* Functions for handling locale-related info
*
*
- * Copyright (c) 1996-2007, PostgreSQL Global Development Group
+ * Copyright (c) 1996-2013, PostgreSQL Global Development Group
*
*
* IDENTIFICATION
- * $PostgreSQL: pgsql/src/port/chklocale.c,v 1.5 2007/10/10 11:50:07 mha Exp $
+ * src/port/chklocale.c
*
*-------------------------------------------------------------------------
*/
{PG_EUC_KR, "IBM-eucKR"},
{PG_EUC_KR, "deckorean"},
{PG_EUC_KR, "5601"},
- {PG_EUC_KR, "CP51949"}, /* or 20949 ? */
+ {PG_EUC_KR, "CP51949"},
{PG_EUC_TW, "EUC-TW"},
{PG_EUC_TW, "eucTW"},
{PG_KOI8R, "KOI8-R"},
{PG_KOI8R, "CP20866"},
+ {PG_KOI8U, "KOI8-U"},
+ {PG_KOI8U, "CP21866"},
+
+ {PG_WIN866, "CP866"},
+ {PG_WIN874, "CP874"},
+ {PG_WIN1250, "CP1250"},
+ {PG_WIN1251, "CP1251"},
+ {PG_WIN1251, "ansi-1251"},
{PG_WIN1252, "CP1252"},
{PG_WIN1253, "CP1253"},
{PG_WIN1254, "CP1254"},
{PG_WIN1256, "CP1256"},
{PG_WIN1257, "CP1257"},
{PG_WIN1258, "CP1258"},
-#ifdef NOT_VERIFIED
- {PG_WIN874, "???"},
-#endif
- {PG_WIN1251, "CP1251"},
- {PG_WIN866, "CP866"},
{PG_ISO_8859_5, "ISO-8859-5"},
{PG_ISO_8859_5, "ISO8859-5"},
{PG_ISO_8859_8, "ISO-8859-8"},
{PG_ISO_8859_8, "ISO8859-8"},
{PG_ISO_8859_8, "iso88598"},
- {PG_ISO_8859_8, "CP28598"},
+ {PG_ISO_8859_8, "CP28598"},
{PG_SJIS, "SJIS"},
{PG_SJIS, "PCK"},
{PG_SJIS, "CP932"},
+ {PG_SJIS, "SHIFT_JIS"},
{PG_BIG5, "BIG5"},
{PG_BIG5, "BIG5HKSCS"},
+ {PG_BIG5, "Big5-HKSCS"},
{PG_BIG5, "CP950"},
{PG_GBK, "GBK"},
{PG_GBK, "CP936"},
{PG_UHC, "UHC"},
+ {PG_UHC, "CP949"},
{PG_JOHAB, "JOHAB"},
{PG_JOHAB, "CP1361"},
{PG_SHIFT_JIS_2004, "SJIS_2004"},
+ {PG_SQL_ASCII, "US-ASCII"},
+
{PG_SQL_ASCII, NULL} /* end marker */
};
#ifdef WIN32
/*
- * On Windows, use CP<codepage number> instead of the nl_langinfo() result
+ * On Windows, use CP<code page number> instead of the nl_langinfo() result
+ *
+ * Visual Studio 2012 expanded the set of valid LC_CTYPE values, so have its
+ * locale machinery determine the code page. See comments at IsoLocaleName().
+ * For other compilers, follow the locale's predictable format.
+ *
+ * Returns a malloc()'d string for the caller to free.
*/
static char *
win32_langinfo(const char *ctype)
{
- char *r;
+ char *r = NULL;
+
+#if (_MSC_VER >= 1700)
+ _locale_t loct = NULL;
+
+ loct = _create_locale(LC_CTYPE, ctype);
+ if (loct != NULL)
+ {
+ r = malloc(16); /* excess */
+ if (r != NULL)
+ sprintf(r, "CP%u", loct->locinfo->lc_codepage);
+ _free_locale(loct);
+ }
+#else
char *codepage;
- int ln;
/*
- * Locale format on Win32 is <Language>_<Country>.<CodePage> .
- * For example, English_USA.1252.
+ * Locale format on Win32 is <Language>_<Country>.<CodePage> . For
+ * example, English_United States.1252.
*/
codepage = strrchr(ctype, '.');
- if (!codepage)
- return NULL;
- codepage++;
- ln = strlen(codepage);
- r = malloc(ln + 3);
- sprintf(r, "CP%s", codepage);
+ if (codepage != NULL)
+ {
+ int ln;
+
+ codepage++;
+ ln = strlen(codepage);
+ r = malloc(ln + 3);
+ if (r != NULL)
+ sprintf(r, "CP%s", codepage);
+ }
+#endif
return r;
}
-#endif /* WIN32 */
+
+#ifndef FRONTEND
+/*
+ * Given a Windows code page identifier, find the corresponding PostgreSQL
+ * encoding. Issue a warning and return -1 if none found.
+ */
+int
+pg_codepage_to_encoding(UINT cp)
+{
+ char sys[16];
+ int i;
+
+ sprintf(sys, "CP%u", cp);
+
+ /* Check the table */
+ for (i = 0; encoding_match_list[i].system_enc_name; i++)
+ if (pg_strcasecmp(sys, encoding_match_list[i].system_enc_name) == 0)
+ return encoding_match_list[i].pg_enc_code;
+
+ ereport(WARNING,
+ (errmsg("could not determine encoding for codeset \"%s\"", sys),
+ errdetail("Please report this to <pgsql-bugs@postgresql.org>.")));
+
+ return -1;
+}
+#endif
+#endif /* WIN32 */
#if (defined(HAVE_LANGINFO_H) && defined(CODESET)) || defined(WIN32)
/*
* Given a setting for LC_CTYPE, return the Postgres ID of the associated
- * encoding, if we can determine it.
+ * encoding, if we can determine it. Return -1 if we can't determine it.
*
* Pass in NULL to get the encoding for the current locale setting.
+ * Pass "" to get the encoding selected by the server's environment.
*
* If the result is PG_SQL_ASCII, callers should treat it as being compatible
- * with any desired encoding. We return this if the locale is C/POSIX or we
- * can't determine the encoding.
+ * with any desired encoding.
+ *
+ * If running in the backend and write_message is false, this function must
+ * cope with the possibility that elog() and palloc() are not yet usable.
*/
int
-pg_get_encoding_from_locale(const char *ctype)
+pg_get_encoding_from_locale(const char *ctype, bool write_message)
{
char *sys;
int i;
char *save;
char *name;
+ /* If locale is C or POSIX, we can allow all encodings */
+ if (pg_strcasecmp(ctype, "C") == 0 ||
+ pg_strcasecmp(ctype, "POSIX") == 0)
+ return PG_SQL_ASCII;
+
save = setlocale(LC_CTYPE, NULL);
if (!save)
- return PG_SQL_ASCII; /* setlocale() broken? */
+ return -1; /* setlocale() broken? */
/* must copy result, or it might change after setlocale */
save = strdup(save);
if (!save)
- return PG_SQL_ASCII; /* out of memory; unlikely */
+ return -1; /* out of memory; unlikely */
name = setlocale(LC_CTYPE, ctype);
if (!name)
{
free(save);
- return PG_SQL_ASCII; /* bogus ctype passed in? */
+ return -1; /* bogus ctype passed in? */
}
#ifndef WIN32
/* much easier... */
ctype = setlocale(LC_CTYPE, NULL);
if (!ctype)
- return PG_SQL_ASCII; /* setlocale() broken? */
+ return -1; /* setlocale() broken? */
+
+ /* If locale is C or POSIX, we can allow all encodings */
+ if (pg_strcasecmp(ctype, "C") == 0 ||
+ pg_strcasecmp(ctype, "POSIX") == 0)
+ return PG_SQL_ASCII;
+
#ifndef WIN32
sys = nl_langinfo(CODESET);
if (sys)
}
if (!sys)
- return PG_SQL_ASCII; /* out of memory; unlikely */
-
- /* If locale is C or POSIX, we can allow all encodings */
- if (pg_strcasecmp(ctype, "C") == 0 || pg_strcasecmp(ctype, "POSIX") == 0)
- {
- free(sys);
- return PG_SQL_ASCII;
- }
+ return -1; /* out of memory; unlikely */
/* Check the table */
for (i = 0; encoding_match_list[i].system_enc_name; i++)
/* Special-case kluges for particular platforms go here */
#ifdef __darwin__
+
/*
* Current OS X has many locales that report an empty string for CODESET,
* but they all seem to actually use UTF-8.
/*
* We print a warning if we got a CODESET string but couldn't recognize
- * it. This means we need another entry in the table.
+ * it. This means we need another entry in the table.
*/
+ if (write_message)
+ {
#ifdef FRONTEND
- fprintf(stderr, _("could not determine encoding for locale \"%s\": codeset is \"%s\""),
- ctype, sys);
- /* keep newline separate so there's only one translatable string */
- fputc('\n', stderr);
+ fprintf(stderr, _("could not determine encoding for locale \"%s\": codeset is \"%s\""),
+ ctype, sys);
+ /* keep newline separate so there's only one translatable string */
+ fputc('\n', stderr);
#else
- ereport(WARNING,
- (errmsg("could not determine encoding for locale \"%s\": codeset is \"%s\"",
- ctype, sys),
- errdetail("Please report this to <pgsql-bugs@postgresql.org>.")));
+ ereport(WARNING,
+ (errmsg("could not determine encoding for locale \"%s\": codeset is \"%s\"",
+ ctype, sys),
+ errdetail("Please report this to <pgsql-bugs@postgresql.org>.")));
#endif
+ }
free(sys);
- return PG_SQL_ASCII;
+ return -1;
}
-
-#else /* (HAVE_LANGINFO_H && CODESET) || WIN32 */
+#else /* (HAVE_LANGINFO_H && CODESET) || WIN32 */
/*
- * stub if no platform support
+ * stub if no multi-language platform support
+ *
+ * Note: we could return -1 here, but that would have the effect of
+ * forcing users to specify an encoding to initdb on such platforms.
+ * It seems better to silently default to SQL_ASCII.
*/
int
-pg_get_encoding_from_locale(const char *ctype)
+pg_get_encoding_from_locale(const char *ctype, bool write_message)
{
return PG_SQL_ASCII;
}
-#endif /* (HAVE_LANGINFO_H && CODESET) || WIN32 */
+#endif /* (HAVE_LANGINFO_H && CODESET) || WIN32 */