postgresql/contrib/isn/isn.c

1140 lines
24 KiB
C
Raw Normal View History

/*-------------------------------------------------------------------------
*
* isn.c
* PostgreSQL type definitions for ISNs (ISBN, ISMN, ISSN, EAN13, UPC)
*
2010-02-26 03:01:40 +01:00
* Author: German Mendez Bravo (Kronuz)
* Portions Copyright (c) 1996-2019, PostgreSQL Global Development Group
*
* IDENTIFICATION
2010-09-20 22:08:53 +02:00
* contrib/isn/isn.c
*
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include "EAN13.h"
#include "ISBN.h"
#include "ISMN.h"
#include "ISSN.h"
#include "UPC.h"
#include "fmgr.h"
#include "isn.h"
#include "utils/builtins.h"
PG_MODULE_MAGIC;
#ifdef USE_ASSERT_CHECKING
#define ISN_DEBUG 1
#else
#define ISN_DEBUG 0
#endif
#define MAXEAN13LEN 18
2006-10-04 02:30:14 +02:00
enum isn_type
{
INVALID, ANY, EAN13, ISBN, ISMN, ISSN, UPC
};
static const char *const isn_names[] = {"EAN13/UPC/ISxN", "EAN13/UPC/ISxN", "EAN13", "ISBN", "ISMN", "ISSN", "UPC"};
static bool g_weak = false;
/***********************************************************************
**
** Routines for EAN13/UPC/ISxNs.
**
** Note:
2006-10-04 02:30:14 +02:00
** In this code, a normalized string is one that is known to be a valid
** ISxN number containing only digits and hyphens and with enough space
** to hold the full 13 digits plus the maximum of four hyphens.
***********************************************************************/
/*----------------------------------------------------------
* Debugging routines.
*---------------------------------------------------------*/
/*
* Check if the table and its index is correct (just for debugging)
*/
pg_attribute_unused()
static bool
2006-10-04 02:30:14 +02:00
check_table(const char *(*TABLE)[2], const unsigned TABLE_index[10][2])
{
2006-10-04 02:30:14 +02:00
const char *aux1,
*aux2;
int a,
b,
x = 0,
y = -1,
i = 0,
j,
init = 0;
if (TABLE == NULL || TABLE_index == NULL)
return true;
while (TABLE[i][0] && TABLE[i][1])
{
aux1 = TABLE[i][0];
aux2 = TABLE[i][1];
/* must always start with a digit: */
if (!isdigit((unsigned char) *aux1) || !isdigit((unsigned char) *aux2))
goto invalidtable;
a = *aux1 - '0';
b = *aux2 - '0';
/* must always have the same format and length: */
2006-10-04 02:30:14 +02:00
while (*aux1 && *aux2)
{
if (!(isdigit((unsigned char) *aux1) &&
isdigit((unsigned char) *aux2)) &&
2006-10-04 02:30:14 +02:00
(*aux1 != *aux2 || *aux1 != '-'))
goto invalidtable;
aux1++;
aux2++;
}
2006-10-04 02:30:14 +02:00
if (*aux1 != *aux2)
goto invalidtable;
/* found a new range */
2006-10-04 02:30:14 +02:00
if (a > y)
{
/* check current range in the index: */
2006-10-04 02:30:14 +02:00
for (j = x; j <= y; j++)
{
if (TABLE_index[j][0] != init)
goto invalidindex;
if (TABLE_index[j][1] != i - init)
goto invalidindex;
}
init = i;
x = a;
2006-10-04 02:30:14 +02:00
}
/* Always get the new limit */
y = b;
2006-10-04 02:30:14 +02:00
if (y < x)
goto invalidtable;
i++;
}
return true;
invalidtable:
elog(DEBUG1, "invalid table near {\"%s\", \"%s\"} (pos: %d)",
TABLE[i][0], TABLE[i][1], i);
return false;
invalidindex:
elog(DEBUG1, "index %d is invalid", j);
return false;
}
/*----------------------------------------------------------
* Formatting and conversion routines.
*---------------------------------------------------------*/
static unsigned
2006-10-04 02:30:14 +02:00
dehyphenate(char *bufO, char *bufI)
{
2006-10-04 02:30:14 +02:00
unsigned ret = 0;
while (*bufI)
{
if (isdigit((unsigned char) *bufI))
{
*bufO++ = *bufI;
ret++;
}
bufI++;
}
*bufO = '\0';
return ret;
}
/*
2006-10-04 02:30:14 +02:00
* hyphenate --- Try to hyphenate, in-place, the string starting at bufI
* into bufO using the given hyphenation range TABLE.
* Assumes the input string to be used is of only digits.
*
* Returns the number of characters actually hyphenated.
*/
static unsigned
2006-10-04 02:30:14 +02:00
hyphenate(char *bufO, char *bufI, const char *(*TABLE)[2], const unsigned TABLE_index[10][2])
{
2006-10-04 02:30:14 +02:00
unsigned ret = 0;
const char *ean_aux1,
*ean_aux2,
*ean_p;
char *firstdig,
*aux1,
*aux2;
unsigned search,
upper,
lower,
step;
bool ean_in1,
ean_in2;
/* just compress the string if no further hyphenation is required */
2006-10-04 02:30:14 +02:00
if (TABLE == NULL || TABLE_index == NULL)
{
while (*bufI)
{
*bufO++ = *bufI++;
ret++;
}
*bufO = '\0';
2006-10-04 02:30:14 +02:00
return (ret + 1);
}
/* add remaining hyphenations */
search = *bufI - '0';
upper = lower = TABLE_index[search][0];
upper += TABLE_index[search][1];
lower--;
step = (upper - lower) / 2;
2006-10-04 02:30:14 +02:00
if (step == 0)
return 0;
search = lower + step;
firstdig = bufI;
ean_in1 = ean_in2 = false;
ean_aux1 = TABLE[search][0];
ean_aux2 = TABLE[search][1];
2006-10-04 02:30:14 +02:00
do
{
if ((ean_in1 || *firstdig >= *ean_aux1) && (ean_in2 || *firstdig <= *ean_aux2))
{
if (*firstdig > *ean_aux1)
ean_in1 = true;
if (*firstdig < *ean_aux2)
ean_in2 = true;
if (ean_in1 && ean_in2)
break;
firstdig++, ean_aux1++, ean_aux2++;
2006-10-04 02:30:14 +02:00
if (!(*ean_aux1 && *ean_aux2 && *firstdig))
break;
if (!isdigit((unsigned char) *ean_aux1))
ean_aux1++, ean_aux2++;
}
else
{
/*
* check in what direction we should go and move the pointer
* accordingly
*/
if (*firstdig < *ean_aux1 && !ean_in1)
upper = search;
else
lower = search;
step = (upper - lower) / 2;
search = lower + step;
/* Initialize stuff again: */
firstdig = bufI;
ean_in1 = ean_in2 = false;
ean_aux1 = TABLE[search][0];
ean_aux2 = TABLE[search][1];
}
2006-10-04 02:30:14 +02:00
} while (step);
if (step)
{
aux1 = bufO;
aux2 = bufI;
ean_p = TABLE[search][0];
2006-10-04 02:30:14 +02:00
while (*ean_p && *aux2)
{
if (*ean_p++ != '-')
*aux1++ = *aux2++;
else
*aux1++ = '-';
ret++;
}
2006-10-04 02:30:14 +02:00
*aux1++ = '-';
*aux1 = *aux2; /* add a lookahead char */
return (ret + 1);
}
return ret;
}
/*
2006-10-04 02:30:14 +02:00
* weight_checkdig -- Receives a buffer with a normalized ISxN string number,
* and the length to weight.
*
* Returns the weight of the number (the check digit value, 0-10)
*/
static unsigned
2006-10-04 02:30:14 +02:00
weight_checkdig(char *isn, unsigned size)
{
2006-10-04 02:30:14 +02:00
unsigned weight = 0;
while (*isn && size > 1)
{
if (isdigit((unsigned char) *isn))
{
weight += size-- * (*isn - '0');
}
isn++;
}
weight = weight % 11;
2006-10-04 02:30:14 +02:00
if (weight != 0)
weight = 11 - weight;
return weight;
}
/*
2006-10-04 02:30:14 +02:00
* checkdig --- Receives a buffer with a normalized ISxN string number,
* and the length to check.
*
* Returns the check digit value (0-9)
*/
static unsigned
2006-10-04 02:30:14 +02:00
checkdig(char *num, unsigned size)
{
2006-10-04 02:30:14 +02:00
unsigned check = 0,
check3 = 0;
unsigned pos = 0;
if (*num == 'M')
{ /* ISMN start with 'M' */
check3 = 3;
pos = 1;
}
2006-10-04 02:30:14 +02:00
while (*num && size > 1)
{
if (isdigit((unsigned char) *num))
{
if (pos++ % 2)
check3 += *num - '0';
else
check += *num - '0';
size--;
}
num++;
}
2006-10-04 02:30:14 +02:00
check = (check + 3 * check3) % 10;
if (check != 0)
check = 10 - check;
return check;
}
/*
* ean2isn --- Try to convert an ean13 number to a UPC/ISxN number.
2006-10-04 02:30:14 +02:00
* This doesn't verify for a valid check digit.
*
* If errorOK is false, ereport a useful error message if the ean13 is bad.
* If errorOK is true, just return "false" for bad input.
*/
static bool
ean2isn(ean13 ean, bool errorOK, ean13 *result, enum isn_type accept)
{
enum isn_type type = INVALID;
2006-10-04 02:30:14 +02:00
char buf[MAXEAN13LEN + 1];
char *aux;
2006-10-04 02:30:14 +02:00
unsigned digval;
unsigned search;
ean13 ret = ean;
ean >>= 1;
/* verify it's in the EAN13 range */
2006-10-04 02:30:14 +02:00
if (ean > UINT64CONST(9999999999999))
goto eantoobig;
/* convert the number */
search = 0;
aux = buf + 13;
2006-10-04 02:30:14 +02:00
*aux = '\0'; /* terminate string; aux points to last digit */
do
{
digval = (unsigned) (ean % 10); /* get the decimal value */
ean /= 10; /* get next digit */
*--aux = (char) (digval + '0'); /* convert to ascii and store */
} while (ean && search++ < 12);
while (search++ < 12)
*--aux = '0'; /* fill the remaining EAN13 with '0' */
/* find out the data type: */
if (strncmp("978", buf, 3) == 0)
2006-10-04 02:30:14 +02:00
{ /* ISBN */
type = ISBN;
2006-10-04 02:30:14 +02:00
}
else if (strncmp("977", buf, 3) == 0)
2006-10-04 02:30:14 +02:00
{ /* ISSN */
type = ISSN;
2006-10-04 02:30:14 +02:00
}
else if (strncmp("9790", buf, 4) == 0)
2006-10-04 02:30:14 +02:00
{ /* ISMN */
type = ISMN;
2006-10-04 02:30:14 +02:00
}
else if (strncmp("979", buf, 3) == 0)
2006-10-04 02:30:14 +02:00
{ /* ISBN-13 */
type = ISBN;
2006-10-04 02:30:14 +02:00
}
else if (*buf == '0')
{ /* UPC */
type = UPC;
2006-10-04 02:30:14 +02:00
}
else
{
type = EAN13;
}
2006-10-04 02:30:14 +02:00
if (accept != ANY && accept != EAN13 && accept != type)
goto eanwrongtype;
*result = ret;
return true;
2006-10-04 02:30:14 +02:00
eanwrongtype:
2006-10-04 02:30:14 +02:00
if (!errorOK)
{
if (type != EAN13)
{
ereport(ERROR,
2006-10-04 02:30:14 +02:00
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("cannot cast EAN13(%s) to %s for number: \"%s\"",
isn_names[type], isn_names[accept], buf)));
}
else
{
ereport(ERROR,
2006-10-04 02:30:14 +02:00
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("cannot cast %s to %s for number: \"%s\"",
isn_names[type], isn_names[accept], buf)));
}
}
return false;
eantoobig:
2006-10-04 02:30:14 +02:00
if (!errorOK)
{
char eanbuf[64];
/*
2006-10-04 02:30:14 +02:00
* Format the number separately to keep the machine-dependent format
* code out of the translatable message text
*/
snprintf(eanbuf, sizeof(eanbuf), EAN13_FORMAT, ean);
ereport(ERROR,
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
errmsg("value \"%s\" is out of range for %s type",
eanbuf, isn_names[type])));
}
return false;
}
/*
2006-10-04 02:30:14 +02:00
* ean2UPC/ISxN --- Convert in-place a normalized EAN13 string to the corresponding
* UPC/ISxN string number. Assumes the input string is normalized.
*/
static inline void
2006-10-04 02:30:14 +02:00
ean2ISBN(char *isn)
{
2006-10-04 02:30:14 +02:00
char *aux;
unsigned check;
/*
2016-06-10 00:02:36 +02:00
* The number should come in this format: 978-0-000-00000-0 or may be an
* ISBN-13 number, 979-..., which does not have a short representation. Do
* the short output version if possible.
*/
if (strncmp("978-", isn, 4) == 0)
{
/* Strip the first part and calculate the new check digit */
hyphenate(isn, isn + 4, NULL, NULL);
check = weight_checkdig(isn, 10);
aux = strchr(isn, '\0');
while (!isdigit((unsigned char) *--aux));
if (check == 10)
*aux = 'X';
else
*aux = check + '0';
}
}
static inline void
2006-10-04 02:30:14 +02:00
ean2ISMN(char *isn)
{
/* the number should come in this format: 979-0-000-00000-0 */
/* Just strip the first part and change the first digit ('0') to 'M' */
2006-10-04 02:30:14 +02:00
hyphenate(isn, isn + 4, NULL, NULL);
isn[0] = 'M';
}
static inline void
2006-10-04 02:30:14 +02:00
ean2ISSN(char *isn)
{
2006-10-04 02:30:14 +02:00
unsigned check;
/* the number should come in this format: 977-0000-000-00-0 */
/* Strip the first part, crop, and calculate the new check digit */
2006-10-04 02:30:14 +02:00
hyphenate(isn, isn + 4, NULL, NULL);
check = weight_checkdig(isn, 8);
2006-10-04 02:30:14 +02:00
if (check == 10)
isn[8] = 'X';
else
isn[8] = check + '0';
isn[9] = '\0';
}
static inline void
2006-10-04 02:30:14 +02:00
ean2UPC(char *isn)
{
/* the number should come in this format: 000-000000000-0 */
/* Strip the first part, crop, and dehyphenate */
2006-10-04 02:30:14 +02:00
dehyphenate(isn, isn + 1);
isn[12] = '\0';
}
/*
* ean2* --- Converts a string of digits into an ean13 number.
2006-10-04 02:30:14 +02:00
* Assumes the input string is a string with only digits
* on it, and that it's within the range of ean13.
*
* Returns the ean13 value of the string.
*/
static ean13
2006-10-04 02:30:14 +02:00
str2ean(const char *num)
{
2006-10-04 02:30:14 +02:00
ean13 ean = 0; /* current ean */
while (*num)
{
if (isdigit((unsigned char) *num))
ean = 10 * ean + (*num - '0');
num++;
}
2006-10-04 02:30:14 +02:00
return (ean << 1); /* also give room to a flag */
}
/*
* ean2string --- Try to convert an ean13 number to a hyphenated string.
2006-10-04 02:30:14 +02:00
* Assumes there's enough space in result to hold
* the string (maximum MAXEAN13LEN+1 bytes)
* This doesn't verify for a valid check digit.
*
* If shortType is true, the returned string is in the old ISxN short format.
* If errorOK is false, ereport a useful error message if the string is bad.
* If errorOK is true, just return "false" for bad input.
*/
static bool
2006-10-04 02:30:14 +02:00
ean2string(ean13 ean, bool errorOK, char *result, bool shortType)
{
const char *(*TABLE)[2];
const unsigned (*TABLE_index)[2];
enum isn_type type = INVALID;
char *aux;
2006-10-04 02:30:14 +02:00
unsigned digval;
unsigned search;
char valid = '\0'; /* was the number initially written with a
* valid check digit? */
TABLE_index = ISBN_index;
2006-10-04 02:30:14 +02:00
if ((ean & 1) != 0)
valid = '!';
ean >>= 1;
/* verify it's in the EAN13 range */
2006-10-04 02:30:14 +02:00
if (ean > UINT64CONST(9999999999999))
goto eantoobig;
/* convert the number */
search = 0;
aux = result + MAXEAN13LEN;
2006-10-04 02:30:14 +02:00
*aux = '\0'; /* terminate string; aux points to last digit */
*--aux = valid; /* append '!' for numbers with invalid but
* corrected check digit */
do
{
digval = (unsigned) (ean % 10); /* get the decimal value */
ean /= 10; /* get next digit */
*--aux = (char) (digval + '0'); /* convert to ascii and store */
if (search == 0)
*--aux = '-'; /* the check digit is always there */
} while (ean && search++ < 13);
while (search++ < 13)
*--aux = '0'; /* fill the remaining EAN13 with '0' */
/* The string should be in this form: ???DDDDDDDDDDDD-D" */
2006-10-04 02:30:14 +02:00
search = hyphenate(result, result + 3, EAN13_range, EAN13_index);
/* verify it's a logically valid EAN13 */
2006-10-04 02:30:14 +02:00
if (search == 0)
{
search = hyphenate(result, result + 3, NULL, NULL);
goto okay;
}
/* find out what type of hyphenation is needed: */
if (strncmp("978-", result, search) == 0)
{ /* ISBN -13 978-range */
/* The string should be in this form: 978-??000000000-0" */
type = ISBN;
TABLE = ISBN_range;
TABLE_index = ISBN_index;
2006-10-04 02:30:14 +02:00
}
else if (strncmp("977-", result, search) == 0)
2006-10-04 02:30:14 +02:00
{ /* ISSN */
/* The string should be in this form: 977-??000000000-0" */
type = ISSN;
TABLE = ISSN_range;
TABLE_index = ISSN_index;
2006-10-04 02:30:14 +02:00
}
else if (strncmp("979-0", result, search + 1) == 0)
2006-10-04 02:30:14 +02:00
{ /* ISMN */
/* The string should be in this form: 979-0?000000000-0" */
type = ISMN;
TABLE = ISMN_range;
TABLE_index = ISMN_index;
2006-10-04 02:30:14 +02:00
}
else if (strncmp("979-", result, search) == 0)
{ /* ISBN-13 979-range */
/* The string should be in this form: 979-??000000000-0" */
type = ISBN;
TABLE = ISBN_range_new;
TABLE_index = ISBN_index_new;
}
2006-10-04 02:30:14 +02:00
else if (*result == '0')
{ /* UPC */
/* The string should be in this form: 000-00000000000-0" */
type = UPC;
TABLE = UPC_range;
TABLE_index = UPC_index;
2006-10-04 02:30:14 +02:00
}
else
{
type = EAN13;
TABLE = NULL;
TABLE_index = NULL;
}
/* verify it's a logically valid EAN13/UPC/ISxN */
digval = search;
2006-10-04 02:30:14 +02:00
search = hyphenate(result + digval, result + digval + 2, TABLE, TABLE_index);
/* verify it's a valid EAN13 */
2006-10-04 02:30:14 +02:00
if (search == 0)
{
search = hyphenate(result + digval, result + digval + 2, NULL, NULL);
goto okay;
}
okay:
/* convert to the old short type: */
2006-10-04 02:30:14 +02:00
if (shortType)
switch (type)
{
case ISBN:
ean2ISBN(result);
break;
case ISMN:
ean2ISMN(result);
break;
case ISSN:
ean2ISSN(result);
break;
case UPC:
ean2UPC(result);
break;
default:
break;
}
return true;
eantoobig:
2006-10-04 02:30:14 +02:00
if (!errorOK)
{
2006-10-04 02:30:14 +02:00
char eanbuf[64];
/*
2006-10-04 02:30:14 +02:00
* Format the number separately to keep the machine-dependent format
* code out of the translatable message text
*/
snprintf(eanbuf, sizeof(eanbuf), EAN13_FORMAT, ean);
ereport(ERROR,
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
errmsg("value \"%s\" is out of range for %s type",
eanbuf, isn_names[type])));
}
return false;
}
/*
* string2ean --- try to parse a string into an ean13.
*
* If errorOK is false, ereport a useful error message if the string is bad.
* If errorOK is true, just return "false" for bad input.
*
* if the input string ends with '!' it will always be treated as invalid
* (even if the check digit is valid)
*/
static bool
string2ean(const char *str, bool errorOK, ean13 *result,
2006-10-04 02:30:14 +02:00
enum isn_type accept)
{
2006-10-04 02:30:14 +02:00
bool digit,
last;
char buf[17] = " ";
char *aux1 = buf + 3; /* leave space for the first part, in case
* it's needed */
const char *aux2 = str;
enum isn_type type = INVALID;
2006-10-04 02:30:14 +02:00
unsigned check = 0,
rcheck = (unsigned) -1;
unsigned length = 0;
bool magic = false,
valid = true;
/* recognize and validate the number: */
2006-10-04 02:30:14 +02:00
while (*aux2 && length <= 13)
{
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
last = (*(aux2 + 1) == '!' || *(aux2 + 1) == '\0'); /* is the last character */
2006-10-04 02:30:14 +02:00
digit = (isdigit((unsigned char) *aux2) != 0); /* is current character
* a digit? */
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
if (*aux2 == '?' && last) /* automagically calculate check digit if
* it's '?' */
magic = digit = true;
2006-10-04 02:30:14 +02:00
if (length == 0 && (*aux2 == 'M' || *aux2 == 'm'))
{
/* only ISMN can be here */
2006-10-04 02:30:14 +02:00
if (type != INVALID)
goto eaninvalid;
type = ISMN;
*aux1++ = 'M';
length++;
2006-10-04 02:30:14 +02:00
}
else if (length == 7 && (digit || *aux2 == 'X' || *aux2 == 'x') && last)
{
/* only ISSN can be here */
2006-10-04 02:30:14 +02:00
if (type != INVALID)
goto eaninvalid;
type = ISSN;
*aux1++ = toupper((unsigned char) *aux2);
length++;
2006-10-04 02:30:14 +02:00
}
else if (length == 9 && (digit || *aux2 == 'X' || *aux2 == 'x') && last)
{
/* only ISBN and ISMN can be here */
2006-10-04 02:30:14 +02:00
if (type != INVALID && type != ISMN)
goto eaninvalid;
if (type == INVALID)
type = ISBN; /* ISMN must start with 'M' */
*aux1++ = toupper((unsigned char) *aux2);
length++;
2006-10-04 02:30:14 +02:00
}
else if (length == 11 && digit && last)
{
/* only UPC can be here */
2006-10-04 02:30:14 +02:00
if (type != INVALID)
goto eaninvalid;
type = UPC;
*aux1++ = *aux2;
length++;
2006-10-04 02:30:14 +02:00
}
else if (*aux2 == '-' || *aux2 == ' ')
{
/* skip, we could validate but I think it's worthless */
2006-10-04 02:30:14 +02:00
}
else if (*aux2 == '!' && *(aux2 + 1) == '\0')
{
/* the invalid check digit suffix was found, set it */
2006-10-04 02:30:14 +02:00
if (!magic)
valid = false;
magic = true;
2006-10-04 02:30:14 +02:00
}
else if (!digit)
{
goto eaninvalid;
2006-10-04 02:30:14 +02:00
}
else
{
*aux1++ = *aux2;
2006-10-04 02:30:14 +02:00
if (++length > 13)
goto eantoobig;
}
aux2++;
}
2006-10-04 02:30:14 +02:00
*aux1 = '\0'; /* terminate the string */
/* find the current check digit value */
2006-10-04 02:30:14 +02:00
if (length == 13)
{
/* only EAN13 can be here */
2006-10-04 02:30:14 +02:00
if (type != INVALID)
goto eaninvalid;
type = EAN13;
2006-10-04 02:30:14 +02:00
check = buf[15] - '0';
}
else if (length == 12)
{
/* only UPC can be here */
2006-10-04 02:30:14 +02:00
if (type != UPC)
goto eaninvalid;
check = buf[14] - '0';
}
else if (length == 10)
{
if (type != ISBN && type != ISMN)
goto eaninvalid;
if (buf[12] == 'X')
check = 10;
else
check = buf[12] - '0';
}
else if (length == 8)
{
if (type != INVALID && type != ISSN)
goto eaninvalid;
type = ISSN;
2006-10-04 02:30:14 +02:00
if (buf[10] == 'X')
check = 10;
else
check = buf[10] - '0';
}
else
goto eaninvalid;
if (type == INVALID)
goto eaninvalid;
/* obtain the real check digit value, validate, and convert to ean13: */
if (accept == EAN13 && type != accept)
goto eanwrongtype;
if (accept != ANY && type != EAN13 && type != accept)
goto eanwrongtype;
switch (type)
{
case EAN13:
2006-10-04 02:30:14 +02:00
valid = (valid && ((rcheck = checkdig(buf + 3, 13)) == check || magic));
/* now get the subtype of EAN13: */
2006-10-04 02:30:14 +02:00
if (buf[3] == '0')
type = UPC;
else if (strncmp("977", buf + 3, 3) == 0)
2006-10-04 02:30:14 +02:00
type = ISSN;
else if (strncmp("978", buf + 3, 3) == 0)
2006-10-04 02:30:14 +02:00
type = ISBN;
else if (strncmp("9790", buf + 3, 4) == 0)
2006-10-04 02:30:14 +02:00
type = ISMN;
else if (strncmp("979", buf + 3, 3) == 0)
2006-10-04 02:30:14 +02:00
type = ISBN;
if (accept != EAN13 && accept != ANY && type != accept)
goto eanwrongtype;
break;
case ISMN:
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
memcpy(buf, "9790", 4); /* this isn't for sure yet, for now ISMN
* it's only 9790 */
valid = (valid && ((rcheck = checkdig(buf, 13)) == check || magic));
break;
case ISBN:
memcpy(buf, "978", 3);
2006-10-04 02:30:14 +02:00
valid = (valid && ((rcheck = weight_checkdig(buf + 3, 10)) == check || magic));
break;
case ISSN:
memcpy(buf + 10, "00", 2); /* append 00 as the normal issue
2006-10-04 02:30:14 +02:00
* publication code */
memcpy(buf, "977", 3);
2006-10-04 02:30:14 +02:00
valid = (valid && ((rcheck = weight_checkdig(buf + 3, 8)) == check || magic));
break;
case UPC:
buf[2] = '0';
2006-10-04 02:30:14 +02:00
valid = (valid && ((rcheck = checkdig(buf + 2, 13)) == check || magic));
default:
break;
}
2006-10-04 02:30:14 +02:00
/* fix the check digit: */
for (aux1 = buf; *aux1 && *aux1 <= ' '; aux1++);
aux1[12] = checkdig(aux1, 13) + '0';
aux1[13] = '\0';
2006-10-04 02:30:14 +02:00
if (!valid && !magic)
goto eanbadcheck;
*result = str2ean(aux1);
2006-10-04 02:30:14 +02:00
*result |= valid ? 0 : 1;
return true;
2006-10-04 02:30:14 +02:00
eanbadcheck:
if (g_weak)
{ /* weak input mode is activated: */
/* set the "invalid-check-digit-on-input" flag */
*result = str2ean(aux1);
*result |= 1;
2006-10-04 02:30:14 +02:00
return true;
}
2006-10-04 02:30:14 +02:00
if (!errorOK)
{
if (rcheck == (unsigned) -1)
{
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
2006-10-04 02:30:14 +02:00
errmsg("invalid %s number: \"%s\"",
isn_names[accept], str)));
2006-10-04 02:30:14 +02:00
}
else
{
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid check digit for %s number: \"%s\", should be %c",
isn_names[accept], str, (rcheck == 10) ? ('X') : (rcheck + '0'))));
}
2006-10-04 02:30:14 +02:00
}
return false;
eaninvalid:
2006-10-04 02:30:14 +02:00
if (!errorOK)
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("invalid input syntax for %s number: \"%s\"",
isn_names[accept], str)));
return false;
eanwrongtype:
2006-10-04 02:30:14 +02:00
if (!errorOK)
ereport(ERROR,
(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
errmsg("cannot cast %s to %s for number: \"%s\"",
isn_names[type], isn_names[accept], str)));
return false;
eantoobig:
2006-10-04 02:30:14 +02:00
if (!errorOK)
ereport(ERROR,
2006-10-04 02:30:14 +02:00
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
errmsg("value \"%s\" is out of range for %s type",
str, isn_names[accept])));
return false;
}
/*----------------------------------------------------------
* Exported routines.
*---------------------------------------------------------*/
void _PG_init(void);
2006-10-04 02:30:14 +02:00
void
_PG_init(void)
{
if (ISN_DEBUG)
{
if (!check_table(EAN13_range, EAN13_index))
elog(ERROR, "EAN13 failed check");
if (!check_table(ISBN_range, ISBN_index))
elog(ERROR, "ISBN failed check");
if (!check_table(ISMN_range, ISMN_index))
elog(ERROR, "ISMN failed check");
if (!check_table(ISSN_range, ISSN_index))
elog(ERROR, "ISSN failed check");
if (!check_table(UPC_range, UPC_index))
elog(ERROR, "UPC failed check");
}
}
/* isn_out
*/
PG_FUNCTION_INFO_V1(isn_out);
Datum
isn_out(PG_FUNCTION_ARGS)
{
ean13 val = PG_GETARG_EAN13(0);
char *result;
char buf[MAXEAN13LEN + 1];
(void) ean2string(val, false, buf, true);
result = pstrdup(buf);
PG_RETURN_CSTRING(result);
}
/* ean13_out
*/
PG_FUNCTION_INFO_V1(ean13_out);
Datum
ean13_out(PG_FUNCTION_ARGS)
{
ean13 val = PG_GETARG_EAN13(0);
char *result;
char buf[MAXEAN13LEN + 1];
(void) ean2string(val, false, buf, false);
result = pstrdup(buf);
PG_RETURN_CSTRING(result);
}
/* ean13_in
*/
PG_FUNCTION_INFO_V1(ean13_in);
Datum
ean13_in(PG_FUNCTION_ARGS)
{
2006-10-04 02:30:14 +02:00
const char *str = PG_GETARG_CSTRING(0);
ean13 result;
(void) string2ean(str, false, &result, EAN13);
PG_RETURN_EAN13(result);
}
/* isbn_in
*/
PG_FUNCTION_INFO_V1(isbn_in);
Datum
isbn_in(PG_FUNCTION_ARGS)
{
2006-10-04 02:30:14 +02:00
const char *str = PG_GETARG_CSTRING(0);
ean13 result;
(void) string2ean(str, false, &result, ISBN);
PG_RETURN_EAN13(result);
}
/* ismn_in
*/
PG_FUNCTION_INFO_V1(ismn_in);
Datum
ismn_in(PG_FUNCTION_ARGS)
{
2006-10-04 02:30:14 +02:00
const char *str = PG_GETARG_CSTRING(0);
ean13 result;
(void) string2ean(str, false, &result, ISMN);
PG_RETURN_EAN13(result);
}
/* issn_in
*/
PG_FUNCTION_INFO_V1(issn_in);
Datum
issn_in(PG_FUNCTION_ARGS)
{
2006-10-04 02:30:14 +02:00
const char *str = PG_GETARG_CSTRING(0);
ean13 result;
(void) string2ean(str, false, &result, ISSN);
PG_RETURN_EAN13(result);
}
/* upc_in
*/
PG_FUNCTION_INFO_V1(upc_in);
Datum
upc_in(PG_FUNCTION_ARGS)
{
2006-10-04 02:30:14 +02:00
const char *str = PG_GETARG_CSTRING(0);
ean13 result;
(void) string2ean(str, false, &result, UPC);
PG_RETURN_EAN13(result);
}
/* casting functions
*/
PG_FUNCTION_INFO_V1(isbn_cast_from_ean13);
Datum
isbn_cast_from_ean13(PG_FUNCTION_ARGS)
{
ean13 val = PG_GETARG_EAN13(0);
ean13 result;
(void) ean2isn(val, false, &result, ISBN);
PG_RETURN_EAN13(result);
}
PG_FUNCTION_INFO_V1(ismn_cast_from_ean13);
Datum
ismn_cast_from_ean13(PG_FUNCTION_ARGS)
{
ean13 val = PG_GETARG_EAN13(0);
ean13 result;
(void) ean2isn(val, false, &result, ISMN);
PG_RETURN_EAN13(result);
}
PG_FUNCTION_INFO_V1(issn_cast_from_ean13);
Datum
issn_cast_from_ean13(PG_FUNCTION_ARGS)
{
ean13 val = PG_GETARG_EAN13(0);
ean13 result;
(void) ean2isn(val, false, &result, ISSN);
PG_RETURN_EAN13(result);
}
PG_FUNCTION_INFO_V1(upc_cast_from_ean13);
Datum
upc_cast_from_ean13(PG_FUNCTION_ARGS)
{
ean13 val = PG_GETARG_EAN13(0);
ean13 result;
(void) ean2isn(val, false, &result, UPC);
PG_RETURN_EAN13(result);
}
/* is_valid - returns false if the "invalid-check-digit-on-input" is set
*/
PG_FUNCTION_INFO_V1(is_valid);
Datum
is_valid(PG_FUNCTION_ARGS)
{
2006-10-04 02:30:14 +02:00
ean13 val = PG_GETARG_EAN13(0);
PG_RETURN_BOOL((val & 1) == 0);
}
/* make_valid - unsets the "invalid-check-digit-on-input" flag
*/
PG_FUNCTION_INFO_V1(make_valid);
Datum
make_valid(PG_FUNCTION_ARGS)
{
2006-10-04 02:30:14 +02:00
ean13 val = PG_GETARG_EAN13(0);
val &= ~((ean13) 1);
PG_RETURN_EAN13(val);
}
2006-10-04 02:30:14 +02:00
/* this function temporarily sets weak input flag
* (to lose the strictness of check digit acceptance)
* It's a helper function, not intended to be used!!
*/
PG_FUNCTION_INFO_V1(accept_weak_input);
Datum
accept_weak_input(PG_FUNCTION_ARGS)
{
#ifdef ISN_WEAK_MODE
g_weak = PG_GETARG_BOOL(0);
#else
/* function has no effect */
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
#endif /* ISN_WEAK_MODE */
PG_RETURN_BOOL(g_weak);
}
PG_FUNCTION_INFO_V1(weak_input_status);
Datum
weak_input_status(PG_FUNCTION_ARGS)
{
PG_RETURN_BOOL(g_weak);
}