]> diplodocus.org Git - nmh/blob - sbr/norm_charmap.c
Escape literal leading full stop in man/new.man.
[nmh] / sbr / norm_charmap.c
1 /*
2 * The Single Unix Specification function nl_langinfo(CODESET)
3 * returns the name of the encoding used by the currently selected
4 * locale:
5 *
6 * http://www.opengroup.org/onlinepubs/7908799/xsh/langinfo.h.html
7 *
8 * Unfortunately the encoding names are not yet standardized.
9 * This function knows about the encoding names used on many
10 * different systems and converts them where possible into
11 * the corresponding MIME charset name registered in
12 *
13 * http://www.iana.org/assignments/character-sets
14 *
15 * Please extend it as needed and suggest improvements to the author.
16 *
17 * Markus.Kuhn@cl.cam.ac.uk -- 2002-03-11
18 * Permission to use, copy, modify, and distribute this software
19 * for any purpose and without fee is hereby granted. The author
20 * disclaims all warranties with regard to this software.
21 *
22 * Latest version:
23 *
24 * http://www.cl.cam.ac.uk/~mgk25/ucs/norm_charmap.c
25 */
26
27 #include <string.h>
28
29 #define digit(x) ((x) >= '0' && (x) <= '9')
30
31 static char buf[16];
32
33 char *
34 norm_charmap(char *name)
35 {
36 char *p;
37
38 if (!name)
39 return name;
40
41 /* Many need no remapping, but they are listed here so you
42 * can see what output to expect, and modify for your needs
43 * as necessary. */
44 if (!strcmp(name, "UTF-8"))
45 return "UTF-8";
46 if (!strcmp(name, "EUC-JP"))
47 return "EUC-JP";
48 if (!strcmp(name, "EUC-KR"))
49 return "EUC-KR";
50 if (!strcmp(name, "EUC-TW"))
51 return "EUC-TW";
52 if (!strcmp(name, "KOI8-R"))
53 return "KOI8-R";
54 if (!strcmp(name, "KOI8-U"))
55 return "KOI8-U";
56 if (!strcmp(name, "GBK"))
57 return "GBK";
58 if (!strcmp(name, "GB2312"))
59 return "GB2312";
60 if (!strcmp(name, "GB18030"))
61 return "GB18030";
62 if (!strcmp(name, "VSCII"))
63 return "VSCII";
64
65 /* ASCII comes in many names */
66 if (!strcmp(name, "ASCII") ||
67 !strcmp(name, "US-ASCII") ||
68 !strcmp(name, "ANSI_X3.4-1968") ||
69 !strcmp(name, "646") ||
70 !strcmp(name, "ISO646") ||
71 !strcmp(name, "ISO_646.IRV"))
72 return "US-ASCII";
73
74 /* ISO 8859 will be converted to "ISO-8859-x" */
75 if ((p = strstr(name, "8859-"))) {
76 memcpy(buf, "ISO-8859-\0\0", 12);
77 p += 5;
78 if (digit(*p)) {
79 buf[9] = *p++;
80 if (digit(*p)) buf[10] = *p++;
81 return buf;
82 }
83 }
84
85 /* Windows code pages will be converted to "WINDOWS-12xx" */
86 if ((p = strstr(name, "CP12"))) {
87 memcpy(buf, "WINDOWS-12\0\0", 13);
88 p += 4;
89 if (digit(*p)) {
90 buf[10] = *p++;
91 if (digit(*p)) buf[11] = *p++;
92 return buf;
93 }
94 }
95
96 /* TIS-620 comes in at least the following two forms */
97 if (!strcmp(name, "TIS-620") ||
98 !strcmp(name, "TIS620.2533"))
99 return "ISO-8859-11";
100
101 /* For some, uppercase/lowercase might differ */
102 if (!strcmp(name, "Big5") || !strcmp(name, "BIG5"))
103 return "Big5";
104 if (!strcmp(name, "Big5HKSCS") || !strcmp(name, "BIG5HKSCS"))
105 return "Big5HKSCS";
106
107 /* I don't know of any implementation of nl_langinfo(CODESET) out
108 * there that returns anything else (and I'm not even certain all of
109 * the above occur in the wild), but just in case, as a fallback,
110 * return the unmodified name. */
111 return name;
112 }