From 9ae85b8c071909a5dceaf50cf14b84916140f7d6 Mon Sep 17 00:00:00 2001 From: Jeffrey Stedfast Date: Tue, 25 Feb 2003 22:37:11 +0000 Subject: New internal function to parse the locale language code. (e_iconv_init): 2003-02-25 Jeffrey Stedfast * gal/util/e-iconv.c (locale_parse_lang): New internal function to parse the locale language code. (e_iconv_init): Parse and set the locale_lang variable. (e_iconv_locale_language): Return the locale_lang. (e_iconv_charset_language): Map some CJKR charsets to a language code. svn path=/trunk/; revision=20063 --- e-util/e-iconv.c | 130 +++++++++++++++++++++++++++++++++++++++++++------------ e-util/e-iconv.h | 5 +++ 2 files changed, 107 insertions(+), 28 deletions(-) diff --git a/e-util/e-iconv.c b/e-util/e-iconv.c index f6823538b4..409224f006 100644 --- a/e-util/e-iconv.c +++ b/e-util/e-iconv.c @@ -35,10 +35,6 @@ #include -#ifdef HAVE_ALLOCA_H -#include -#endif - #ifdef HAVE_CODESET #include #endif @@ -97,6 +93,7 @@ static unsigned int iconv_cache_size = 0; static GHashTable *iconv_charsets = NULL; static char *locale_charset = NULL; +static char *locale_lang = NULL; struct { char *charset; @@ -189,21 +186,42 @@ static EDListNode *e_dlist_remove(EDListNode *n) return n; } -/* - we don't want to lower charset names in native locale (look what does in such case ISO-8859-9), - but rather lower them in "pure" C locale - this fixes bug #24179 -*/ static void -C_g_strdown (gchar *string) +locale_parse_lang (const char *locale) { - gchar *old_locale; - - old_locale = g_strdup (setlocale (LC_CTYPE, NULL)); - setlocale (LC_CTYPE, "C"); - g_strdown (string); - setlocale (LC_CTYPE, old_locale); - g_free (old_locale); + char *codeset, *lang; + + if ((codeset = strchr (locale, '.'))) + lang = g_strndup (locale, codeset - locale); + else + lang = g_strdup (locale); + + /* validate the language */ + if (strlen (lang) >= 2) { + if (lang[2] == '-' || lang[2] == '_') { + /* canonicalise the lang */ + g_ascii_strdown (lang, 2); + + /* validate the country code */ + if (strlen (lang + 3) > 2) { + /* invalid country code */ + lang[2] = '\0'; + } else { + lang[2] = '-'; + g_ascii_strup (lang + 3, 2); + } + } else if (lang[2] != '\0') { + /* invalid language */ + g_free (lang); + lang = NULL; + } + + locale_lang = lang; + } else { + /* invalid language */ + locale_lang = NULL; + g_free (lang); + } } /* NOTE: Owns the lock on return if keep is TRUE ! */ @@ -226,7 +244,7 @@ e_iconv_init(int keep) for (i = 0; known_iconv_charsets[i].charset != NULL; i++) { from = g_strdup(known_iconv_charsets[i].charset); to = g_strdup(known_iconv_charsets[i].iconv_name); - C_g_strdown(from); + g_ascii_strdown (from, -1); g_hash_table_insert(iconv_charsets, from, to); } @@ -235,7 +253,7 @@ e_iconv_init(int keep) iconv_cache_open = g_hash_table_new(NULL, NULL); locale = setlocale (LC_ALL, NULL); - + if (!locale || !strcmp (locale, "C") || !strcmp (locale, "POSIX")) { /* The locale "C" or "POSIX" is a portable locale; its * LC_CTYPE part corresponds to the 7-bit ASCII character @@ -243,10 +261,11 @@ e_iconv_init(int keep) */ locale_charset = NULL; + locale_lang = NULL; } else { #ifdef HAVE_CODESET - locale_charset = g_strdup(nl_langinfo(CODESET)); - C_g_strdown(locale_charset); + locale_charset = g_strdup (nl_langinfo (CODESET)); + g_ascii_strdown (locale_charset, -1); #else /* A locale name is typically of the form language[_terri- * tory][.codeset][@modifier], where language is an ISO 639 @@ -263,12 +282,16 @@ e_iconv_init(int keep) /* ; is a hack for debian systems and / is a hack for Solaris systems */ for (p = codeset; *p && !strchr ("@;/", *p); p++); locale_charset = g_strndup (codeset, p - codeset); - C_g_strdown (locale_charset); + g_ascii_strdown (locale_charset, -1); } else { /* charset unknown */ locale_charset = NULL; } -#endif +#endif + + /* parse the locale lang */ + locale_parse_lang (locale); + } if (!keep) @@ -282,10 +305,10 @@ const char *e_iconv_charset_name(const char *charset) if (charset == NULL) return NULL; - name = alloca(strlen(charset)+1); - strcpy(name, charset); - C_g_strdown(name); - + name = g_alloca (strlen (charset) + 1); + strcpy (name, charset); + g_ascii_strdown (name, -1); + e_iconv_init(TRUE); ret = g_hash_table_lookup(iconv_charsets, name); if (ret != NULL) { @@ -385,7 +408,7 @@ iconv_t e_iconv_open(const char *oto, const char *ofrom) to = e_iconv_charset_name (oto); from = e_iconv_charset_name (ofrom); - tofrom = alloca(strlen(to) +strlen(from) + 2); + tofrom = g_alloca (strlen (to) + strlen (from) + 2); sprintf(tofrom, "%s%%%s", to, from); LOCK(); @@ -496,3 +519,54 @@ const char *e_iconv_locale_charset(void) return locale_charset; } + +const char * +e_iconv_locale_language (void) +{ + e_iconv_init (FALSE); + + return locale_lang; +} + +/* map CJKR charsets to their language code */ +/* NOTE: only support charset names that will be returned by + * e_iconv_charset_name() so that we don't have to keep track of all + * the aliases too. */ +static struct { + char *charset; + char *lang; +} cjkr_lang_map[] = { + { "Big5", "zh" }, + { "BIG5HKCS", "zh" }, + { "gb2312", "zh" }, + { "gb18030", "zh" }, + { "gbk", "zh" }, + { "euc-tw", "zh" }, + { "iso-2022-jp", "ja" }, + { "sjis", "ja" }, + { "ujis", "ja" }, + { "eucJP", "ja" }, + { "euc-jp", "ja" }, + { "euc-kr", "ko" }, + { "koi8-r", "ru" }, + { "koi8-u", "uk" } +}; + +#define NUM_CJKR_LANGS (sizeof (cjkr_lang_map) / sizeof (cjkr_lang_map[0])) + +const char * +e_iconv_charset_language (const char *charset) +{ + int i; + + if (!charset) + return NULL; + + charset = e_iconv_charset_name (charset); + for (i = 0; i < NUM_CJKR_LANGS; i++) { + if (!strcasecmp (cjkr_lang_map[i].charset, charset)) + return cjkr_lang_map[i].lang; + } + + return NULL; +} diff --git a/e-util/e-iconv.h b/e-util/e-iconv.h index daf97a30ab..14b93853d5 100644 --- a/e-util/e-iconv.h +++ b/e-util/e-iconv.h @@ -5,6 +5,7 @@ * * Authors: * Michael Zucchi + * Jeffrey Stedfast * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Library General Public @@ -37,6 +38,10 @@ size_t e_iconv(iconv_t cd, const char **inbuf, size_t *inbytesleft, char ** outb void e_iconv_close(iconv_t ip); const char *e_iconv_locale_charset(void); +/* languages */ +const char *e_iconv_locale_language (void); +const char *e_iconv_charset_language (const char *charset); + #ifdef __cplusplus } #endif /* __cplusplus */ -- cgit v1.2.3