2 * This file is part of DisOrder
3 * Copyright (C) 2005, 2007 Richard Kettlewell
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 2 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful, but
11 * WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * 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, write to the Free Software
17 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
21 * @brief Support for MIME and allied protocols
39 /** @brief Match whitespace characters */
40 static int whitespace(int c
) {
52 /** @brief Match RFC2045 tspecial characters */
53 static int tspecial(int c
) {
76 /** @brief Match RFC2616 seprator characters */
77 static int http_separator(int c
) {
104 /** @brief Match CRLF */
105 static int iscrlf(const char *ptr
) {
106 return ptr
[0] == '\r' && ptr
[1] == '\n';
109 /** @brief Skip whitespace
110 * @param rfc822_comments If true, skip RFC822 nested comments
112 static const char *skipwhite(const char *s
, int rfc822_comments
) {
131 case '(': ++depth
; break;
132 case ')': --depth
; break;
147 /** @brief Test for a word character
148 * @param c Character to test
149 * @param special tspecial() (MIME/RFC2405) or http_separator() (HTTP/RFC2616)
150 * @return 1 if @p c is a word character, else 0
152 static int iswordchar(int c
, int (*special
)(int)) {
153 return !(c
<= ' ' || c
> '~' || special(c
));
156 /** @brief Parse an RFC1521/RFC2616 word
157 * @param s Pointer to start of word
158 * @param valuep Where to store value
159 * @param special tspecial() (MIME/RFC2405) or http_separator() (HTTP/RFC2616)
160 * @return Pointer just after end of word or NULL if there's no word
162 * A word is a token or a quoted-string.
164 static const char *parseword(const char *s
, char **valuep
,
165 int (*special
)(int)) {
166 struct dynstr value
[1];
172 while((c
= *s
++) != '"') {
175 if(!(c
= *s
++)) return 0;
177 dynstr_append(value
, c
);
183 if(!iswordchar((unsigned char)*s
, special
))
186 while(iswordchar((unsigned char)*s
, special
))
187 dynstr_append(value
, *s
++);
189 dynstr_terminate(value
);
190 *valuep
= value
->vec
;
194 /** @brief Parse an RFC1521/RFC2616 token
195 * @param s Pointer to start of token
196 * @param valuep Where to store value
197 * @param special tspecial() (MIME/RFC2405) or http_separator() (HTTP/RFC2616)
198 * @return Pointer just after end of token or NULL if there's no token
200 static const char *parsetoken(const char *s
, char **valuep
,
201 int (*special
)(int)) {
202 if(*s
== '"') return 0;
203 return parseword(s
, valuep
, special
);
206 /** @brief Parse a MIME content-type field
207 * @param s Start of field
208 * @param typep Where to store type
209 * @param parameternamep Where to store parameter name
210 * @param parameternvaluep Wher to store parameter value
211 * @return 0 on success, non-0 on error
213 * See <a href="http://tools.ietf.org/html/rfc2045#section-5">RFC 2045 s5</a>.
215 int mime_content_type(const char *s
,
217 char **parameternamep
,
218 char **parametervaluep
) {
219 struct dynstr type
, parametername
;
222 if(!(s
= skipwhite(s
, 1))) return -1;
224 while(*s
&& !tspecial(*s
) && !whitespace(*s
))
225 dynstr_append(&type
, tolower((unsigned char)*s
++));
226 if(!(s
= skipwhite(s
, 1))) return -1;
227 if(*s
++ != '/') return -1;
228 dynstr_append(&type
, '/');
229 if(!(s
= skipwhite(s
, 1))) return -1;
230 while(*s
&& !tspecial(*s
) && !whitespace(*s
))
231 dynstr_append(&type
, tolower((unsigned char)*s
++));
232 if(!(s
= skipwhite(s
, 1))) return -1;
235 dynstr_init(¶metername
);
237 if(!(s
= skipwhite(s
, 1))) return -1;
239 while(*s
&& !tspecial(*s
) && !whitespace(*s
))
240 dynstr_append(¶metername
, tolower((unsigned char)*s
++));
241 if(!(s
= skipwhite(s
, 1))) return -1;
242 if(*s
++ != '=') return -1;
243 if(!(s
= skipwhite(s
, 1))) return -1;
244 if(!(s
= parseword(s
, parametervaluep
, tspecial
))) return -1;
245 if(!(s
= skipwhite(s
, 1))) return -1;
246 dynstr_terminate(¶metername
);
247 *parameternamep
= parametername
.vec
;
249 *parametervaluep
= *parameternamep
= 0;
250 dynstr_terminate(&type
);
255 /** @brief Parse a MIME message
256 * @param s Start of message
257 * @param callback Called for each header field
258 * @param u Passed to callback
259 * @return Pointer to decoded body (might be in original string), or NULL on error
261 * This does an RFC 822-style parse and honors Content-Transfer-Encoding as
262 * described in <a href="http://tools.ietf.org/html/rfc2045#section-6">RFC 2045
263 * s6</a>. @p callback is called for each header field encountered, in order,
264 * with ASCII characters in the header name forced to lower case.
266 const char *mime_parse(const char *s
,
267 int (*callback
)(const char *name
, const char *value
,
270 struct dynstr name
, value
;
273 while(*s
&& !iscrlf(s
)) {
276 while(*s
&& !tspecial(*s
) && !whitespace(*s
))
277 dynstr_append(&name
, tolower((unsigned char)*s
++));
278 if(!(s
= skipwhite(s
, 1))) return 0;
279 if(*s
!= ':') return 0;
281 while(*s
&& !(*s
== '\n' && !(s
[1] == ' ' || s
[1] == '\t')))
282 dynstr_append(&value
, *s
++);
284 dynstr_terminate(&name
);
285 dynstr_terminate(&value
);
286 if(!strcmp(name
.vec
, "content-transfer-encoding")) {
287 cte
= xstrdup(value
.vec
);
288 for(p
= cte
; *p
; p
++)
289 *p
= tolower((unsigned char)*p
);
291 if(callback(name
.vec
, value
.vec
, u
)) return 0;
295 if(!strcmp(cte
, "base64")) return mime_base64(s
, 0);
296 if(!strcmp(cte
, "quoted-printable")) return mime_qp(s
);
301 /** @brief Match the boundary string */
302 static int isboundary(const char *ptr
, const char *boundary
, size_t bl
) {
303 return (ptr
[0] == '-'
305 && !strncmp(ptr
+ 2, boundary
, bl
)
306 && (iscrlf(ptr
+ bl
+ 2)
307 || (ptr
[bl
+ 2] == '-'
308 && ptr
[bl
+ 3] == '-'
309 && (iscrlf(ptr
+ bl
+ 4) || *(ptr
+ bl
+ 4) == 0))));
312 /** @brief Match the final boundary string */
313 static int isfinal(const char *ptr
, const char *boundary
, size_t bl
) {
314 return (ptr
[0] == '-'
316 && !strncmp(ptr
+ 2, boundary
, bl
)
317 && ptr
[bl
+ 2] == '-'
318 && ptr
[bl
+ 3] == '-'
319 && (iscrlf(ptr
+ bl
+ 4) || *(ptr
+ bl
+ 4) == 0));
322 /** @brief Parse a multipart MIME body
323 * @param s Start of message
324 * @param callback Callback for each part
325 * @param boundary Boundary string
326 * @param u Passed to callback
327 * @return 0 on success, non-0 on error
329 * See <a href="http://tools.ietf.org/html/rfc2046#section-5.1">RFC 2046
330 * s5.1</a>. @p callback is called for each part (not yet decoded in any way)
331 * in succession; you should probably call mime_parse() for each part.
333 int mime_multipart(const char *s
,
334 int (*callback
)(const char *s
, void *u
),
335 const char *boundary
,
337 size_t bl
= strlen(boundary
);
338 const char *start
, *e
;
341 /* We must start with a boundary string */
342 if(!isboundary(s
, boundary
, bl
))
344 /* Keep going until we hit a final boundary */
345 while(!isfinal(s
, boundary
, bl
)) {
346 s
= strstr(s
, "\r\n") + 2;
348 while(!isboundary(s
, boundary
, bl
)) {
349 if(!(e
= strstr(s
, "\r\n")))
353 if((ret
= callback(xstrndup(start
,
354 s
== start ?
0 : s
- start
- 2),
361 /** @brief Parse an RFC2388-style content-disposition field
362 * @param s Start of field
363 * @param typep Where to store type
364 * @param parameternamep Where to store parameter name
365 * @param parameternvaluep Wher to store parameter value
366 * @return 0 on success, non-0 on error
368 * See <a href="http://tools.ietf.org/html/rfc2388#section-3">RFC 2388 s3</a>
369 * and <a href="http://tools.ietf.org/html/rfc2183">RFC 2183</a>.
371 int mime_rfc2388_content_disposition(const char *s
,
373 char **parameternamep
,
374 char **parametervaluep
) {
375 struct dynstr disposition
, parametername
;
377 dynstr_init(&disposition
);
378 if(!(s
= skipwhite(s
, 1))) return -1;
380 while(*s
&& !tspecial(*s
) && !whitespace(*s
))
381 dynstr_append(&disposition
, tolower((unsigned char)*s
++));
382 if(!(s
= skipwhite(s
, 1))) return -1;
385 dynstr_init(¶metername
);
387 if(!(s
= skipwhite(s
, 1))) return -1;
389 while(*s
&& !tspecial(*s
) && !whitespace(*s
))
390 dynstr_append(¶metername
, tolower((unsigned char)*s
++));
391 if(!(s
= skipwhite(s
, 1))) return -1;
392 if(*s
++ != '=') return -1;
393 if(!(s
= skipwhite(s
, 1))) return -1;
394 if(!(s
= parseword(s
, parametervaluep
, tspecial
))) return -1;
395 if(!(s
= skipwhite(s
, 1))) return -1;
396 dynstr_terminate(¶metername
);
397 *parameternamep
= parametername
.vec
;
399 *parametervaluep
= *parameternamep
= 0;
400 dynstr_terminate(&disposition
);
401 *dispositionp
= disposition
.vec
;
405 /** @brief Convert MIME quoted-printable
406 * @param s Quoted-printable data
407 * @return Decoded data
409 * See <a href="http://tools.ietf.org/html/rfc2045#section-6.7">RFC 2045
412 char *mime_qp(const char *s
) {
421 if((a
= unhexdigitq(s
[0])) != -1
422 && (b
= unhexdigitq(s
[1])) != -1) {
423 dynstr_append(&d
, a
* 16 + b
);
427 while(*t
== ' ' || *t
== '\t') ++t
;
429 /* soft line break */
438 while(*t
== ' ' || *t
== '\t') ++t
;
440 /* trailing space is always eliminated */
443 dynstr_append(&d
, c
);
446 dynstr_append(&d
, c
);
450 dynstr_terminate(&d
);
454 /** @brief Parse a RFC2109 Cookie: header
455 * @param s Header field value
456 * @param cd Where to store result
457 * @return 0 on success, non-0 on error
459 int parse_cookie(const char *s
,
460 struct cookiedata
*cd
) {
463 memset(cd
, 0, sizeof *cd
);
466 /* Skip separators */
467 if(*s
== ';' || *s
== ',') {
472 if(!(s
= parsetoken(s
, &n
, http_separator
))) return -1;
474 if(*s
++ != '=') return -1;
476 if(!(s
= parseword(s
, &v
, http_separator
))) return -1;
478 /* Some bit of meta-information */
479 if(!strcmp(n
, "$Version"))
481 else if(!strcmp(n
, "$Path")) {
482 if(cd
->ncookies
> 0 && cd
->cookies
[cd
->ncookies
-1].path
== 0)
483 cd
->cookies
[cd
->ncookies
-1].path
= v
;
485 error(0, "redundant $Path in Cookie: header");
488 } else if(!strcmp(n
, "$Domain")) {
489 if(cd
->ncookies
> 0 && cd
->cookies
[cd
->ncookies
-1].domain
== 0)
490 cd
->cookies
[cd
->ncookies
-1].domain
= v
;
492 error(0, "redundant $Domain in Cookie: header");
497 /* It's a new cookie */
498 cd
->cookies
= xrealloc(cd
->cookies
,
499 (cd
->ncookies
+ 1) * sizeof (struct cookie
));
500 cd
->cookies
[cd
->ncookies
].name
= n
;
501 cd
->cookies
[cd
->ncookies
].value
= v
;
502 cd
->cookies
[cd
->ncookies
].path
= 0;
503 cd
->cookies
[cd
->ncookies
].domain
= 0;
507 if(*s
&& (*s
!= ',' && *s
!= ';')) {
508 error(0, "missing separator in Cookie: header");
515 /** @brief Find a named cookie
516 * @param cd Parse cookie data
517 * @param name Name of cookie
518 * @return Cookie structure or NULL if not found
520 const struct cookie
*find_cookie(const struct cookiedata
*cd
,
524 for(n
= 0; n
< cd
->ncookies
; ++n
)
525 if(!strcmp(cd
->cookies
[n
].name
, name
))
526 return &cd
->cookies
[n
];