objectSubId = 0;
recordDependencyOn(&myself, &referenced, DEPENDENCY_NORMAL);
/* dependency on indexed datatype */
referenced.classId = RelOid_pg_type;
referenced.objectId = typeoid;
referenced.objectSubId = 0;
recordDependencyOn(&myself, &referenced, DEPENDENCY_NORMAL);
/* dependency on storage datatype */
if (OidIsValid(storageoid))
{
referenced.classId = RelOid_pg_type;
referenced.objectId = storageoid;
referenced.objectSubId = 0;
recordDependencyOn(&myself, &referenced, DEPENDENCY_NORMAL);
}
/* dependencies on operators */
referenced.classId = get_system_catalog_relid(OperatorRelationName);
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
foreach(l, operators)
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
{
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
OpClassMember *op = (OpClassMember *) lfirst(l);
referenced.objectId = op->object;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
referenced.objectSubId = 0;
recordDependencyOn(&myself, &referenced, DEPENDENCY_NORMAL);
}
/* dependencies on procedures */
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
foreach(l, procedures)
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
{
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
OpClassMember *proc = (OpClassMember *) lfirst(l);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
referenced.classId = RelOid_pg_proc;
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
referenced.objectId = proc->object;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
referenced.objectSubId = 0;
recordDependencyOn(&myself, &referenced, DEPENDENCY_NORMAL);
}
heap_close(rel, RowExclusiveLock);
}
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
/*
* Determine the subtype to assign to an operator, and do any validity
* checking we can manage
*
* Currently this is done using hardwired rules; we don't let the user
* specify it directly.
*/
static Oid
assignOperSubtype(Oid amoid, Oid typeoid, Oid operOid)
{
Oid subtype;
Operator optup;
Form_pg_operator opform;
/* Subtypes are currently only supported by btree, others use 0 */
if (amoid != BTREE_AM_OID)
return InvalidOid;
optup = SearchSysCache(OPEROID,
ObjectIdGetDatum(operOid),
0, 0, 0);
if (optup == NULL)
elog(ERROR, "cache lookup failed for operator %u", operOid);
opform = (Form_pg_operator) GETSTRUCT(optup);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
/*
* btree operators must be binary ops returning boolean, and the
* left-side input type must match the operator class' input type.
*/
if (opform->oprkind != 'b')
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
errmsg("btree operators must be binary")));
if (opform->oprresult != BOOLOID)
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
errmsg("btree operators must return boolean")));
if (opform->oprleft != typeoid)
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
errmsg("btree operators must have index type as left input")));
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
/*
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
* The subtype is "default" (0) if oprright matches the operator
* class, otherwise it is oprright.
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
*/
if (opform->oprright == typeoid)
subtype = InvalidOid;
else
subtype = opform->oprright;
ReleaseSysCache(optup);
return subtype;
}
/*
* Determine the subtype to assign to a support procedure, and do any validity
* checking we can manage
*
* Currently this is done using hardwired rules; we don't let the user
* specify it directly.
*/
static Oid
assignProcSubtype(Oid amoid, Oid typeoid, Oid procOid)
{
Oid subtype;
HeapTuple proctup;
Form_pg_proc procform;
/* Subtypes are currently only supported by btree, others use 0 */
if (amoid != BTREE_AM_OID)
return InvalidOid;
proctup = SearchSysCache(PROCOID,
ObjectIdGetDatum(procOid),
0, 0, 0);
if (proctup == NULL)
elog(ERROR, "cache lookup failed for function %u", procOid);
procform = (Form_pg_proc) GETSTRUCT(proctup);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
/*
* btree support procs must be 2-arg procs returning int4, and the
* first input type must match the operator class' input type.
*/
if (procform->pronargs != 2)
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
errmsg("btree procedures must have two arguments")));
if (procform->prorettype != INT4OID)
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
errmsg("btree procedures must return integer")));
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
if (procform->proargtypes.values[0] != typeoid)
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
errmsg("btree procedures must have index type as first input")));
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
/*
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
* The subtype is "default" (0) if second input type matches the
* operator class, otherwise it is the second input type.
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
*/
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
if (procform->proargtypes.values[1] == typeoid)
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
subtype = InvalidOid;
else
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
subtype = procform->proargtypes.values[1];
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ReleaseSysCache(proctup);
return subtype;
}
/*
* Add a new class member to the appropriate list, after checking for
* duplicated strategy or proc number.
*/
static void
addClassMember(List **list, OpClassMember *member, bool isProc)
{
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
ListCell *l;
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
foreach(l, *list)
{
OpClassMember *old = (OpClassMember *) lfirst(l);
if (old->number == member->number &&
old->subtype == member->subtype)
{
if (isProc)
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
errmsg("procedure number %d appears more than once",
member->number)));
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
else
ereport(ERROR,
(errcode(ERRCODE_INVALID_OBJECT_DEFINITION),
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
errmsg("operator number %d appears more than once",
member->number)));
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
}
}
*list = lappend(*list, member);
}
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/*
* Dump the operators to pg_amop
*/
static void
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
storeOperators(Oid opclassoid, List *operators)
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
{
pgindent run.
2002-09-04 22:31:48 +02:00
Relation rel;
Datum values[Natts_pg_amop];
char nulls[Natts_pg_amop];
HeapTuple tup;
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
ListCell *l;
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
int i;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
rel = heap_openr(AccessMethodOperatorRelationName, RowExclusiveLock);
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
foreach(l, operators)
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
{
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
OpClassMember *op = (OpClassMember *) lfirst(l);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
for (i = 0; i < Natts_pg_amop; ++i)
{
nulls[i] = ' ';
values[i] = (Datum) NULL;
}
i = 0;
pgindent run.
2002-09-04 22:31:48 +02:00
values[i++] = ObjectIdGetDatum(opclassoid); /* amopclaid */
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
values[i++] = ObjectIdGetDatum(op->subtype); /* amopsubtype */
values[i++] = Int16GetDatum(op->number); /* amopstrategy */
values[i++] = BoolGetDatum(op->recheck); /* amopreqcheck */
values[i++] = ObjectIdGetDatum(op->object); /* amopopr */
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
tup = heap_formtuple(rel->rd_att, values, nulls);
simple_heap_insert(rel, tup);
Restructure system-catalog index updating logic. Instead of having hardwired lists of index names for each catalog, use the relcache's mechanism for caching lists of OIDs of indexes of any table. This reduces the common case of updating system catalog indexes to a single line, makes it much easier to add a new system index (in fact, you can now do so on-the-fly if you want to), and as a nice side benefit improves performance a little. Per recent pghackers discussion.
2002-08-05 05:29:17 +02:00
CatalogUpdateIndexes(rel, tup);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
heap_freetuple(tup);
}
heap_close(rel, RowExclusiveLock);
}
/*
* Dump the procedures (support routines) to pg_amproc
*/
static void
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
storeProcedures(Oid opclassoid, List *procedures)
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
{
pgindent run.
2002-09-04 22:31:48 +02:00
Relation rel;
Datum values[Natts_pg_amproc];
char nulls[Natts_pg_amproc];
HeapTuple tup;
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
ListCell *l;
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
int i;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
rel = heap_openr(AccessMethodProcedureRelationName, RowExclusiveLock);
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
foreach(l, procedures)
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
{
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
OpClassMember *proc = (OpClassMember *) lfirst(l);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
for (i = 0; i < Natts_pg_amproc; ++i)
{
nulls[i] = ' ';
values[i] = (Datum) NULL;
}
i = 0;
pgindent run.
2002-09-04 22:31:48 +02:00
values[i++] = ObjectIdGetDatum(opclassoid); /* amopclaid */
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
values[i++] = ObjectIdGetDatum(proc->subtype); /* amprocsubtype */
values[i++] = Int16GetDatum(proc->number); /* amprocnum */
values[i++] = ObjectIdGetDatum(proc->object); /* amproc */
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
tup = heap_formtuple(rel->rd_att, values, nulls);
simple_heap_insert(rel, tup);
Restructure system-catalog index updating logic. Instead of having hardwired lists of index names for each catalog, use the relcache's mechanism for caching lists of OIDs of indexes of any table. This reduces the common case of updating system catalog indexes to a single line, makes it much easier to add a new system index (in fact, you can now do so on-the-fly if you want to), and as a nice side benefit improves performance a little. Per recent pghackers discussion.
2002-08-05 05:29:17 +02:00
CatalogUpdateIndexes(rel, tup);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
heap_freetuple(tup);
}
heap_close(rel, RowExclusiveLock);
}
/*
* RemoveOpClass
* Deletes an opclass.
*/
void
RemoveOpClass(RemoveOpClassStmt *stmt)
{
pgindent run.
2002-09-04 22:31:48 +02:00
Oid amID,
opcID;
Centralize code for interpreting schema references, which had gotten copied more places than I first thought it would. This fixes a bug: a couple of these places were neglecting to enforce USAGE access on explicitly-referenced schemas.
2002-07-30 01:46:35 +02:00
char *schemaname;
char *opcname;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
HeapTuple tuple;
ObjectAddress object;
/*
pgindent run.
2002-09-04 22:31:48 +02:00
* Get the access method's OID.
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
*/
amID = GetSysCacheOid(AMNAME,
CStringGetDatum(stmt->amname),
0, 0, 0);
if (!OidIsValid(amID))
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("access method \"%s\" does not exist",
stmt->amname)));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/*
* Look up the opclass.
*/
/* deconstruct the name list */
Centralize code for interpreting schema references, which had gotten copied more places than I first thought it would. This fixes a bug: a couple of these places were neglecting to enforce USAGE access on explicitly-referenced schemas.
2002-07-30 01:46:35 +02:00
DeconstructQualifiedName(stmt->opclassname, &schemaname, &opcname);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
if (schemaname)
{
/* Look in specific schema only */
pgindent run.
2002-09-04 22:31:48 +02:00
Oid namespaceId;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
Centralize code for interpreting schema references, which had gotten copied more places than I first thought it would. This fixes a bug: a couple of these places were neglecting to enforce USAGE access on explicitly-referenced schemas.
2002-07-30 01:46:35 +02:00
namespaceId = LookupExplicitNamespace(schemaname);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
tuple = SearchSysCache(CLAAMNAMENSP,
ObjectIdGetDatum(amID),
PointerGetDatum(opcname),
ObjectIdGetDatum(namespaceId),
0);
}
else
{
/* Unqualified opclass name, so search the search path */
opcID = OpclassnameGetOpcid(amID, opcname);
if (!OidIsValid(opcID))
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("operator class \"%s\" does not exist for access method \"%s\"",
opcname, stmt->amname)));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
tuple = SearchSysCache(CLAOID,
ObjectIdGetDatum(opcID),
0, 0, 0);
}
if (!HeapTupleIsValid(tuple))
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("operator class \"%s\" does not exist for access method \"%s\"",
pgindent run.
2003-08-04 02:43:34 +02:00
NameListToString(stmt->opclassname), stmt->amname)));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
opcID = HeapTupleGetOid(tuple);
/* Permission check: must own opclass or its namespace */
if (!pg_opclass_ownercheck(opcID, GetUserId()) &&
!pg_namespace_ownercheck(((Form_pg_opclass) GETSTRUCT(tuple))->opcnamespace,
GetUserId()))
Adjust 'permission denied' messages to be more useful and consistent.
2003-08-01 02:15:26 +02:00
aclcheck_error(ACLCHECK_NOT_OWNER, ACL_KIND_OPCLASS,
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
NameListToString(stmt->opclassname));
ReleaseSysCache(tuple);
/*
* Do the deletion
*/
object.classId = get_system_catalog_relid(OperatorClassRelationName);
object.objectId = opcID;
object.objectSubId = 0;
performDeletion(&object, stmt->behavior);
}
/*
* Guts of opclass deletion.
*/
void
RemoveOpClassById(Oid opclassOid)
{
pgindent run.
2002-09-04 22:31:48 +02:00
Relation rel;
HeapTuple tup;
ScanKeyData skey[1];
SysScanDesc scan;
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
/*
* First remove the pg_opclass entry itself.
*/
rel = heap_openr(OperatorClassRelationName, RowExclusiveLock);
tup = SearchSysCache(CLAOID,
ObjectIdGetDatum(opclassOid),
0, 0, 0);
pgindent run.
2002-09-04 22:31:48 +02:00
if (!HeapTupleIsValid(tup)) /* should not happen */
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
elog(ERROR, "cache lookup failed for opclass %u", opclassOid);
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
simple_heap_delete(rel, &tup->t_self);
ReleaseSysCache(tup);
heap_close(rel, RowExclusiveLock);
/*
* Remove associated entries in pg_amop.
*/
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey[0],
Anum_pg_amop_amopclaid,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(opclassOid));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
rel = heap_openr(AccessMethodOperatorRelationName, RowExclusiveLock);
scan = systable_beginscan(rel, AccessMethodStrategyIndex, true,
SnapshotNow, 1, skey);
while (HeapTupleIsValid(tup = systable_getnext(scan)))
simple_heap_delete(rel, &tup->t_self);
systable_endscan(scan);
heap_close(rel, RowExclusiveLock);
/*
* Remove associated entries in pg_amproc.
*/
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey[0],
Anum_pg_amproc_amopclaid,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(opclassOid));
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more documentation (xindex.sgml should be rewritten), need to teach pg_dump about it, need to update contrib modules that currently build pg_opclass entries by hand. Original patch by Bill Studenmund, grammar adjustments and general update for 7.3 by Tom Lane.
2002-07-30 00:14:11 +02:00
rel = heap_openr(AccessMethodProcedureRelationName, RowExclusiveLock);
scan = systable_beginscan(rel, AccessMethodProcedureIndex, true,
SnapshotNow, 1, skey);
while (HeapTupleIsValid(tup = systable_getnext(scan)))
simple_heap_delete(rel, &tup->t_self);
systable_endscan(scan);
heap_close(rel, RowExclusiveLock);
}
First batch of object rename commands.
2003-06-27 16:45:32 +02:00
/*
* Rename opclass
*/
void
RenameOpClass(List *name, const char *access_method, const char *newname)
{
Oid opcOid;
Oid amOid;
Oid namespaceOid;
char *schemaname;
char *opcname;
HeapTuple tup;
Relation rel;
AclResult aclresult;
amOid = GetSysCacheOid(AMNAME,
CStringGetDatum(access_method),
0, 0, 0);
if (!OidIsValid(amOid))
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("access method \"%s\" does not exist",
access_method)));
First batch of object rename commands.
2003-06-27 16:45:32 +02:00
rel = heap_openr(OperatorClassRelationName, RowExclusiveLock);
/*
* Look up the opclass
*/
DeconstructQualifiedName(name, &schemaname, &opcname);
if (schemaname)
{
namespaceOid = LookupExplicitNamespace(schemaname);
tup = SearchSysCacheCopy(CLAAMNAMENSP,
ObjectIdGetDatum(amOid),
PointerGetDatum(opcname),
ObjectIdGetDatum(namespaceOid),
0);
if (!HeapTupleIsValid(tup))
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("operator class \"%s\" does not exist for access method \"%s\"",
opcname, access_method)));
First batch of object rename commands.
2003-06-27 16:45:32 +02:00
opcOid = HeapTupleGetOid(tup);
}
else
{
opcOid = OpclassnameGetOpcid(amOid, opcname);
if (!OidIsValid(opcOid))
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("operator class \"%s\" does not exist for access method \"%s\"",
opcname, access_method)));
First batch of object rename commands.
2003-06-27 16:45:32 +02:00
tup = SearchSysCacheCopy(CLAOID,
ObjectIdGetDatum(opcOid),
0, 0, 0);
pgindent run.
2003-08-04 02:43:34 +02:00
if (!HeapTupleIsValid(tup)) /* should not happen */
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
elog(ERROR, "cache lookup failed for opclass %u", opcOid);
First batch of object rename commands.
2003-06-27 16:45:32 +02:00
namespaceOid = ((Form_pg_opclass) GETSTRUCT(tup))->opcnamespace;
}
/* make sure the new name doesn't exist */
if (SearchSysCacheExists(CLAAMNAMENSP,
ObjectIdGetDatum(amOid),
CStringGetDatum(newname),
ObjectIdGetDatum(namespaceOid),
0))
{
ereport(ERROR,
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
(errcode(ERRCODE_DUPLICATE_OBJECT),
First batch of object rename commands.
2003-06-27 16:45:32 +02:00
errmsg("operator class \"%s\" for access method \"%s\" already exists in schema \"%s\"",
First bits of work on error message editing.
2003-07-19 01:20:33 +02:00
newname, access_method,
get_namespace_name(namespaceOid))));
First batch of object rename commands.
2003-06-27 16:45:32 +02:00
}
/* must be owner */
if (!pg_opclass_ownercheck(opcOid, GetUserId()))
Adjust 'permission denied' messages to be more useful and consistent.
2003-08-01 02:15:26 +02:00
aclcheck_error(ACLCHECK_NOT_OWNER, ACL_KIND_OPCLASS,
NameListToString(name));
First batch of object rename commands.
2003-06-27 16:45:32 +02:00
/* must have CREATE privilege on namespace */
aclresult = pg_namespace_aclcheck(namespaceOid, GetUserId(), ACL_CREATE);
if (aclresult != ACLCHECK_OK)
Adjust 'permission denied' messages to be more useful and consistent.
2003-08-01 02:15:26 +02:00
aclcheck_error(aclresult, ACL_KIND_NAMESPACE,
get_namespace_name(namespaceOid));
First batch of object rename commands.
2003-06-27 16:45:32 +02:00
/* rename */
namestrcpy(&(((Form_pg_opclass) GETSTRUCT(tup))->opcname), newname);
simple_heap_update(rel, &tup->t_self, tup);
CatalogUpdateIndexes(rel, tup);
heap_close(rel, NoLock);
heap_freetuple(tup);
}
Support renaming of tablespaces, and changing the owners of aggregates, conversions, functions, operators, operator classes, schemas, types, and tablespaces. Fold the existing implementations of alter domain owner and alter database owner in with these. Christopher Kings-Lynne
2004-06-25 23:55:59 +02:00
/*
* Change opclass owner
*/
void
AlterOpClassOwner(List *name, const char *access_method, AclId newOwnerSysId)
{
Oid opcOid;
Oid amOid;
Oid namespaceOid;
char *schemaname;
char *opcname;
HeapTuple tup;
Relation rel;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
Form_pg_opclass opcForm;
Support renaming of tablespaces, and changing the owners of aggregates, conversions, functions, operators, operator classes, schemas, types, and tablespaces. Fold the existing implementations of alter domain owner and alter database owner in with these. Christopher Kings-Lynne
2004-06-25 23:55:59 +02:00
amOid = GetSysCacheOid(AMNAME,
CStringGetDatum(access_method),
0, 0, 0);
if (!OidIsValid(amOid))
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("access method \"%s\" does not exist",
access_method)));
rel = heap_openr(OperatorClassRelationName, RowExclusiveLock);
/*
* Look up the opclass
*/
DeconstructQualifiedName(name, &schemaname, &opcname);
if (schemaname)
{
namespaceOid = LookupExplicitNamespace(schemaname);
tup = SearchSysCacheCopy(CLAAMNAMENSP,
ObjectIdGetDatum(amOid),
PointerGetDatum(opcname),
ObjectIdGetDatum(namespaceOid),
0);
if (!HeapTupleIsValid(tup))
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("operator class \"%s\" does not exist for access method \"%s\"",
opcname, access_method)));
}
else
{
opcOid = OpclassnameGetOpcid(amOid, opcname);
if (!OidIsValid(opcOid))
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_OBJECT),
errmsg("operator class \"%s\" does not exist for access method \"%s\"",
opcname, access_method)));
tup = SearchSysCacheCopy(CLAOID,
ObjectIdGetDatum(opcOid),
0, 0, 0);
if (!HeapTupleIsValid(tup)) /* should not happen */
elog(ERROR, "cache lookup failed for opclass %u", opcOid);
namespaceOid = ((Form_pg_opclass) GETSTRUCT(tup))->opcnamespace;
}
opcForm = (Form_pg_opclass) GETSTRUCT(tup);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
/*
Support renaming of tablespaces, and changing the owners of aggregates, conversions, functions, operators, operator classes, schemas, types, and tablespaces. Fold the existing implementations of alter domain owner and alter database owner in with these. Christopher Kings-Lynne
2004-06-25 23:55:59 +02:00
* If the new owner is the same as the existing owner, consider the
* command to have succeeded. This is for dump restoration purposes.
*/
if (opcForm->opcowner != newOwnerSysId)
{
/* Otherwise, must be superuser to change object ownership */
if (!superuser())
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
errmsg("must be superuser to change owner")));
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
/*
* Modify the owner --- okay to scribble on tup because it's a
* copy
*/
Support renaming of tablespaces, and changing the owners of aggregates, conversions, functions, operators, operator classes, schemas, types, and tablespaces. Fold the existing implementations of alter domain owner and alter database owner in with these. Christopher Kings-Lynne
2004-06-25 23:55:59 +02:00
opcForm->opcowner = newOwnerSysId;
simple_heap_update(rel, &tup->t_self, tup);
CatalogUpdateIndexes(rel, tup);
}
heap_close(rel, NoLock);
heap_freetuple(tup);
}