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 int mime_content_type(const char *s,
215 char **parameternamep,
216 char **parametervaluep) {
217 struct dynstr type, parametername;
220 if(!(s = skipwhite(s, 1))) return -1;
222 while(*s && !tspecial(*s) && !whitespace(*s))
223 dynstr_append(&type, tolower((unsigned char)*s++));
224 if(!(s = skipwhite(s, 1))) return -1;
225 if(*s++ != '/') return -1;
226 dynstr_append(&type, '/');
227 if(!(s = skipwhite(s, 1))) return -1;
228 while(*s && !tspecial(*s) && !whitespace(*s))
229 dynstr_append(&type, tolower((unsigned char)*s++));
230 if(!(s = skipwhite(s, 1))) return -1;
233 dynstr_init(¶metername);
235 if(!(s = skipwhite(s, 1))) return -1;
237 while(*s && !tspecial(*s) && !whitespace(*s))
238 dynstr_append(¶metername, tolower((unsigned char)*s++));
239 if(!(s = skipwhite(s, 1))) return -1;
240 if(*s++ != '=') return -1;
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;
244 dynstr_terminate(¶metername);
245 *parameternamep = parametername.vec;
247 *parametervaluep = *parameternamep = 0;
248 dynstr_terminate(&type);
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)
259 const char *mime_parse(const char *s,
260 int (*callback)(const char *name, const char *value,
263 struct dynstr name, value;
266 while(*s && !iscrlf(s)) {
269 while(*s && !tspecial(*s) && !whitespace(*s))
270 dynstr_append(&name, tolower((unsigned char)*s++));
271 if(!(s = skipwhite(s, 1))) return 0;
272 if(*s != ':') return 0;
274 while(*s && !(*s == '\n' && !(s[1] == ' ' || s[1] == '\t')))
275 dynstr_append(&value, *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);
284 if(callback(name.vec, value.vec, u)) return 0;
288 if(!strcmp(cte, "base64")) return mime_base64(s, 0);
289 if(!strcmp(cte, "quoted-printable")) return mime_qp(s);
294 static int isboundary(const char *ptr, const char *boundary, size_t bl) {
295 return (ptr[0] == '-'
297 && !strncmp(ptr + 2, boundary, bl)
298 && (iscrlf(ptr + bl + 2)
299 || (ptr[bl + 2] == '-'
300 && ptr[bl + 3] == '-'
301 && (iscrlf(ptr + bl + 4) || *(ptr + bl + 4) == 0))));
304 static int isfinal(const char *ptr, const char *boundary, size_t bl) {
305 return (ptr[0] == '-'
307 && !strncmp(ptr + 2, boundary, bl)
308 && ptr[bl + 2] == '-'
309 && ptr[bl + 3] == '-'
310 && (iscrlf(ptr + bl + 4) || *(ptr + bl + 4) == 0));
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
320 int mime_multipart(const char *s,
321 int (*callback)(const char *s, void *u),
322 const char *boundary,
324 size_t bl = strlen(boundary);
325 const char *start, *e;
328 /* We must start with a boundary string */
329 if(!isboundary(s, boundary, bl))
331 /* Keep going until we hit a final boundary */
332 while(!isfinal(s, boundary, bl)) {
333 s = strstr(s, "\r\n") + 2;
335 while(!isboundary(s, boundary, bl)) {
336 if(!(e = strstr(s, "\r\n")))
340 if((ret = callback(xstrndup(start,
341 s == start ? 0 : s - start - 2),
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
355 int mime_rfc2388_content_disposition(const char *s,
357 char **parameternamep,
358 char **parametervaluep) {
359 struct dynstr disposition, parametername;
361 dynstr_init(&disposition);
362 if(!(s = skipwhite(s, 1))) return -1;
364 while(*s && !tspecial(*s) && !whitespace(*s))
365 dynstr_append(&disposition, tolower((unsigned char)*s++));
366 if(!(s = skipwhite(s, 1))) return -1;
369 dynstr_init(¶metername);
371 if(!(s = skipwhite(s, 1))) return -1;
373 while(*s && !tspecial(*s) && !whitespace(*s))
374 dynstr_append(¶metername, tolower((unsigned char)*s++));
375 if(!(s = skipwhite(s, 1))) return -1;
376 if(*s++ != '=') return -1;
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;
380 dynstr_terminate(¶metername);
381 *parameternamep = parametername.vec;
383 *parametervaluep = *parameternamep = 0;
384 dynstr_terminate(&disposition);
385 *dispositionp = disposition.vec;
389 /** @brief Convert MIME quoted-printable
390 * @param s Quoted-printable data
391 * @return Decoded data
393 char *mime_qp(const char *s) {
402 if((a = unhexdigitq(s[0])) != -1
403 && (b = unhexdigitq(s[1])) != -1) {
404 dynstr_append(&d, a * 16 + b);
408 while(*t == ' ' || *t == '\t') ++t;
410 /* soft line break */
419 while(*t == ' ' || *t == '\t') ++t;
421 /* trailing space is always eliminated */
424 dynstr_append(&d, c);
427 dynstr_append(&d, c);
431 dynstr_terminate(&d);
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
440 int parse_cookie(const char *s,
441 struct cookiedata *cd) {
444 memset(cd, 0, sizeof *cd);
447 /* Skip separators */
448 if(*s == ';' || *s == ',') {
453 if(!(s = parsetoken(s, &n, http_separator))) return -1;
455 if(*s++ != '=') return -1;
457 if(!(s = parseword(s, &v, http_separator))) return -1;
459 /* Some bit of meta-information */
460 if(!strcmp(n, "$Version"))
462 else if(!strcmp(n, "$Path")) {
463 if(cd->ncookies > 0 && cd->cookies[cd->ncookies-1].path == 0)
464 cd->cookies[cd->ncookies-1].path = v;
466 error(0, "redundant $Path in Cookie: header");
469 } else if(!strcmp(n, "$Domain")) {
470 if(cd->ncookies > 0 && cd->cookies[cd->ncookies-1].domain == 0)
471 cd->cookies[cd->ncookies-1].domain = v;
473 error(0, "redundant $Domain in Cookie: header");
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;
488 if(*s && (*s != ',' && *s != ';')) {
489 error(0, "missing separator in Cookie: header");
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
501 const struct cookie *find_cookie(const struct cookiedata *cd,
505 for(n = 0; n < cd->ncookies; ++n)
506 if(!strcmp(cd->cookies[n].name, name))
507 return &cd->cookies[n];