1 /* Unicode character output to streams with locale dependent encoding.
3 Copyright (C) 2000-2003, 2006, 2008 Free Software Foundation, Inc.
5 This program is free software: you can redistribute it and/or modify
6 it under the terms of the GNU General Public License as published by
7 the Free Software Foundation; either version 3 of the License, or
8 (at your option) any later version.
10 This program is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 GNU General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with this program. If not, see <http://www.gnu.org/licenses/>. */
18 /* Written by Bruno Haible <haible@clisp.cons.org>. */
23 #include "unicodeio.h"
36 #define _(msgid) gettext (msgid)
37 #define N_(msgid) msgid
39 #include "localcharset.h"
42 /* When we pass a Unicode character to iconv(), we must pass it in a
43 suitable encoding. The standardized Unicode encodings are
44 UTF-8, UCS-2, UCS-4, UTF-16, UTF-16BE, UTF-16LE, UTF-7.
45 UCS-2 supports only characters up to \U0000FFFF.
46 UTF-16 and variants support only characters up to \U0010FFFF.
47 UTF-7 is way too complex and not supported by glibc-2.1.
48 UCS-4 specification leaves doubts about endianness and byte order
49 mark. glibc currently interprets it as big endian without byte order
50 mark, but this is not backed by an RFC.
51 So we use UTF-8. It supports characters up to \U7FFFFFFF and is
52 unambiguously defined. */
54 /* Luckily, the encoding's name is platform independent. */
55 #define UTF8_NAME "UTF-8"
57 /* Converts the Unicode character CODE to its multibyte representation
58 in the current locale and calls the SUCCESS callback on the resulting
59 byte sequence. If an error occurs, invokes the FAILURE callback instead,
60 passing it CODE and an English error string.
61 Returns whatever the callback returned.
62 Assumes that the locale doesn't change between two calls. */
64 unicode_to_mb (unsigned int code,
65 long (*success) (const char *buf, size_t buflen,
67 long (*failure) (unsigned int code, const char *msg,
71 static int initialized;
74 static iconv_t utf8_to_local;
82 const char *charset = locale_charset ();
84 is_utf8 = !strcmp (charset, UTF8_NAME);
88 utf8_to_local = iconv_open (charset, UTF8_NAME);
89 if (utf8_to_local == (iconv_t)(-1))
90 /* For an unknown encoding, assume ASCII. */
91 utf8_to_local = iconv_open ("ASCII", UTF8_NAME);
97 /* Test whether the utf8_to_local converter is available at all. */
101 if (utf8_to_local == (iconv_t)(-1))
102 return failure (code, N_("iconv function not usable"), callback_arg);
104 return failure (code, N_("iconv function not available"), callback_arg);
108 /* Convert the character to UTF-8. */
109 count = u8_uctomb ((unsigned char *) inbuf, code, sizeof (inbuf));
111 return failure (code, N_("character out of range"), callback_arg);
126 outbytesleft = sizeof (outbuf);
128 /* Convert the character from UTF-8 to the locale's charset. */
129 res = iconv (utf8_to_local,
130 (ICONV_CONST char **)&inptr, &inbytesleft,
131 &outptr, &outbytesleft);
132 if (inbytesleft > 0 || res == (size_t)(-1)
133 /* Irix iconv() inserts a NUL byte if it cannot convert. */
134 # if !defined _LIBICONV_VERSION && (defined sgi || defined __sgi)
135 || (res > 0 && code != 0 && outptr - outbuf == 1 && *outbuf == '\0')
138 return failure (code, NULL, callback_arg);
140 /* Avoid glibc-2.1 bug and Solaris 7 bug. */
141 # if defined _LIBICONV_VERSION \
142 || !((__GLIBC__ - 0 == 2 && __GLIBC_MINOR__ - 0 <= 1) || defined __sun)
144 /* Get back to the initial shift state. */
145 res = iconv (utf8_to_local, NULL, NULL, &outptr, &outbytesleft);
146 if (res == (size_t)(-1))
147 return failure (code, NULL, callback_arg);
150 return success (outbuf, outptr - outbuf, callback_arg);
154 /* At this point, is_utf8 is true, so no conversion is needed. */
155 return success (inbuf, count, callback_arg);
158 /* Simple success callback that outputs the converted string.
159 The STREAM is passed as callback_arg. */
161 fwrite_success_callback (const char *buf, size_t buflen, void *callback_arg)
163 FILE *stream = (FILE *) callback_arg;
165 fwrite (buf, 1, buflen, stream);
169 /* Simple failure callback that displays an error and exits. */
171 exit_failure_callback (unsigned int code, const char *msg,
172 void *callback_arg _UNUSED_PARAMETER_)
175 error (1, 0, _("cannot convert U+%04X to local character set"), code);
177 error (1, 0, _("cannot convert U+%04X to local character set: %s"), code,
182 /* Simple failure callback that displays a fallback representation in plain
183 ASCII, using the same notation as ISO C99 strings. */
185 fallback_failure_callback (unsigned int code,
186 const char *msg _UNUSED_PARAMETER_,
189 FILE *stream = (FILE *) callback_arg;
192 fprintf (stream, "\\u%04X", code);
194 fprintf (stream, "\\U%08X", code);
198 /* Outputs the Unicode character CODE to the output stream STREAM.
199 Upon failure, exit if exit_on_error is true, otherwise output a fallback
202 print_unicode_char (FILE *stream, unsigned int code, int exit_on_error)
204 unicode_to_mb (code, fwrite_success_callback,
206 ? exit_failure_callback
207 : fallback_failure_callback,