]>
Dogcows Code - chaz/tar/blob - lib/unicodeio.c
1 /* Unicode character output to streams with locale dependent encoding.
3 Copyright (C) 2000, 2001 Free Software Foundation, Inc.
5 This program is free software; you can redistribute it and/or modify it
6 under the terms of the GNU Library General Public License as published
7 by the Free Software Foundation; either version 2, or (at your option)
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 GNU
13 Library General Public License for more details.
15 You should have received a copy of the GNU Library General Public
16 License along with this program; if not, write to the Free Software
17 Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
20 /* Written by Bruno Haible <haible@clisp.cons.org>. */
50 # define _(Text) gettext (Text)
55 #include "unicodeio.h"
57 /* When we pass a Unicode character to iconv(), we must pass it in a
58 suitable encoding. The standardized Unicode encodings are
59 UTF-8, UCS-2, UCS-4, UTF-16, UTF-16BE, UTF-16LE, UTF-7.
60 UCS-2 supports only characters up to \U0000FFFF.
61 UTF-16 and variants support only characters up to \U0010FFFF.
62 UTF-7 is way too complex and not supported by glibc-2.1.
63 UCS-4 specification leaves doubts about endianness and byte order
64 mark. glibc currently interprets it as big endian without byte order
65 mark, but this is not backed by an RFC.
66 So we use UTF-8. It supports characters up to \U7FFFFFFF and is
67 unambiguously defined. */
69 /* Stores the UTF-8 representation of the Unicode character wc in r[0..5].
70 Returns the number of bytes stored, or -1 if wc is out of range. */
72 utf8_wctomb (unsigned char *r
, unsigned int wc
)
80 else if (wc
< 0x10000)
82 else if (wc
< 0x200000)
84 else if (wc
< 0x4000000)
86 else if (wc
<= 0x7fffffff)
93 /* Note: code falls through cases! */
94 case 6: r
[5] = 0x80 | (wc
& 0x3f); wc
= wc
>> 6; wc
|= 0x4000000;
95 case 5: r
[4] = 0x80 | (wc
& 0x3f); wc
= wc
>> 6; wc
|= 0x200000;
96 case 4: r
[3] = 0x80 | (wc
& 0x3f); wc
= wc
>> 6; wc
|= 0x10000;
97 case 3: r
[2] = 0x80 | (wc
& 0x3f); wc
= wc
>> 6; wc
|= 0x800;
98 case 2: r
[1] = 0x80 | (wc
& 0x3f); wc
= wc
>> 6; wc
|= 0xc0;
105 /* Luckily, the encoding's name is platform independent. */
106 #define UTF8_NAME "UTF-8"
108 /* Converts the Unicode character CODE to its multibyte representation
109 in the current locale and calls the CALLBACK on the resulting byte
111 Assumes that the locale doesn't change between two calls. */
113 unicode_to_mb (unsigned int code
,
114 void (*callback
) PARAMS((const char *buf
, size_t buflen
,
115 void *callback_arg
)),
118 static int initialized
;
121 static iconv_t utf8_to_local
;
129 extern const char *locale_charset
PARAMS ((void));
130 const char *charset
= locale_charset ();
132 is_utf8
= !strcmp (charset
, UTF8_NAME
);
136 utf8_to_local
= iconv_open (charset
, UTF8_NAME
);
137 if (utf8_to_local
== (iconv_t
)(-1))
139 /* For an unknown encoding, assume ASCII. */
140 utf8_to_local
= iconv_open ("ASCII", UTF8_NAME
);
141 if (utf8_to_local
== (iconv_t
)(-1))
143 _("cannot convert U+%04X to local character set: iconv function not usable"),
151 /* Convert the character to UTF-8. */
152 count
= utf8_wctomb ((unsigned char *) inbuf
, code
);
154 error (1, 0, _("U+%04X: character out of range"), code
);
158 callback (inbuf
, count
, callback_arg
);
173 outbytesleft
= sizeof (outbuf
);
175 /* Convert the character from UTF-8 to the locale's charset. */
176 res
= iconv (utf8_to_local
,
177 (ICONV_CONST
char **)&inptr
, &inbytesleft
,
178 &outptr
, &outbytesleft
);
179 if (inbytesleft
> 0 || res
== (size_t)(-1)
180 /* Irix iconv() inserts a NUL byte if it cannot convert. */
181 # if !defined _LIBICONV_VERSION && (defined sgi || defined __sgi)
182 || (res
> 0 && code
!= 0 && outptr
- outbuf
== 1 && *outbuf
== '\0')
185 error (1, res
== (size_t)(-1) ? errno
: 0,
186 _("cannot convert U+%04X to local character set"), code
);
188 /* Avoid glibc-2.1 bug and Solaris 2.7 bug. */
189 # if defined _LIBICONV_VERSION \
190 || !((__GLIBC__ - 0 == 2 && __GLIBC_MINOR__ - 0 <= 1) || defined __sun)
192 /* Get back to the initial shift state. */
193 res
= iconv (utf8_to_local
, NULL
, NULL
, &outptr
, &outbytesleft
);
194 if (res
== (size_t)(-1))
195 error (1, errno
, _("cannot convert U+%04X to local character set"),
199 callback (outbuf
, outptr
- outbuf
, callback_arg
);
202 _("cannot convert U+%04X to local character set: iconv function not available"),
208 /* Simple callback that outputs the converted string.
209 The STREAM is passed as callback_arg. */
211 fprintf_callback (const char *buf
, size_t buflen
, void *callback_arg
)
213 FILE *stream
= (FILE *) callback_arg
;
215 fwrite (buf
, 1, buflen
, stream
);
218 /* Outputs the Unicode character CODE to the output stream STREAM.
219 Assumes that the locale doesn't change between two calls. */
221 print_unicode_char (FILE *stream
, unsigned int code
)
223 unicode_to_mb (code
, fprintf_callback
, stream
);
This page took 0.041571 seconds and 4 git commands to generate.