1 /* IRI related functions.
2 Copyright (C) 2008 Free Software Foundation, Inc.
4 This file is part of GNU Wget.
6 GNU Wget is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 3 of the License, or (at
9 your option) any later version.
11 GNU Wget is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
16 You should have received a copy of the GNU General Public License
17 along with Wget. If not, see <http://www.gnu.org/licenses/>.
19 Additional permission under GNU GPL version 3 section 7
21 If you modify this program, or any covered work, by linking or
22 combining it with the OpenSSL project's OpenSSL library (or a
23 modified version of that library), containing parts covered by the
24 terms of the OpenSSL or SSLeay licenses, the Free Software Foundation
25 grants you additional permission to convey the resulting work.
26 Corresponding Source for a non-source form of such a combination
27 shall include the source code for the parts of OpenSSL used as well
28 as that of the covered work. */
37 #include <stringprep.h>
45 static iconv_t locale2utf8;
48 static bool open_locale_to_utf8 (void);
49 static bool do_conversion (iconv_t cd, char *in, size_t inlen, char **out);
52 /* Given a string containing "charset=XXX", return the encoding if found,
55 parse_charset (char *str)
62 str = strcasestr (str, "charset=");
69 /* sXXXav: which chars should be banned ??? */
70 while (*charset && !c_isspace (*charset))
73 /* sXXXav: could strdupdelim return NULL ? */
74 charset = strdupdelim (str, charset);
76 /* Do a minimum check on the charset value */
77 if (!check_encoding_name (charset))
83 logprintf (LOG_VERBOSE, "parse_charset: %s\n", quote (charset));
88 /* Find the locale used, or fall back on a default value */
92 /* sXXXav, made our own function or use libidn one ?! */
93 return (char *) stringprep_locale_charset ();
96 /* Basic check of an encoding name. */
98 check_encoding_name (char *encoding)
104 if (!c_isascii (*s) || c_isspace (*s))
106 logprintf (LOG_VERBOSE, "Encoding %s isn't valid\n", quote (encoding));
116 /* Try opening an iconv_t descriptor for conversion from locale to UTF-8 */
118 open_locale_to_utf8 (void)
123 /* sXXXav : That shouldn't happen, just in case */
126 logprintf (LOG_VERBOSE, "open_locale_to_utf8: locale is unset\n");
127 opt.locale = find_locale ();
133 locale2utf8 = iconv_open ("UTF-8", opt.locale);
134 if (locale2utf8 != (iconv_t)(-1))
137 logprintf (LOG_VERBOSE, "Conversion from %s to %s isn't supported\n",
138 quote (opt.locale), quote ("UTF-8"));
143 /* Return a new string */
145 locale_to_utf8 (const char *str)
149 if (!strcasecmp (opt.locale, "utf-8"))
152 if (!open_locale_to_utf8 ())
155 if (do_conversion (locale2utf8, (char *) str, strlen ((char *) str), &new))
156 return (const char *) new;
163 do_conversion (iconv_t cd, char *in, size_t inlen, char **out)
165 /* sXXXav : hummm hard to guess... */
166 size_t len, done, outlen = inlen * 2;
167 int invalid = 0, tooshort = 0;
170 s = xmalloc (outlen + 1);
175 /* sXXXav : put a maximum looping factor ??? */
178 if (iconv (cd, &in, &inlen, out, &outlen) != (size_t)(-1))
181 *(s + len - outlen - done) = '\0';
185 /* Incomplete or invalid multibyte sequence */
186 if (errno == EINVAL || errno == EILSEQ)
189 logprintf (LOG_VERBOSE,
190 "Incomplete or invalide multibyte sequence encountered\n");
199 else if (errno == E2BIG) /* Output buffer full */
205 outlen = done + inlen * 2;
206 new = xmalloc (outlen + 1);
207 memcpy (new, s, done);
213 else /* Weird, we got an unspecified error */
215 logprintf (LOG_VERBOSE, "Unhandled errno %d\n", errno);
223 /* Try to encode UTF-8 host to ASCII. Return the new domain on success or NULL
225 char *idn_encode (char *host)
230 /* toASCII UTF-8 NULL terminated string */
231 ret = idna_to_ascii_8z (host, &new, 0);
232 if (ret != IDNA_SUCCESS)
234 logprintf (LOG_VERBOSE, "idn_encode failed (%d): %s\n", ret,
235 quote (idna_strerror (ret)));