database/postgres
database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-05-08 07:21:33 +03:00
Code

Move libc-specific code from pg_locale.c into pg_locale_libc.c.

Browse Source 
Move implementation of pg_locale_t code for libc collations into
pg_locale_libc.c. Other locale-related code, such as
pg_perm_setlocale(), remains in pg_locale.c for now.

Discussion: https://postgr.es/m/flat/2830211e1b6e6a2e26d845780b03e125281ea17b.camel@j-davis.com
This commit is contained in:
Jeff Davis 2024-10-14 12:22:13 -07:00
parent 9812138593
commit 66ac94cdc7
4 changed files with 514 additions and 464 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
src/backend/utils/adt/Makefile
View File

@ -80,6 +80,7 @@ OBJS = \
partitionfuncs.o \
pg_locale.o \
pg_locale_icu.o \
pg_locale_libc.o \
pg_lsn.o \
pg_upgrade_support.o \
pgstatfuncs.o \

1
src/backend/utils/adt/meson.build
View File

@ -67,6 +67,7 @@ backend_sources += files(
'partitionfuncs.c',
'pg_locale.c',
'pg_locale_icu.c',
'pg_locale_libc.c',
'pg_lsn.c',
'pg_upgrade_support.c',
'pgstatfuncs.c',

474
src/backend/utils/adt/pg_locale.c
View File

@ -105,6 +105,16 @@ extern size_t strnxfrm_prefix_icu(char *dest, size_t destsize,
pg_locale_t locale);
#endif
/* pg_locale_libc.c */
extern locale_t make_libc_collator(const char *collate,
const char *ctype);
extern int strncoll_libc(const char *arg1, ssize_t len1,
const char *arg2, ssize_t len2,
pg_locale_t locale);
extern size_t strnxfrm_libc(char *dest, size_t destsize,
const char *src, ssize_t srclen,
pg_locale_t locale);
/* GUC settings */
char *locale_messages;
char *locale_monetary;
@ -173,43 +183,6 @@ static pg_locale_t last_collation_cache_locale = NULL;
static char *IsoLocaleName(const char *);
#endif
/*
* POSIX doesn't define _l-variants of these functions, but several systems
* have them. We provide our own replacements here.
*/
#ifndef HAVE_MBSTOWCS_L
static size_t
mbstowcs_l(wchar_t *dest, const char *src, size_t n, locale_t loc)
{
#ifdef WIN32
return _mbstowcs_l(dest, src, n, loc);
#else
size_t result;
locale_t save_locale = uselocale(loc);
result = mbstowcs(dest, src, n);
uselocale(save_locale);
return result;
#endif
}
#endif
#ifndef HAVE_WCSTOMBS_L
static size_t
wcstombs_l(char *dest, const wchar_t *src, size_t n, locale_t loc)
{
#ifdef WIN32
return _wcstombs_l(dest, src, n, loc);
#else
size_t result;
locale_t save_locale = uselocale(loc);
result = wcstombs(dest, src, n);
uselocale(save_locale);
return result;
#endif
}
#endif
/*
* pg_perm_setlocale
*
@ -1280,108 +1253,6 @@ lookup_collation_cache(Oid collation)
return cache_entry;
}
/* simple subroutine for reporting errors from newlocale() */
static void
report_newlocale_failure(const char *localename)
{
int save_errno;
/*
* Windows doesn't provide any useful error indication from
* _create_locale(), and BSD-derived platforms don't seem to feel they
* need to set errno either (even though POSIX is pretty clear that
* newlocale should do so). So, if errno hasn't been set, assume ENOENT
* is what to report.
*/
if (errno == 0)
errno = ENOENT;
/*
* ENOENT means "no such locale", not "no such file", so clarify that
* errno with an errdetail message.
*/
save_errno = errno; /* auxiliary funcs might change errno */
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("could not create locale \"%s\": %m",
localename),
(save_errno == ENOENT ?
errdetail("The operating system could not find any locale data for the locale name \"%s\".",
localename) : 0)));
}
/*
* Create a locale_t with the given collation and ctype.
*
* The "C" and "POSIX" locales are not actually handled by libc, so return
* NULL.
*
* Ensure that no path leaks a locale_t.
*/
static locale_t
make_libc_collator(const char *collate, const char *ctype)
{
locale_t loc = 0;
if (strcmp(collate, ctype) == 0)
{
if (strcmp(ctype, "C") != 0 && strcmp(ctype, "POSIX") != 0)
{
/* Normal case where they're the same */
errno = 0;
#ifndef WIN32
loc = newlocale(LC_COLLATE_MASK | LC_CTYPE_MASK, collate,
NULL);
#else
loc = _create_locale(LC_ALL, collate);
#endif
if (!loc)
report_newlocale_failure(collate);
}
}
else
{
#ifndef WIN32
/* We need two newlocale() steps */
locale_t loc1 = 0;
if (strcmp(collate, "C") != 0 && strcmp(collate, "POSIX") != 0)
{
errno = 0;
loc1 = newlocale(LC_COLLATE_MASK, collate, NULL);
if (!loc1)
report_newlocale_failure(collate);
}
if (strcmp(ctype, "C") != 0 && strcmp(ctype, "POSIX") != 0)
{
errno = 0;
loc = newlocale(LC_CTYPE_MASK, ctype, loc1);
if (!loc)
{
if (loc1)
freelocale(loc1);
report_newlocale_failure(ctype);
}
}
else
loc = loc1;
#else
/*
* XXX The _create_locale() API doesn't appear to support this. Could
* perhaps be worked around by changing pg_locale_t to contain two
* separate fields.
*/
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("collations with different collate and ctype values are not supported on this platform")));
#endif
}
return loc;
}
/*
* Initialize default_locale with database locale settings.
*/
@ -1746,150 +1617,6 @@ get_collation_actual_version(char collprovider, const char *collcollate)
return collversion;
}
/*
* strncoll_libc_win32_utf8
*
* Win32 does not have UTF-8. Convert UTF8 arguments to wide characters and
* invoke wcscoll_l().
*
* An input string length of -1 means that it's NUL-terminated.
*/
#ifdef WIN32
static int
strncoll_libc_win32_utf8(const char *arg1, ssize_t len1, const char *arg2,
ssize_t len2, pg_locale_t locale)
{
char sbuf[TEXTBUFLEN];
char *buf = sbuf;
char *a1p,
*a2p;
int a1len;
int a2len;
int r;
int result;
Assert(locale->provider == COLLPROVIDER_LIBC);
Assert(GetDatabaseEncoding() == PG_UTF8);
if (len1 == -1)
len1 = strlen(arg1);
if (len2 == -1)
len2 = strlen(arg2);
a1len = len1 * 2 + 2;
a2len = len2 * 2 + 2;
if (a1len + a2len > TEXTBUFLEN)
buf = palloc(a1len + a2len);
a1p = buf;
a2p = buf + a1len;
/* API does not work for zero-length input */
if (len1 == 0)
r = 0;
else
{
r = MultiByteToWideChar(CP_UTF8, 0, arg1, len1,
(LPWSTR) a1p, a1len / 2);
if (!r)
ereport(ERROR,
(errmsg("could not convert string to UTF-16: error code %lu",
GetLastError())));
}
((LPWSTR) a1p)[r] = 0;
if (len2 == 0)
r = 0;
else
{
r = MultiByteToWideChar(CP_UTF8, 0, arg2, len2,
(LPWSTR) a2p, a2len / 2);
if (!r)
ereport(ERROR,
(errmsg("could not convert string to UTF-16: error code %lu",
GetLastError())));
}
((LPWSTR) a2p)[r] = 0;
errno = 0;
result = wcscoll_l((LPWSTR) a1p, (LPWSTR) a2p, locale->info.lt);
if (result == 2147483647) /* _NLSCMPERROR; missing from mingw headers */
ereport(ERROR,
(errmsg("could not compare Unicode strings: %m")));
if (buf != sbuf)
pfree(buf);
return result;
}
#endif /* WIN32 */
/*
* strncoll_libc
*
* NUL-terminate arguments, if necessary, and pass to strcoll_l().
*
* An input string length of -1 means that it's already NUL-terminated.
*/
static int
strncoll_libc(const char *arg1, ssize_t len1, const char *arg2, ssize_t len2,
pg_locale_t locale)
{
char sbuf[TEXTBUFLEN];
char *buf = sbuf;
size_t bufsize1 = (len1 == -1) ? 0 : len1 + 1;
size_t bufsize2 = (len2 == -1) ? 0 : len2 + 1;
const char *arg1n;
const char *arg2n;
int result;
Assert(locale->provider == COLLPROVIDER_LIBC);
#ifdef WIN32
/* check for this case before doing the work for nul-termination */
if (GetDatabaseEncoding() == PG_UTF8)
return strncoll_libc_win32_utf8(arg1, len1, arg2, len2, locale);
#endif /* WIN32 */
if (bufsize1 + bufsize2 > TEXTBUFLEN)
buf = palloc(bufsize1 + bufsize2);
/* nul-terminate arguments if necessary */
if (len1 == -1)
{
arg1n = arg1;
}
else
{
char *buf1 = buf;
memcpy(buf1, arg1, len1);
buf1[len1] = '\0';
arg1n = buf1;
}
if (len2 == -1)
{
arg2n = arg2;
}
else
{
char *buf2 = buf + bufsize1;
memcpy(buf2, arg2, len2);
buf2[len2] = '\0';
arg2n = buf2;
}
result = strcoll_l(arg1n, arg2n, locale->info.lt);
if (buf != sbuf)
pfree(buf);
return result;
}
/*
* pg_strcoll
*
@ -1946,45 +1673,6 @@ pg_strncoll(const char *arg1, ssize_t len1, const char *arg2, ssize_t len2,
return result;
}
/*
* strnxfrm_libc
*
* NUL-terminate src, if necessary, and pass to strxfrm_l().
*
* A source length of -1 means that it's already NUL-terminated.
*/
static size_t
strnxfrm_libc(char *dest, size_t destsize, const char *src, ssize_t srclen,
pg_locale_t locale)
{
char sbuf[TEXTBUFLEN];
char *buf = sbuf;
size_t bufsize = srclen + 1;
size_t result;
Assert(locale->provider == COLLPROVIDER_LIBC);
if (srclen == -1)
return strxfrm_l(dest, src, destsize, locale->info.lt);
if (bufsize > TEXTBUFLEN)
buf = palloc(bufsize);
/* nul-terminate argument */
memcpy(buf, src, srclen);
buf[srclen] = '\0';
result = strxfrm_l(dest, buf, destsize, locale->info.lt);
if (buf != sbuf)
pfree(buf);
/* if dest is defined, it should be nul-terminated */
Assert(result >= destsize || dest[result] == '\0');
return result;
}
/*
* Return true if the collation provider supports pg_strxfrm() and
* pg_strnxfrm(); otherwise false.
@ -2333,145 +2021,3 @@ icu_validate_locale(const char *loc_str)
errmsg("ICU is not supported in this build")));
#endif /* not USE_ICU */
}
/*
* These functions convert from/to libc's wchar_t, *not* pg_wchar_t.
* Therefore we keep them here rather than with the mbutils code.
*/
/*
* wchar2char --- convert wide characters to multibyte format
*
* This has the same API as the standard wcstombs_l() function; in particular,
* tolen is the maximum number of bytes to store at *to, and *from must be
* zero-terminated. The output will be zero-terminated iff there is room.
*/
size_t
wchar2char(char *to, const wchar_t *from, size_t tolen, pg_locale_t locale)
{
size_t result;
Assert(!locale || locale->provider == COLLPROVIDER_LIBC);
if (tolen == 0)
return 0;
#ifdef WIN32
/*
* On Windows, the "Unicode" locales assume UTF16 not UTF8 encoding, and
* for some reason mbstowcs and wcstombs won't do this for us, so we use
* MultiByteToWideChar().
*/
if (GetDatabaseEncoding() == PG_UTF8)
{
result = WideCharToMultiByte(CP_UTF8, 0, from, -1, to, tolen,
NULL, NULL);
/* A zero return is failure */
if (result <= 0)
result = -1;
else
{
Assert(result <= tolen);
/* Microsoft counts the zero terminator in the result */
result--;
}
}
else
#endif /* WIN32 */
if (locale == (pg_locale_t) 0)
{
/* Use wcstombs directly for the default locale */
result = wcstombs(to, from, tolen);
}
else
{
/* Use wcstombs_l for nondefault locales */
result = wcstombs_l(to, from, tolen, locale->info.lt);
}
return result;
}
/*
* char2wchar --- convert multibyte characters to wide characters
*
* This has almost the API of mbstowcs_l(), except that *from need not be
* null-terminated; instead, the number of input bytes is specified as
* fromlen. Also, we ereport() rather than returning -1 for invalid
* input encoding. tolen is the maximum number of wchar_t's to store at *to.
* The output will be zero-terminated iff there is room.
*/
size_t
char2wchar(wchar_t *to, size_t tolen, const char *from, size_t fromlen,
pg_locale_t locale)
{
size_t result;
Assert(!locale || locale->provider == COLLPROVIDER_LIBC);
if (tolen == 0)
return 0;
#ifdef WIN32
/* See WIN32 "Unicode" comment above */
if (GetDatabaseEncoding() == PG_UTF8)
{
/* Win32 API does not work for zero-length input */
if (fromlen == 0)
result = 0;
else
{
result = MultiByteToWideChar(CP_UTF8, 0, from, fromlen, to, tolen - 1);
/* A zero return is failure */
if (result == 0)
result = -1;
}
if (result != -1)
{
Assert(result < tolen);
/* Append trailing null wchar (MultiByteToWideChar() does not) */
to[result] = 0;
}
}
else
#endif /* WIN32 */
{
/* mbstowcs requires ending '\0' */
char *str = pnstrdup(from, fromlen);
if (locale == (pg_locale_t) 0)
{
/* Use mbstowcs directly for the default locale */
result = mbstowcs(to, str, tolen);
}
else
{
/* Use mbstowcs_l for nondefault locales */
result = mbstowcs_l(to, str, tolen, locale->info.lt);
}
pfree(str);
}
if (result == -1)
{
/*
* Invalid multibyte character encountered. We try to give a useful
* error message by letting pg_verifymbstr check the string. But it's
* possible that the string is OK to us, and not OK to mbstowcs ---
* this suggests that the LC_CTYPE locale is different from the
* database encoding. Give a generic error message if pg_verifymbstr
* can't find anything wrong.
*/
pg_verifymbstr(from, fromlen, false); /* might not return */
/* but if it does ... */
ereport(ERROR,
(errcode(ERRCODE_CHARACTER_NOT_IN_REPERTOIRE),
errmsg("invalid multibyte character for locale"),
errhint("The server's LC_CTYPE locale is probably incompatible with the database encoding.")));
}
return result;
}

502
src/backend/utils/adt/pg_locale_libc.c Normal file
View File

@ -0,0 +1,502 @@
/*-----------------------------------------------------------------------
*
* PostgreSQL locale utilities for libc
*
* Portions Copyright (c) 2002-2024, PostgreSQL Global Development Group
*
* src/backend/utils/adt/pg_locale_libc.c
*
*-----------------------------------------------------------------------
*/
#include "postgres.h"
#include "catalog/pg_collation.h"
#include "mb/pg_wchar.h"
#include "utils/formatting.h"
#include "utils/pg_locale.h"
/*
* Size of stack buffer to use for string transformations, used to avoid heap
* allocations in typical cases. This should be large enough that most strings
* will fit, but small enough that we feel comfortable putting it on the
* stack.
*/
#define TEXTBUFLEN 1024
extern locale_t make_libc_collator(const char *collate,
const char *ctype);
extern int strncoll_libc(const char *arg1, ssize_t len1,
const char *arg2, ssize_t len2,
pg_locale_t locale);
extern size_t strnxfrm_libc(char *dest, size_t destsize,
const char *src, ssize_t srclen,
pg_locale_t locale);
static void report_newlocale_failure(const char *localename);
#ifdef WIN32
static int strncoll_libc_win32_utf8(const char *arg1, ssize_t len1,
const char *arg2, ssize_t len2,
pg_locale_t locale);
#endif
/*
* Create a locale_t with the given collation and ctype.
*
* The "C" and "POSIX" locales are not actually handled by libc, so return
* NULL.
*
* Ensure that no path leaks a locale_t.
*/
locale_t
make_libc_collator(const char *collate, const char *ctype)
{
locale_t loc = 0;
if (strcmp(collate, ctype) == 0)
{
if (strcmp(ctype, "C") != 0 && strcmp(ctype, "POSIX") != 0)
{
/* Normal case where they're the same */
errno = 0;
#ifndef WIN32
loc = newlocale(LC_COLLATE_MASK | LC_CTYPE_MASK, collate,
NULL);
#else
loc = _create_locale(LC_ALL, collate);
#endif
if (!loc)
report_newlocale_failure(collate);
}
}
else
{
#ifndef WIN32
/* We need two newlocale() steps */
locale_t loc1 = 0;
if (strcmp(collate, "C") != 0 && strcmp(collate, "POSIX") != 0)
{
errno = 0;
loc1 = newlocale(LC_COLLATE_MASK, collate, NULL);
if (!loc1)
report_newlocale_failure(collate);
}
if (strcmp(ctype, "C") != 0 && strcmp(ctype, "POSIX") != 0)
{
errno = 0;
loc = newlocale(LC_CTYPE_MASK, ctype, loc1);
if (!loc)
{
if (loc1)
freelocale(loc1);
report_newlocale_failure(ctype);
}
}
else
loc = loc1;
#else
/*
* XXX The _create_locale() API doesn't appear to support this. Could
* perhaps be worked around by changing pg_locale_t to contain two
* separate fields.
*/
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("collations with different collate and ctype values are not supported on this platform")));
#endif
}
return loc;
}
/*
* strncoll_libc
*
* NUL-terminate arguments, if necessary, and pass to strcoll_l().
*
* An input string length of -1 means that it's already NUL-terminated.
*/
int
strncoll_libc(const char *arg1, ssize_t len1, const char *arg2, ssize_t len2,
pg_locale_t locale)
{
char sbuf[TEXTBUFLEN];
char *buf = sbuf;
size_t bufsize1 = (len1 == -1) ? 0 : len1 + 1;
size_t bufsize2 = (len2 == -1) ? 0 : len2 + 1;
const char *arg1n;
const char *arg2n;
int result;
Assert(locale->provider == COLLPROVIDER_LIBC);
#ifdef WIN32
/* check for this case before doing the work for nul-termination */
if (GetDatabaseEncoding() == PG_UTF8)
return strncoll_libc_win32_utf8(arg1, len1, arg2, len2, locale);
#endif /* WIN32 */
if (bufsize1 + bufsize2 > TEXTBUFLEN)
buf = palloc(bufsize1 + bufsize2);
/* nul-terminate arguments if necessary */
if (len1 == -1)
{
arg1n = arg1;
}
else
{
char *buf1 = buf;
memcpy(buf1, arg1, len1);
buf1[len1] = '\0';
arg1n = buf1;
}
if (len2 == -1)
{
arg2n = arg2;
}
else
{
char *buf2 = buf + bufsize1;
memcpy(buf2, arg2, len2);
buf2[len2] = '\0';
arg2n = buf2;
}
result = strcoll_l(arg1n, arg2n, locale->info.lt);
if (buf != sbuf)
pfree(buf);
return result;
}
/*
* strnxfrm_libc
*
* NUL-terminate src, if necessary, and pass to strxfrm_l().
*
* A source length of -1 means that it's already NUL-terminated.
*/
size_t
strnxfrm_libc(char *dest, size_t destsize, const char *src, ssize_t srclen,
pg_locale_t locale)
{
char sbuf[TEXTBUFLEN];
char *buf = sbuf;
size_t bufsize = srclen + 1;
size_t result;
Assert(locale->provider == COLLPROVIDER_LIBC);
if (srclen == -1)
return strxfrm_l(dest, src, destsize, locale->info.lt);
if (bufsize > TEXTBUFLEN)
buf = palloc(bufsize);
/* nul-terminate argument */
memcpy(buf, src, srclen);
buf[srclen] = '\0';
result = strxfrm_l(dest, buf, destsize, locale->info.lt);
if (buf != sbuf)
pfree(buf);
/* if dest is defined, it should be nul-terminated */
Assert(result >= destsize || dest[result] == '\0');
return result;
}
/*
* strncoll_libc_win32_utf8
*
* Win32 does not have UTF-8. Convert UTF8 arguments to wide characters and
* invoke wcscoll_l().
*
* An input string length of -1 means that it's NUL-terminated.
*/
#ifdef WIN32
static int
strncoll_libc_win32_utf8(const char *arg1, ssize_t len1, const char *arg2,
ssize_t len2, pg_locale_t locale)
{
char sbuf[TEXTBUFLEN];
char *buf = sbuf;
char *a1p,
*a2p;
int a1len;
int a2len;
int r;
int result;
Assert(locale->provider == COLLPROVIDER_LIBC);
Assert(GetDatabaseEncoding() == PG_UTF8);
if (len1 == -1)
len1 = strlen(arg1);
if (len2 == -1)
len2 = strlen(arg2);
a1len = len1 * 2 + 2;
a2len = len2 * 2 + 2;
if (a1len + a2len > TEXTBUFLEN)
buf = palloc(a1len + a2len);
a1p = buf;
a2p = buf + a1len;
/* API does not work for zero-length input */
if (len1 == 0)
r = 0;
else
{
r = MultiByteToWideChar(CP_UTF8, 0, arg1, len1,
(LPWSTR) a1p, a1len / 2);
if (!r)
ereport(ERROR,
(errmsg("could not convert string to UTF-16: error code %lu",
GetLastError())));
}
((LPWSTR) a1p)[r] = 0;
if (len2 == 0)
r = 0;
else
{
r = MultiByteToWideChar(CP_UTF8, 0, arg2, len2,
(LPWSTR) a2p, a2len / 2);
if (!r)
ereport(ERROR,
(errmsg("could not convert string to UTF-16: error code %lu",
GetLastError())));
}
((LPWSTR) a2p)[r] = 0;
errno = 0;
result = wcscoll_l((LPWSTR) a1p, (LPWSTR) a2p, locale->info.lt);
if (result == 2147483647) /* _NLSCMPERROR; missing from mingw headers */
ereport(ERROR,
(errmsg("could not compare Unicode strings: %m")));
if (buf != sbuf)
pfree(buf);
return result;
}
#endif /* WIN32 */
/* simple subroutine for reporting errors from newlocale() */
static void
report_newlocale_failure(const char *localename)
{
int save_errno;
/*
* Windows doesn't provide any useful error indication from
* _create_locale(), and BSD-derived platforms don't seem to feel they
* need to set errno either (even though POSIX is pretty clear that
* newlocale should do so). So, if errno hasn't been set, assume ENOENT
* is what to report.
*/
if (errno == 0)
errno = ENOENT;
/*
* ENOENT means "no such locale", not "no such file", so clarify that
* errno with an errdetail message.
*/
save_errno = errno; /* auxiliary funcs might change errno */
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("could not create locale \"%s\": %m",
localename),
(save_errno == ENOENT ?
errdetail("The operating system could not find any locale data for the locale name \"%s\".",
localename) : 0)));
}
/*
* POSIX doesn't define _l-variants of these functions, but several systems
* have them. We provide our own replacements here.
*/
#ifndef HAVE_MBSTOWCS_L
static size_t
mbstowcs_l(wchar_t *dest, const char *src, size_t n, locale_t loc)
{
#ifdef WIN32
return _mbstowcs_l(dest, src, n, loc);
#else
size_t result;
locale_t save_locale = uselocale(loc);
result = mbstowcs(dest, src, n);
uselocale(save_locale);
return result;
#endif
}
#endif
#ifndef HAVE_WCSTOMBS_L
static size_t
wcstombs_l(char *dest, const wchar_t *src, size_t n, locale_t loc)
{
#ifdef WIN32
return _wcstombs_l(dest, src, n, loc);
#else
size_t result;
locale_t save_locale = uselocale(loc);
result = wcstombs(dest, src, n);
uselocale(save_locale);
return result;
#endif
}
#endif
/*
* These functions convert from/to libc's wchar_t, *not* pg_wchar_t.
* Therefore we keep them here rather than with the mbutils code.
*/
/*
* wchar2char --- convert wide characters to multibyte format
*
* This has the same API as the standard wcstombs_l() function; in particular,
* tolen is the maximum number of bytes to store at *to, and *from must be
* zero-terminated. The output will be zero-terminated iff there is room.
*/
size_t
wchar2char(char *to, const wchar_t *from, size_t tolen, pg_locale_t locale)
{
size_t result;
if (tolen == 0)
return 0;
#ifdef WIN32
/*
* On Windows, the "Unicode" locales assume UTF16 not UTF8 encoding, and
* for some reason mbstowcs and wcstombs won't do this for us, so we use
* MultiByteToWideChar().
*/
if (GetDatabaseEncoding() == PG_UTF8)
{
result = WideCharToMultiByte(CP_UTF8, 0, from, -1, to, tolen,
NULL, NULL);
/* A zero return is failure */
if (result <= 0)
result = -1;
else
{
Assert(result <= tolen);
/* Microsoft counts the zero terminator in the result */
result--;
}
}
else
#endif /* WIN32 */
if (locale == (pg_locale_t) 0)
{
/* Use wcstombs directly for the default locale */
result = wcstombs(to, from, tolen);
}
else
{
/* Use wcstombs_l for nondefault locales */
result = wcstombs_l(to, from, tolen, locale->info.lt);
}
return result;
}
/*
* char2wchar --- convert multibyte characters to wide characters
*
* This has almost the API of mbstowcs_l(), except that *from need not be
* null-terminated; instead, the number of input bytes is specified as
* fromlen. Also, we ereport() rather than returning -1 for invalid
* input encoding. tolen is the maximum number of wchar_t's to store at *to.
* The output will be zero-terminated iff there is room.
*/
size_t
char2wchar(wchar_t *to, size_t tolen, const char *from, size_t fromlen,
pg_locale_t locale)
{
size_t result;
if (tolen == 0)
return 0;
#ifdef WIN32
/* See WIN32 "Unicode" comment above */
if (GetDatabaseEncoding() == PG_UTF8)
{
/* Win32 API does not work for zero-length input */
if (fromlen == 0)
result = 0;
else
{
result = MultiByteToWideChar(CP_UTF8, 0, from, fromlen, to, tolen - 1);
/* A zero return is failure */
if (result == 0)
result = -1;
}
if (result != -1)
{
Assert(result < tolen);
/* Append trailing null wchar (MultiByteToWideChar() does not) */
to[result] = 0;
}
}
else
#endif /* WIN32 */
{
/* mbstowcs requires ending '\0' */
char *str = pnstrdup(from, fromlen);
if (locale == (pg_locale_t) 0)
{
/* Use mbstowcs directly for the default locale */
result = mbstowcs(to, str, tolen);
}
else
{
/* Use mbstowcs_l for nondefault locales */
result = mbstowcs_l(to, str, tolen, locale->info.lt);
}
pfree(str);
}
if (result == -1)
{
/*
* Invalid multibyte character encountered. We try to give a useful
* error message by letting pg_verifymbstr check the string. But it's
* possible that the string is OK to us, and not OK to mbstowcs ---
* this suggests that the LC_CTYPE locale is different from the
* database encoding. Give a generic error message if pg_verifymbstr
* can't find anything wrong.
*/
pg_verifymbstr(from, fromlen, false); /* might not return */
/* but if it does ... */
ereport(ERROR,
(errcode(ERRCODE_CHARACTER_NOT_IN_REPERTOIRE),
errmsg("invalid multibyte character for locale"),
errhint("The server's LC_CTYPE locale is probably incompatible with the database encoding.")));
}
return result;
}