| 1 |
/* IRI related functions. |
| 2 |
Copyright (C) 2008-2011, 2015, 2018-2024 Free Software Foundation, |
| 3 |
Inc. |
| 4 |
|
| 5 |
This file is part of GNU Wget. |
| 6 |
|
| 7 |
GNU Wget is free software; you can redistribute it and/or modify |
| 8 |
it under the terms of the GNU General Public License as published by |
| 9 |
the Free Software Foundation; either version 3 of the License, or (at |
| 10 |
your option) any later version. |
| 11 |
|
| 12 |
GNU Wget is distributed in the hope that it will be useful, |
| 13 |
but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 |
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 15 |
GNU General Public License for more details. |
| 16 |
|
| 17 |
You should have received a copy of the GNU General Public License |
| 18 |
along with Wget. If not, see <http://www.gnu.org/licenses/>. |
| 19 |
|
| 20 |
Additional permission under GNU GPL version 3 section 7 |
| 21 |
|
| 22 |
If you modify this program, or any covered work, by linking or |
| 23 |
combining it with the OpenSSL project's OpenSSL library (or a |
| 24 |
modified version of that library), containing parts covered by the |
| 25 |
terms of the OpenSSL or SSLeay licenses, the Free Software Foundation |
| 26 |
grants you additional permission to convey the resulting work. |
| 27 |
Corresponding Source for a non-source form of such a combination |
| 28 |
shall include the source code for the parts of OpenSSL used as well |
| 29 |
as that of the covered work. */ |
| 30 |
|
| 31 |
#include "wget.h" |
| 32 |
|
| 33 |
#include <stdio.h> |
| 34 |
#include <stdlib.h> |
| 35 |
#include <string.h> |
| 36 |
#include <langinfo.h> |
| 37 |
#include <errno.h> |
| 38 |
#ifdef HAVE_ICONV |
| 39 |
# include <iconv.h> |
| 40 |
#endif |
| 41 |
#include <idn2.h> |
| 42 |
#if IDN2_VERSION_NUMBER < 0x00140000 |
| 43 |
# include <unicase.h> |
| 44 |
# include <unistr.h> |
| 45 |
#endif |
| 46 |
|
| 47 |
#include "utils.h" |
| 48 |
#include "url.h" |
| 49 |
#include "c-strcase.h" |
| 50 |
#include "c-strcasestr.h" |
| 51 |
#include "xstrndup.h" |
| 52 |
|
| 53 |
/* Note: locale encoding is kept in options struct (opt.locale) */ |
| 54 |
|
| 55 |
/* Given a string containing "charset=XXX", return the encoding if found, |
| 56 |
or NULL otherwise */ |
| 57 |
char * |
| 58 |
parse_charset (const char *str) |
| 59 |
{ |
| 60 |
const char *end; |
| 61 |
char *charset; |
| 62 |
|
| 63 |
if (!str || !*str) |
| 64 |
return NULL; |
| 65 |
|
| 66 |
str = c_strcasestr (str, "charset="); |
| 67 |
if (!str) |
| 68 |
return NULL; |
| 69 |
|
| 70 |
str += 8; |
| 71 |
end = str; |
| 72 |
|
| 73 |
/* sXXXav: which chars should be banned ??? */ |
| 74 |
while (*end && !c_isspace (*end)) |
| 75 |
end++; |
| 76 |
|
| 77 |
/* sXXXav: could strdupdelim return NULL ? */ |
| 78 |
charset = strdupdelim (str, end); |
| 79 |
|
| 80 |
/* Do a minimum check on the charset value */ |
| 81 |
if (!check_encoding_name (charset)) |
| 82 |
{ |
| 83 |
xfree (charset); |
| 84 |
return NULL; |
| 85 |
} |
| 86 |
|
| 87 |
/*logprintf (LOG_VERBOSE, "parse_charset: %s\n", quote (charset));*/ |
| 88 |
|
| 89 |
return charset; |
| 90 |
} |
| 91 |
|
| 92 |
/* Find the locale used, or fall back on a default value */ |
| 93 |
const char * |
| 94 |
find_locale (void) |
| 95 |
{ |
| 96 |
const char *encoding = nl_langinfo(CODESET); |
| 97 |
|
| 98 |
if (!encoding || !*encoding) |
| 99 |
return xstrdup("ASCII"); |
| 100 |
|
| 101 |
return xstrdup(encoding); |
| 102 |
} |
| 103 |
|
| 104 |
/* Basic check of an encoding name. */ |
| 105 |
bool |
| 106 |
check_encoding_name (const char *encoding) |
| 107 |
{ |
| 108 |
const char *s = encoding; |
| 109 |
|
| 110 |
while (*s) |
| 111 |
{ |
| 112 |
if (!c_isascii (*s) || c_isspace (*s)) |
| 113 |
{ |
| 114 |
logprintf (LOG_VERBOSE, _("Encoding %s isn't valid\n"), quote (encoding)); |
| 115 |
return false; |
| 116 |
} |
| 117 |
|
| 118 |
s++; |
| 119 |
} |
| 120 |
|
| 121 |
return true; |
| 122 |
} |
| 123 |
|
| 124 |
#ifdef HAVE_ICONV |
| 125 |
/* Do the conversion according to the passed conversion descriptor cd. *out |
| 126 |
will contain the transcoded string on success. *out content is |
| 127 |
unspecified otherwise. */ |
| 128 |
static bool |
| 129 |
do_conversion (const char *tocode, const char *fromcode, char const *in_org, size_t inlen, char **out) |
| 130 |
{ |
| 131 |
iconv_t cd; |
| 132 |
/* sXXXav : hummm hard to guess... */ |
| 133 |
size_t len, done, outlen; |
| 134 |
int invalid = 0, tooshort = 0; |
| 135 |
char *s, *in, *in_save; |
| 136 |
|
| 137 |
cd = iconv_open (tocode, fromcode); |
| 138 |
if (cd == (iconv_t)(-1)) |
| 139 |
{ |
| 140 |
logprintf (LOG_VERBOSE, _("Conversion from %s to %s isn't supported\n"), |
| 141 |
quote_n (0, fromcode), quote_n (1, tocode)); |
| 142 |
*out = NULL; |
| 143 |
return false; |
| 144 |
} |
| 145 |
|
| 146 |
/* iconv() has to work on an unescaped string */ |
| 147 |
in_save = in = xstrndup (in_org, inlen); |
| 148 |
url_unescape_except_reserved (in); |
| 149 |
inlen = strlen(in); |
| 150 |
|
| 151 |
len = outlen = inlen * 2; |
| 152 |
*out = s = xmalloc (outlen + 1); |
| 153 |
done = 0; |
| 154 |
|
| 155 |
for (;;) |
| 156 |
{ |
| 157 |
if (iconv (cd, (ICONV_CONST char **) &in, &inlen, out, &outlen) != (size_t)(-1) && |
| 158 |
iconv (cd, NULL, NULL, out, &outlen) != (size_t)(-1)) |
| 159 |
{ |
| 160 |
*out = s; |
| 161 |
*(s + len - outlen - done) = '\0'; |
| 162 |
xfree(in_save); |
| 163 |
iconv_close(cd); |
| 164 |
IF_DEBUG |
| 165 |
{ |
| 166 |
/* not not print out embedded passwords, in_org might be an URL */ |
| 167 |
if (!strchr(in_org, '@') && !strchr(*out, '@')) |
| 168 |
debug_logprintf ("converted '%s' (%s) -> '%s' (%s)\n", in_org, fromcode, *out, tocode); |
| 169 |
else |
| 170 |
debug_logprintf ("logging suppressed, strings may contain password\n"); |
| 171 |
} |
| 172 |
return true; |
| 173 |
} |
| 174 |
|
| 175 |
/* Incomplete or invalid multibyte sequence */ |
| 176 |
if (errno == EINVAL || errno == EILSEQ) |
| 177 |
{ |
| 178 |
if (!invalid) |
| 179 |
logprintf (LOG_VERBOSE, |
| 180 |
_("Incomplete or invalid multibyte sequence encountered\n")); |
| 181 |
|
| 182 |
invalid++; |
| 183 |
**out = *in; |
| 184 |
in++; |
| 185 |
inlen--; |
| 186 |
(*out)++; |
| 187 |
outlen--; |
| 188 |
} |
| 189 |
else if (errno == E2BIG) /* Output buffer full */ |
| 190 |
{ |
| 191 |
tooshort++; |
| 192 |
done = len; |
| 193 |
len = done + inlen * 2; |
| 194 |
char *s_new = xrealloc (s, len + 1); |
| 195 |
if (s_new != s) |
| 196 |
{ |
| 197 |
/* Help valgrind by initializing additional memory. */ |
| 198 |
s = s_new; |
| 199 |
memset(s + done, 0, inlen * 2 + 1); |
| 200 |
} |
| 201 |
*out = s + done - outlen; |
| 202 |
outlen += inlen * 2; |
| 203 |
} |
| 204 |
else /* Weird, we got an unspecified error */ |
| 205 |
{ |
| 206 |
logprintf (LOG_VERBOSE, _("Unhandled errno %d\n"), errno); |
| 207 |
break; |
| 208 |
} |
| 209 |
} |
| 210 |
|
| 211 |
xfree(in_save); |
| 212 |
iconv_close(cd); |
| 213 |
IF_DEBUG |
| 214 |
{ |
| 215 |
/* not not print out embedded passwords, in_org might be an URL */ |
| 216 |
if (!strchr(in_org, '@') && !strchr(*out, '@')) |
| 217 |
debug_logprintf ("converted '%s' (%s) -> '%s' (%s)\n", in_org, fromcode, *out, tocode); |
| 218 |
else |
| 219 |
debug_logprintf ("logging suppressed, strings may contain password\n"); |
| 220 |
} |
| 221 |
return false; |
| 222 |
} |
| 223 |
#else |
| 224 |
static bool |
| 225 |
do_conversion (const char *tocode _GL_UNUSED, const char *fromcode _GL_UNUSED, |
| 226 |
char const *in_org _GL_UNUSED, size_t inlen _GL_UNUSED, char **out) |
| 227 |
{ |
| 228 |
*out = NULL; |
| 229 |
return false; |
| 230 |
} |
| 231 |
#endif |
| 232 |
|
| 233 |
/* Try converting string str from locale to UTF-8. Return a new string |
| 234 |
on success, or str on error or if conversion isn't needed. */ |
| 235 |
const char * |
| 236 |
locale_to_utf8 (const char *str) |
| 237 |
{ |
| 238 |
char *new; |
| 239 |
|
| 240 |
/* That shouldn't happen, just in case */ |
| 241 |
if (!opt.locale) |
| 242 |
{ |
| 243 |
logprintf (LOG_VERBOSE, _("locale_to_utf8: locale is unset\n")); |
| 244 |
opt.locale = find_locale (); |
| 245 |
} |
| 246 |
|
| 247 |
if (!opt.locale || !c_strcasecmp (opt.locale, "utf-8")) |
| 248 |
return str; |
| 249 |
|
| 250 |
if (do_conversion ("UTF-8", opt.locale, (char *) str, strlen ((char *) str), &new)) |
| 251 |
return (const char *) new; |
| 252 |
|
| 253 |
xfree (new); |
| 254 |
return str; |
| 255 |
} |
| 256 |
|
| 257 |
/* Try to "ASCII encode" UTF-8 host. Return the new domain on success or NULL |
| 258 |
on error. */ |
| 259 |
char * |
| 260 |
idn_encode (const struct iri *i, const char *host) |
| 261 |
{ |
| 262 |
int ret; |
| 263 |
char *ascii_encoded; |
| 264 |
char *utf8_encoded = NULL; |
| 265 |
const char *src; |
| 266 |
#if IDN2_VERSION_NUMBER < 0x00140000 |
| 267 |
uint8_t *lower; |
| 268 |
size_t len = 0; |
| 269 |
#endif |
| 270 |
|
| 271 |
/* Encode to UTF-8 if not done */ |
| 272 |
if (!i->utf8_encode) |
| 273 |
{ |
| 274 |
if (!remote_to_utf8 (i, host, &utf8_encoded)) |
| 275 |
return NULL; /* Nothing to encode or an error occurred */ |
| 276 |
src = utf8_encoded; |
| 277 |
} |
| 278 |
else |
| 279 |
src = host; |
| 280 |
|
| 281 |
#if IDN2_VERSION_NUMBER >= 0x00140000 |
| 282 |
/* IDN2_TRANSITIONAL implies input NFC encoding */ |
| 283 |
ret = idn2_lookup_u8 ((uint8_t *) src, (uint8_t **) &ascii_encoded, IDN2_NONTRANSITIONAL); |
| 284 |
if (ret != IDN2_OK) |
| 285 |
/* fall back to TR46 Transitional mode, max IDNA2003 compatibility */ |
| 286 |
ret = idn2_lookup_u8 ((uint8_t *) src, (uint8_t **) &ascii_encoded, IDN2_TRANSITIONAL); |
| 287 |
|
| 288 |
if (ret != IDN2_OK) |
| 289 |
logprintf (LOG_VERBOSE, _("idn_encode failed (%d): %s\n"), ret, |
| 290 |
quote (idn2_strerror (ret))); |
| 291 |
#else |
| 292 |
/* we need a conversion to lowercase */ |
| 293 |
lower = u8_tolower ((uint8_t *) src, u8_strlen ((uint8_t *) src) + 1, 0, UNINORM_NFKC, NULL, &len); |
| 294 |
if (!lower) |
| 295 |
{ |
| 296 |
logprintf (LOG_VERBOSE, _("Failed to convert to lower: %d: %s\n"), |
| 297 |
errno, quote (src)); |
| 298 |
xfree (utf8_encoded); |
| 299 |
return NULL; |
| 300 |
} |
| 301 |
|
| 302 |
if ((ret = idn2_lookup_u8 (lower, (uint8_t **) &ascii_encoded, IDN2_NFC_INPUT)) != IDN2_OK) |
| 303 |
{ |
| 304 |
logprintf (LOG_VERBOSE, _("idn_encode failed (%d): %s\n"), ret, |
| 305 |
quote (idn2_strerror (ret))); |
| 306 |
} |
| 307 |
|
| 308 |
xfree (lower); |
| 309 |
#endif |
| 310 |
|
| 311 |
xfree (utf8_encoded); |
| 312 |
|
| 313 |
if (ret == IDN2_OK && ascii_encoded) |
| 314 |
{ |
| 315 |
char *tmp = xstrdup (ascii_encoded); |
| 316 |
idn2_free (ascii_encoded); |
| 317 |
ascii_encoded = tmp; |
| 318 |
} |
| 319 |
|
| 320 |
return ret == IDN2_OK ? ascii_encoded : NULL; |
| 321 |
} |
| 322 |
|
| 323 |
/* Try to decode an "ASCII encoded" host. Return the new domain in the locale |
| 324 |
on success or NULL on error. */ |
| 325 |
char * |
| 326 |
idn_decode (const char *host) |
| 327 |
{ |
| 328 |
/* |
| 329 |
char *new; |
| 330 |
int ret; |
| 331 |
|
| 332 |
ret = idn2_register_u8 (NULL, host, (uint8_t **) &new, 0); |
| 333 |
if (ret != IDN2_OK) |
| 334 |
{ |
| 335 |
logprintf (LOG_VERBOSE, _("idn2_register_u8 failed (%d): %s: %s\n"), ret, |
| 336 |
quote (idn2_strerror (ret)), host); |
| 337 |
return NULL; |
| 338 |
} |
| 339 |
|
| 340 |
return new; |
| 341 |
*/ |
| 342 |
/* idn2_register_u8() just works label by label. |
| 343 |
* That is pretty much overhead for just displaying the original ulabels. |
| 344 |
* To keep at least the debug output format, return a cloned host. */ |
| 345 |
return xstrdup(host); |
| 346 |
} |
| 347 |
|
| 348 |
/* Try to transcode string str from remote encoding to UTF-8. On success, *new |
| 349 |
contains the transcoded string. *new content is unspecified otherwise. */ |
| 350 |
bool |
| 351 |
remote_to_utf8 (const struct iri *iri, const char *str, char **new) |
| 352 |
{ |
| 353 |
bool ret = false; |
| 354 |
|
| 355 |
if (!iri->uri_encoding) |
| 356 |
return false; |
| 357 |
|
| 358 |
/* When `i->uri_encoding' == "UTF-8" there is nothing to convert. But we must |
| 359 |
test for non-ASCII symbols for correct hostname processing in `idn_encode' |
| 360 |
function. */ |
| 361 |
if (!c_strcasecmp (iri->uri_encoding, "UTF-8")) |
| 362 |
{ |
| 363 |
const unsigned char *p; |
| 364 |
for (p = (unsigned char *) str; *p; p++) |
| 365 |
if (*p > 127) |
| 366 |
{ |
| 367 |
*new = strdup (str); |
| 368 |
return true; |
| 369 |
} |
| 370 |
return false; |
| 371 |
} |
| 372 |
|
| 373 |
if (do_conversion ("UTF-8", iri->uri_encoding, str, strlen (str), new)) |
| 374 |
ret = true; |
| 375 |
|
| 376 |
/* Test if something was converted */ |
| 377 |
if (*new && !strcmp (str, *new)) |
| 378 |
{ |
| 379 |
xfree (*new); |
| 380 |
return false; |
| 381 |
} |
| 382 |
|
| 383 |
return ret; |
| 384 |
} |
| 385 |
|
| 386 |
/* Allocate a new iri structure and return a pointer to it. */ |
| 387 |
struct iri * |
| 388 |
iri_new (void) |
| 389 |
{ |
| 390 |
struct iri *i = xmalloc (sizeof *i); |
| 391 |
i->uri_encoding = opt.encoding_remote ? xstrdup (opt.encoding_remote) : NULL; |
| 392 |
i->content_encoding = NULL; |
| 393 |
i->orig_url = NULL; |
| 394 |
i->utf8_encode = opt.enable_iri; |
| 395 |
return i; |
| 396 |
} |
| 397 |
|
| 398 |
struct iri *iri_dup (const struct iri *src) |
| 399 |
{ |
| 400 |
struct iri *i = xmalloc (sizeof *i); |
| 401 |
i->uri_encoding = src->uri_encoding ? xstrdup (src->uri_encoding) : NULL; |
| 402 |
i->content_encoding = (src->content_encoding ? |
| 403 |
xstrdup (src->content_encoding) : NULL); |
| 404 |
i->orig_url = src->orig_url ? xstrdup (src->orig_url) : NULL; |
| 405 |
i->utf8_encode = src->utf8_encode; |
| 406 |
return i; |
| 407 |
} |
| 408 |
|
| 409 |
/* Completely free an iri structure. */ |
| 410 |
void |
| 411 |
iri_free (struct iri *i) |
| 412 |
{ |
| 413 |
if (i) |
| 414 |
{ |
| 415 |
xfree (i->uri_encoding); |
| 416 |
xfree (i->content_encoding); |
| 417 |
xfree (i->orig_url); |
| 418 |
xfree (i); |
| 419 |
} |
| 420 |
} |
| 421 |
|
| 422 |
/* Set uri_encoding of struct iri i. If a remote encoding was specified, use |
| 423 |
it unless force is true. */ |
| 424 |
void |
| 425 |
set_uri_encoding (struct iri *i, const char *charset, bool force) |
| 426 |
{ |
| 427 |
DEBUGP (("URI encoding = %s\n", charset ? quote (charset) : "None")); |
| 428 |
if (!force && opt.encoding_remote) |
| 429 |
return; |
| 430 |
if (i->uri_encoding) |
| 431 |
{ |
| 432 |
if (charset && !c_strcasecmp (i->uri_encoding, charset)) |
| 433 |
return; |
| 434 |
xfree (i->uri_encoding); |
| 435 |
} |
| 436 |
|
| 437 |
i->uri_encoding = charset ? xstrdup (charset) : NULL; |
| 438 |
} |
| 439 |
|
| 440 |
/* Set content_encoding of struct iri i. */ |
| 441 |
void |
| 442 |
set_content_encoding (struct iri *i, const char *charset) |
| 443 |
{ |
| 444 |
DEBUGP (("URI content encoding = %s\n", charset ? quote (charset) : "None")); |
| 445 |
if (opt.encoding_remote) |
| 446 |
return; |
| 447 |
if (i->content_encoding) |
| 448 |
{ |
| 449 |
if (charset && !c_strcasecmp (i->content_encoding, charset)) |
| 450 |
return; |
| 451 |
xfree (i->content_encoding); |
| 452 |
} |
| 453 |
|
| 454 |
i->content_encoding = charset ? xstrdup (charset) : NULL; |
| 455 |
} |