diplodocus.org Git - nmh/blob - sbr/check_charset.c

   1
   2 /*
   3  * check_charset.c -- routines for character sets
   4  *
   5  * This code is Copyright (c) 2002, by the authors of nmh.  See the
   6  * COPYRIGHT file in the root directory of the nmh distribution for
   7  * complete copyright information.
   8  */
   9
  10 #include <h/mh.h>
  11
  12 #include <langinfo.h>
  13
  14
  15 /*
  16  * Get the current character set
  17  */
  18 char *
  19 get_charset ()
  20 {
  21     return norm_charmap(nl_langinfo (CODESET));
  22 }
  23
  24
  25 /*
  26  * Check if we can display a given character set natively.
  27  * We are passed the length of the initial part of the
  28  * string to check, since we want to allow the name of the
  29  * character set to be a substring of a larger string.
  30  */
  31
  32 int
  33 check_charset (char *str, int len)
  34 {
  35     static char *mm_charset = NULL;
  36     static char *alt_charset = NULL;
  37     static int mm_len;
  38     static int alt_len;
  39
  40     /* Cache the name of our default character set */
  41     if (!mm_charset) {
  42         if (!(mm_charset = get_charset ()))
  43             mm_charset = "US-ASCII";
  44         mm_len = strlen (mm_charset);
  45
  46         /* US-ASCII is a subset of the ISO-8859-X and UTF-8 character sets */
  47         if (!strncasecmp("ISO-8859-", mm_charset, 9) ||
  48                 !strcasecmp("UTF-8", mm_charset)) {
  49             alt_charset = "US-ASCII";
  50             alt_len = strlen (alt_charset);
  51         }
  52     }
  53
  54     /* Check if character set is OK */
  55     if ((len == mm_len) && !strncasecmp(str, mm_charset, mm_len))
  56         return 1;
  57     if (alt_charset && (len == alt_len) && !strncasecmp(str, alt_charset, alt_len))
  58         return 1;
  59
  60     return 0;
  61 }
  62
  63
  64 /*
  65  * Return the name of the character set we are
  66  * using for 8bit text.
  67  */
  68 char *
  69 write_charset_8bit (void)
  70 {
  71     static char *mm_charset = NULL;
  72
  73     /*
  74      * Cache the name of the character set to
  75      * use for 8bit text.
  76      */
  77     if (!mm_charset && !(mm_charset = get_charset ()))
  78             mm_charset = "x-unknown";
  79
  80     return mm_charset;
  81 }