chiark / gitweb /
Split base64 code into a separate .c so that the server doesn't pull
[disorder] / lib / mime.c
CommitLineData
460b9539 1/*
2 * This file is part of DisOrder
39d4aa6b 3 * Copyright (C) 2005, 2007 Richard Kettlewell
460b9539 4 *
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.
9 *
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.
14 *
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
18 * USA
19 */
39d4aa6b
RK
20/** @file lib/mime.c
21 * @brief Support for MIME and allied protocols
22 */
460b9539 23
24#include <config.h>
25#include "types.h"
26
27#include <string.h>
28#include <ctype.h>
29
22896b25
RK
30#include <stdio.h>
31
460b9539 32#include "mem.h"
33#include "mime.h"
34#include "vector.h"
35#include "hex.h"
39d4aa6b 36#include "log.h"
fce810c2 37#include "base64.h"
460b9539 38
39d4aa6b 39/** @brief Match whitespace characters */
460b9539 40static int whitespace(int c) {
41 switch(c) {
42 case ' ':
43 case '\t':
44 case '\r':
45 case '\n':
46 return 1;
47 default:
48 return 0;
49 }
50}
51
39d4aa6b 52/** @brief Match RFC2045 tspecial characters */
460b9539 53static int tspecial(int c) {
54 switch(c) {
55 case '(':
56 case ')':
57 case '<':
58 case '>':
59 case '@':
60 case ',':
61 case ';':
62 case ':':
63 case '\\':
64 case '"':
65 case '/':
66 case '[':
67 case ']':
68 case '?':
69 case '=':
70 return 1;
71 default:
72 return 0;
73 }
74}
75
5818980a 76/** @brief Match RFC2616 seprator characters */
39d4aa6b
RK
77static int http_separator(int c) {
78 switch(c) {
79 case '(':
80 case ')':
81 case '<':
82 case '>':
83 case '@':
84 case ',':
85 case ';':
86 case ':':
87 case '\\':
88 case '"':
89 case '/':
90 case '[':
91 case ']':
92 case '?':
93 case '=':
94 case '{':
95 case '}':
96 case ' ':
97 case '\t':
98 return 1;
99 default:
100 return 0;
101 }
102}
103
104/** @brief Match CRLF */
105static int iscrlf(const char *ptr) {
106 return ptr[0] == '\r' && ptr[1] == '\n';
107}
108
109/** @brief Skip whitespace
110 * @param rfc822_comments If true, skip RFC822 nested comments
111 */
112static const char *skipwhite(const char *s, int rfc822_comments) {
460b9539 113 int c, depth;
114
115 for(;;) {
116 switch(c = *s) {
117 case ' ':
118 case '\t':
119 case '\r':
120 case '\n':
121 ++s;
122 break;
123 case '(':
39d4aa6b
RK
124 if(!rfc822_comments)
125 return s;
460b9539 126 ++s;
127 depth = 1;
128 while(*s && depth) {
129 c = *s++;
130 switch(c) {
131 case '(': ++depth; break;
132 case ')': --depth; break;
133 case '\\':
134 if(!*s) return 0;
135 ++s;
136 break;
137 }
138 }
139 if(depth) return 0;
140 break;
141 default:
142 return s;
143 }
144 }
145}
146
39d4aa6b
RK
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
151 */
152static int iswordchar(int c, int (*special)(int)) {
153 return !(c <= ' ' || c > '~' || special(c));
154}
155
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
161 *
162 * A word is a token or a quoted-string.
163 */
164static const char *parseword(const char *s, char **valuep,
165 int (*special)(int)) {
166 struct dynstr value[1];
460b9539 167 int c;
168
39d4aa6b
RK
169 dynstr_init(value);
170 if(*s == '"') {
171 ++s;
172 while((c = *s++) != '"') {
173 switch(c) {
174 case '\\':
175 if(!(c = *s++)) return 0;
176 default:
177 dynstr_append(value, c);
178 break;
179 }
460b9539 180 }
39d4aa6b
RK
181 if(!c) return 0;
182 } else {
183 if(!iswordchar((unsigned char)*s, special))
184 return NULL;
185 dynstr_init(value);
186 while(iswordchar((unsigned char)*s, special))
187 dynstr_append(value, *s++);
460b9539 188 }
39d4aa6b
RK
189 dynstr_terminate(value);
190 *valuep = value->vec;
460b9539 191 return s;
192}
193
39d4aa6b
RK
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
199 */
200static const char *parsetoken(const char *s, char **valuep,
201 int (*special)(int)) {
202 if(*s == '"') return 0;
203 return parseword(s, valuep, special);
204}
205
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
212 */
460b9539 213int mime_content_type(const char *s,
214 char **typep,
215 char **parameternamep,
216 char **parametervaluep) {
39d4aa6b 217 struct dynstr type, parametername;
460b9539 218
219 dynstr_init(&type);
39d4aa6b 220 if(!(s = skipwhite(s, 1))) return -1;
460b9539 221 if(!*s) return -1;
222 while(*s && !tspecial(*s) && !whitespace(*s))
223 dynstr_append(&type, tolower((unsigned char)*s++));
39d4aa6b 224 if(!(s = skipwhite(s, 1))) return -1;
460b9539 225 if(*s++ != '/') return -1;
226 dynstr_append(&type, '/');
39d4aa6b 227 if(!(s = skipwhite(s, 1))) return -1;
460b9539 228 while(*s && !tspecial(*s) && !whitespace(*s))
229 dynstr_append(&type, tolower((unsigned char)*s++));
39d4aa6b 230 if(!(s = skipwhite(s, 1))) return -1;
460b9539 231
232 if(*s == ';') {
233 dynstr_init(&parametername);
234 ++s;
39d4aa6b 235 if(!(s = skipwhite(s, 1))) return -1;
460b9539 236 if(!*s) return -1;
237 while(*s && !tspecial(*s) && !whitespace(*s))
238 dynstr_append(&parametername, tolower((unsigned char)*s++));
39d4aa6b 239 if(!(s = skipwhite(s, 1))) return -1;
460b9539 240 if(*s++ != '=') return -1;
39d4aa6b
RK
241 if(!(s = skipwhite(s, 1))) return -1;
242 if(!(s = parseword(s, parametervaluep, tspecial))) return -1;
243 if(!(s = skipwhite(s, 1))) return -1;
460b9539 244 dynstr_terminate(&parametername);
245 *parameternamep = parametername.vec;
246 } else
247 *parametervaluep = *parameternamep = 0;
248 dynstr_terminate(&type);
249 *typep = type.vec;
250 return 0;
251}
252
39d4aa6b
RK
253/** @brief Parse a MIME message
254 * @param s Start of message
255 * @param callback Called for each header field
256 * @param u Passed to callback
257 * @return Pointer to decoded body (might be in original string)
258 */
460b9539 259const char *mime_parse(const char *s,
260 int (*callback)(const char *name, const char *value,
261 void *u),
262 void *u) {
263 struct dynstr name, value;
264 char *cte = 0, *p;
265
266 while(*s && !iscrlf(s)) {
267 dynstr_init(&name);
268 dynstr_init(&value);
269 while(*s && !tspecial(*s) && !whitespace(*s))
270 dynstr_append(&name, tolower((unsigned char)*s++));
39d4aa6b 271 if(!(s = skipwhite(s, 1))) return 0;
460b9539 272 if(*s != ':') return 0;
273 ++s;
274 while(*s && !(*s == '\n' && !(s[1] == ' ' || s[1] == '\t')))
275 dynstr_append(&value, *s++);
276 if(*s) ++s;
277 dynstr_terminate(&name);
278 dynstr_terminate(&value);
279 if(!strcmp(name.vec, "content-transfer-encoding")) {
280 cte = xstrdup(value.vec);
281 for(p = cte; *p; p++)
282 *p = tolower((unsigned char)*p);
283 }
284 if(callback(name.vec, value.vec, u)) return 0;
285 }
286 if(*s) s += 2;
287 if(cte) {
8a7ccdfe 288 if(!strcmp(cte, "base64")) return mime_base64(s, 0);
460b9539 289 if(!strcmp(cte, "quoted-printable")) return mime_qp(s);
290 }
291 return s;
292}
293
294static int isboundary(const char *ptr, const char *boundary, size_t bl) {
295 return (ptr[0] == '-'
296 && ptr[1] == '-'
297 && !strncmp(ptr + 2, boundary, bl)
298 && (iscrlf(ptr + bl + 2)
299 || (ptr[bl + 2] == '-'
300 && ptr[bl + 3] == '-'
22896b25 301 && (iscrlf(ptr + bl + 4) || *(ptr + bl + 4) == 0))));
460b9539 302}
303
304static int isfinal(const char *ptr, const char *boundary, size_t bl) {
305 return (ptr[0] == '-'
306 && ptr[1] == '-'
307 && !strncmp(ptr + 2, boundary, bl)
308 && ptr[bl + 2] == '-'
309 && ptr[bl + 3] == '-'
22896b25 310 && (iscrlf(ptr + bl + 4) || *(ptr + bl + 4) == 0));
460b9539 311}
312
39d4aa6b
RK
313/** @brief Parse a multipart MIME body
314 * @param s Start of message
315 * @param callback CAllback for each part
316 * @param boundary Boundary string
317 * @param u Passed to callback
318 * @return 0 on success, non-0 on error
319 */
460b9539 320int mime_multipart(const char *s,
321 int (*callback)(const char *s, void *u),
322 const char *boundary,
323 void *u) {
324 size_t bl = strlen(boundary);
325 const char *start, *e;
326 int ret;
327
22896b25
RK
328 /* We must start with a boundary string */
329 if(!isboundary(s, boundary, bl))
330 return -1;
331 /* Keep going until we hit a final boundary */
460b9539 332 while(!isfinal(s, boundary, bl)) {
333 s = strstr(s, "\r\n") + 2;
334 start = s;
335 while(!isboundary(s, boundary, bl)) {
22896b25
RK
336 if(!(e = strstr(s, "\r\n")))
337 return -1;
460b9539 338 s = e + 2;
339 }
340 if((ret = callback(xstrndup(start,
341 s == start ? 0 : s - start - 2),
342 u)))
343 return ret;
344 }
345 return 0;
346}
347
39d4aa6b
RK
348/** @brief Parse an RFC2388-style content-disposition field
349 * @param s Start of field
350 * @param typep Where to store type
351 * @param parameternamep Where to store parameter name
352 * @param parameternvaluep Wher to store parameter value
353 * @return 0 on success, non-0 on error
354 */
460b9539 355int mime_rfc2388_content_disposition(const char *s,
356 char **dispositionp,
357 char **parameternamep,
358 char **parametervaluep) {
39d4aa6b 359 struct dynstr disposition, parametername;
460b9539 360
361 dynstr_init(&disposition);
39d4aa6b 362 if(!(s = skipwhite(s, 1))) return -1;
460b9539 363 if(!*s) return -1;
364 while(*s && !tspecial(*s) && !whitespace(*s))
365 dynstr_append(&disposition, tolower((unsigned char)*s++));
39d4aa6b 366 if(!(s = skipwhite(s, 1))) return -1;
460b9539 367
368 if(*s == ';') {
369 dynstr_init(&parametername);
370 ++s;
39d4aa6b 371 if(!(s = skipwhite(s, 1))) return -1;
460b9539 372 if(!*s) return -1;
373 while(*s && !tspecial(*s) && !whitespace(*s))
374 dynstr_append(&parametername, tolower((unsigned char)*s++));
39d4aa6b 375 if(!(s = skipwhite(s, 1))) return -1;
460b9539 376 if(*s++ != '=') return -1;
39d4aa6b
RK
377 if(!(s = skipwhite(s, 1))) return -1;
378 if(!(s = parseword(s, parametervaluep, tspecial))) return -1;
379 if(!(s = skipwhite(s, 1))) return -1;
460b9539 380 dynstr_terminate(&parametername);
381 *parameternamep = parametername.vec;
382 } else
383 *parametervaluep = *parameternamep = 0;
384 dynstr_terminate(&disposition);
385 *dispositionp = disposition.vec;
386 return 0;
387}
388
39d4aa6b
RK
389/** @brief Convert MIME quoted-printable
390 * @param s Quoted-printable data
391 * @return Decoded data
392 */
460b9539 393char *mime_qp(const char *s) {
394 struct dynstr d;
395 int c, a, b;
396 const char *t;
397
398 dynstr_init(&d);
399 while((c = *s++)) {
400 switch(c) {
401 case '=':
402 if((a = unhexdigitq(s[0])) != -1
403 && (b = unhexdigitq(s[1])) != -1) {
404 dynstr_append(&d, a * 16 + b);
405 s += 2;
406 } else {
407 t = s;
408 while(*t == ' ' || *t == '\t') ++t;
409 if(iscrlf(t)) {
410 /* soft line break */
411 s = t + 2;
412 } else
413 return 0;
414 }
415 break;
416 case ' ':
417 case '\t':
418 t = s;
419 while(*t == ' ' || *t == '\t') ++t;
420 if(iscrlf(t))
421 /* trailing space is always eliminated */
422 s = t;
423 else
424 dynstr_append(&d, c);
425 break;
426 default:
427 dynstr_append(&d, c);
428 break;
429 }
430 }
431 dynstr_terminate(&d);
432 return d.vec;
433}
434
39d4aa6b
RK
435/** @brief Parse a RFC2109 Cookie: header
436 * @param s Header field value
437 * @param cd Where to store result
438 * @return 0 on success, non-0 on error
439 */
440int parse_cookie(const char *s,
441 struct cookiedata *cd) {
442 char *n = 0, *v = 0;
443
444 memset(cd, 0, sizeof *cd);
445 s = skipwhite(s, 0);
446 while(*s) {
447 /* Skip separators */
448 if(*s == ';' || *s == ',') {
449 ++s;
450 s = skipwhite(s, 0);
451 continue;
452 }
453 if(!(s = parsetoken(s, &n, http_separator))) return -1;
454 s = skipwhite(s, 0);
455 if(*s++ != '=') return -1;
456 s = skipwhite(s, 0);
457 if(!(s = parseword(s, &v, http_separator))) return -1;
458 if(n[0] == '$') {
459 /* Some bit of meta-information */
460 if(!strcmp(n, "$Version"))
461 cd->version = v;
462 else if(!strcmp(n, "$Path")) {
463 if(cd->ncookies > 0 && cd->cookies[cd->ncookies-1].path == 0)
464 cd->cookies[cd->ncookies-1].path = v;
465 else {
466 error(0, "redundant $Path in Cookie: header");
467 return -1;
468 }
469 } else if(!strcmp(n, "$Domain")) {
470 if(cd->ncookies > 0 && cd->cookies[cd->ncookies-1].domain == 0)
471 cd->cookies[cd->ncookies-1].domain = v;
472 else {
473 error(0, "redundant $Domain in Cookie: header");
474 return -1;
475 }
476 }
477 } else {
478 /* It's a new cookie */
479 cd->cookies = xrealloc(cd->cookies,
480 (cd->ncookies + 1) * sizeof (struct cookie));
481 cd->cookies[cd->ncookies].name = n;
482 cd->cookies[cd->ncookies].value = v;
483 cd->cookies[cd->ncookies].path = 0;
484 cd->cookies[cd->ncookies].domain = 0;
485 ++cd->ncookies;
486 }
487 s = skipwhite(s, 0);
488 if(*s && (*s != ',' && *s != ';')) {
489 error(0, "missing separator in Cookie: header");
490 return -1;
491 }
492 }
493 return 0;
494}
495
496/** @brief Find a named cookie
497 * @param cd Parse cookie data
498 * @param name Name of cookie
499 * @return Cookie structure or NULL if not found
500 */
501const struct cookie *find_cookie(const struct cookiedata *cd,
502 const char *name) {
503 int n;
504
505 for(n = 0; n < cd->ncookies; ++n)
506 if(!strcmp(cd->cookies[n].name, name))
507 return &cd->cookies[n];
508 return 0;
509}
510
460b9539 511/*
512Local Variables:
513c-basic-offset:2
514comment-column:40
515fill-column:79
516End:
517*/