1: /*************************************************************************
2: * (C) 2012 AITNET ltd - Sofia/Bulgaria - <misho@aitnet.org>
3: * by Michael Pounov <misho@elwix.org>
4: *
5: * $Author: misho $
6: * $Id: mime.c,v 1.3.4.1 2012/07/31 11:56:16 misho Exp $
7: *
8: **************************************************************************
9: The ELWIX and AITNET software is distributed under the following
10: terms:
11:
12: All of the documentation and software included in the ELWIX and AITNET
13: Releases is copyrighted by ELWIX - Sofia/Bulgaria <info@elwix.org>
14:
15: Copyright 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012
16: by Michael Pounov <misho@elwix.org>. All rights reserved.
17:
18: Redistribution and use in source and binary forms, with or without
19: modification, are permitted provided that the following conditions
20: are met:
21: 1. Redistributions of source code must retain the above copyright
22: notice, this list of conditions and the following disclaimer.
23: 2. Redistributions in binary form must reproduce the above copyright
24: notice, this list of conditions and the following disclaimer in the
25: documentation and/or other materials provided with the distribution.
26: 3. All advertising materials mentioning features or use of this software
27: must display the following acknowledgement:
28: This product includes software developed by Michael Pounov <misho@elwix.org>
29: ELWIX - Embedded LightWeight unIX and its contributors.
30: 4. Neither the name of AITNET nor the names of its contributors
31: may be used to endorse or promote products derived from this software
32: without specific prior written permission.
33:
34: THIS SOFTWARE IS PROVIDED BY AITNET AND CONTRIBUTORS ``AS IS'' AND
35: ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
36: IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
37: ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
38: FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
39: DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
40: OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
41: HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
42: LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
43: OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
44: SUCH DAMAGE.
45: */
46: #include "global.h"
47: #include "mime.h"
48:
49:
50: static int decode_quoted(char *, int, char *);
51: static int decode_base64(char *, int, char *);
52:
53: static const char *n_encode[] = { "7bit", "8bit", "binary" };
54: static struct _tagEncode {
55: char *name;
56: float mul;
57:
58: int (*decode)(char *, int, char *);
59: } encode[] = {
60: { "quoted-printable", 1, decode_quoted },
61: { "base64", (float) 3 / 4, decode_base64 }
62: };
63:
64:
65: static inline char *
66: bd_begin(const char *str)
67: {
68: char *s;
69: int len = strlen(str) + 6;
70:
71: s = malloc(len + 1);
72: if (!s) {
73: LOGERR;
74: return NULL;
75: } else {
76: snprintf(s, len + 1, "\r\n--%s\r\n", str);
77: s[len] = 0;
78: }
79:
80: return s;
81: }
82:
83: static inline char *
84: bd_end(const char *str)
85: {
86: char *s;
87: int len = strlen(str) + 8;
88:
89: s = malloc(len + 1);
90: if (!s) {
91: LOGERR;
92: return NULL;
93: } else {
94: snprintf(s, len + 1, "\r\n--%s--\r\n", str);
95: s[len] = 0;
96: }
97:
98: return s;
99: }
100:
101: static u_int
102: powmod(int x, int y, int q)
103: {
104: u_int ret = 1;
105:
106: while (y) {
107: if (y & 1)
108: ret = ((unsigned long long)ret * x) % q;
109: x = (unsigned long long) x * x % q;
110: y = y / 2;
111: }
112: return ret;
113: }
114:
115: static const char *
116: findtextpos(const char *T, size_t tlen, const char *P, size_t plen)
117: {
118: const u_int q = 4294967291u;
119: const u_int d = 256;
120: u_int hash, p = 0, t = 0;
121: register int i;
122:
123: hash = powmod(d, plen - 1, q);
124:
125: /* calculate initial hash tags */
126: for (i = 0; i < plen; i++) {
127: p = (d * p + P[i]) % q;
128: t = (d * t + T[i]) % q;
129: }
130:
131: tlen -= plen;
132: for (i = 0; i <= tlen; i++) {
133: if (p == t) {
134: /* match pattern */
135: if (!memcmp(P, T + i, plen))
136: return T + i;
137: }
138:
139: /* rehashing */
140: if (i < tlen)
141: t = (d * (t - T[i] * hash) + T[i + plen]) % q;
142: }
143:
144: return NULL;
145: }
146:
147: static inline void
148: freeHeader(struct tagMIME * __restrict m)
149: {
150: struct tagCGI *c;
151:
152: while ((c = SLIST_FIRST(&m->mime_header))) {
153: if (c->cgi_name)
154: free(c->cgi_name);
155: if (c->cgi_value)
156: free(c->cgi_value);
157: SLIST_REMOVE_HEAD(&m->mime_header, cgi_node);
158: free(c);
159: }
160: }
161:
162: static char *
163: hdrValue(const char *str, size_t len, const char **end)
164: {
165: const char *e, *crlf = NULL;
166: char *tmp, *s = NULL;
167: int off = 0;
168:
169: e = str + len;
170: while (str < e) {
171: if (!(crlf = findtextpos(str, e - str, CRLF, strlen(CRLF)))) {
172: www_SetErr(EFAULT, "Bad header format of MIME part");
173: return NULL;
174: }
175:
176: tmp = realloc(s, crlf - str + off + 1);
177: if (!tmp) {
178: LOGERR;
179: free(s);
180: return NULL;
181: } else
182: s = tmp;
183:
184: memcpy(s + off, str, crlf - str);
185: s[crlf - str + off] = 0;
186: off += crlf - str;
187:
188: /* if is multi part header value */
189: tmp = (char*) crlf + strlen(CRLF);
190: if (*tmp == ' ' || *tmp == '\t')
191: str = ++tmp;
192: else
193: break;
194: }
195:
196: *end = crlf + strlen(CRLF);
197: return s;
198: }
199:
200: static inline int
201: hexdigit(char a)
202: {
203: if (a >= '0' && a <= '9')
204: return a - '0';
205: if (a >= 'a' && a <= 'f')
206: return a - 'a' + 10;
207: if (a >= 'A' && a <= 'F')
208: return a - 'A' + 10;
209: /* error!!! */
210: return -1;
211: }
212:
213: static int
214: decode_quoted(char *in, int len, char *out)
215: {
216: register int i, cx;
217:
218: for (i = cx = 0; i < len; i++)
219: if (in[i] == '=') {
220: /* special handling */
221: i++;
222: if ((in[i] >= '0' && in[i] <= '9') ||
223: (in[i] >= 'A' && in[i] <= 'F') ||
224: (in[i] >= 'a' && in[i] <= 'f')) {
225: /* encoding a special char */
226: *out++ = hexdigit(in[i]) << 4 | hexdigit(in[i+ 1]);
227: cx++;
228: } else
229: i += strlen(CRLF);
230: } else {
231: *out++ = in[i++];
232: cx++;
233: }
234:
235: return cx;
236: }
237:
238: static int
239: decode_base64(char *in, int len, char *out)
240: {
241: register int cx, i, j;
242: int bits, eqc;
243:
244: for (cx = i = eqc = bits = 0; i < len && !eqc; bits = 0) {
245: for (j = 0; i < len && j < 4; i++) {
246: switch (in[i]) {
247: case 'A': case 'B': case 'C': case 'D': case 'E':
248: case 'F': case 'G': case 'H': case 'I': case 'J':
249: case 'K': case 'L': case 'M': case 'N': case 'O':
250: case 'P': case 'Q': case 'R': case 'S': case 'T':
251: case 'U': case 'V': case 'W': case 'X': case 'Y':
252: case 'Z':
253: bits = (bits << 6) | (in[i] - 'A');
254: j++;
255: break;
256: case 'a': case 'b': case 'c': case 'd': case 'e':
257: case 'f': case 'g': case 'h': case 'i': case 'j':
258: case 'k': case 'l': case 'm': case 'n': case 'o':
259: case 'p': case 'q': case 'r': case 's': case 't':
260: case 'u': case 'v': case 'w': case 'x': case 'y':
261: case 'z':
262: bits = (bits << 6) | (in[i] - 'a' + 26);
263: j++;
264: break;
265: case '0': case '1': case '2': case '3': case '4':
266: case '5': case '6': case '7': case '8': case '9':
267: bits = (bits << 6) | (in[i] - '0' + 52);
268: j++;
269: break;
270: case '+':
271: bits = (bits << 6) | 62;
272: j++;
273: break;
274: case '/':
275: bits = (bits << 6) | 63;
276: j++;
277: break;
278: case '=':
279: bits <<= 6;
280: j++;
281: eqc++;
282: break;
283: default:
284: break;
285: }
286: }
287:
288: if (!j && i >= len)
289: continue;
290:
291: switch (eqc) {
292: case 0:
293: *out++ = (bits >> 16) & 0xff;
294: *out++ = (bits >> 8) & 0xff;
295: *out++ = bits & 0xff;
296: cx += 3;
297: break;
298: case 1:
299: *out++ = (bits >> 16) & 0xff;
300: *out++ = (bits >> 8) & 0xff;
301: cx += 2;
302: break;
303: case 2:
304: *out++ = (bits >> 16) & 0xff;
305: cx += 1;
306: break;
307: }
308: }
309:
310: return cx;
311: }
312:
313: /* ------------------------------------------------------------------ */
314:
315: /*
316: * mime_parseMultiPart() - Parse multi part MIME message
317: *
318: * @str = String
319: * @len = String length
320: * @bd = Boundary tag
321: * @end = End of parsed part
322: * return: NULL error or !=NULL allocated MIME session
323: */
324: mime_t *
325: mime_parseMultiPart(const char *str, size_t len, const char *bdtag, const char **end)
326: {
327: mime_t *mime = NULL;
328: struct iovec bd[2];
329: struct tagMIME *m, *old = NULL;
330: const char *next = NULL;
331:
332: if (!str | !bdtag) {
333: www_SetErr(EINVAL, "String or boundary tag is NULL");
334: return NULL;
335: }
336:
337: /* init MIME */
338: mime = malloc(sizeof(mime_t));
339: if (!mime) {
340: LOGERR;
341: return NULL;
342: } else {
343: memset(mime, 0, sizeof(mime_t));
344: SLIST_INIT(mime);
345: }
346:
347: /* prepare boundary format */
348: bd[0].iov_base = bd_begin(bdtag);
349: if (!bd[0].iov_base) {
350: free(mime);
351: return NULL;
352: } else
353: bd[0].iov_len = strlen(bd[0].iov_base);
354: bd[1].iov_base = bd_end(bdtag);
355: if (!bd[1].iov_base) {
356: free(bd[0].iov_base);
357: free(mime);
358: return NULL;
359: } else
360: bd[1].iov_len = strlen(bd[1].iov_base);
361:
362: /* check boundary tag */
363: if (memcmp(str, strstr(bd[0].iov_base, "--"), strlen(strstr(bd[0].iov_base, "--")))) {
364: www_SetErr(EFAULT, "Bad content data, not found boundary tag");
365: free(bd[1].iov_base);
366: free(bd[0].iov_base);
367: free(mime);
368: return NULL;
369: } else {
370: str += strlen(strstr(bd[0].iov_base, "--"));
371: len -= strlen(strstr(bd[0].iov_base, "--"));
372: }
373:
374: while (len > 0) {
375: m = malloc(sizeof(struct tagMIME));
376: if (!m) {
377: LOGERR;
378: mime_close(&mime);
379: free(bd[1].iov_base);
380: free(bd[0].iov_base);
381: return NULL;
382: } else {
383: memset(m, 0, sizeof(struct tagMIME));
384: SLIST_INIT(&m->mime_header);
385: }
386:
387: if (!(next = findtextpos(str, len, bd[0].iov_base, bd[0].iov_len)))
388: next = findtextpos(str, len, bd[1].iov_base, bd[1].iov_len);
389:
390: /* parse message between tags */
391: if (mime_readPart(m, str, next - str)) {
392: mime_close(&mime);
393: free(bd[1].iov_base);
394: free(bd[0].iov_base);
395: return NULL;
396: }
397:
398: str += next - str;
399: len -= next - str;
400:
401: /* add to mime session */
402: if (!old)
403: SLIST_INSERT_HEAD(mime, m, mime_node);
404: else
405: SLIST_INSERT_AFTER(old, m, mime_node);
406: old = m;
407:
408: /* match part termination tag */
409: if (!memcmp(str, bd[1].iov_base, bd[1].iov_len))
410: break;
411:
412: str += bd[0].iov_len;
413: len -= bd[0].iov_len;
414: }
415:
416: str += bd[0].iov_len;
417: /* LLVM static code analyzer said for this - unusable
418: *
419: len -= bd[0].iov_len;
420: */
421:
422: free(bd[1].iov_base);
423: free(bd[0].iov_base);
424:
425: if (end)
426: *end = str;
427: return mime;
428: }
429:
430: static inline void
431: freeMIME(struct tagMIME * __restrict m)
432: {
433: if (m->mime_body.iov_base)
434: free(m->mime_body.iov_base);
435: if (m->mime_prolog.iov_base)
436: free(m->mime_prolog.iov_base);
437: if (m->mime_epilog.iov_base)
438: free(m->mime_epilog.iov_base);
439:
440: freeHeader(m);
441: mime_close(&m->mime_attach);
442: }
443:
444: /*
445: * mime_close() - Close MIME session and free all resources
446: *
447: * @mime = Inited mime session
448: * return: none
449: */
450: void
451: mime_close(mime_t ** __restrict mime)
452: {
453: struct tagMIME *m;
454:
455: if (!mime || !*mime)
456: return;
457:
458: while ((m = SLIST_FIRST(*mime))) {
459: SLIST_REMOVE_HEAD(*mime, mime_node);
460: freeMIME(m);
461: free(m);
462: }
463:
464: free(*mime);
465: *mime = NULL;
466: }
467:
468: /*
469: * mime_parseHeader() - Parse MIME header pairs
470: *
471: * @m = Mime part
472: * @str = String
473: * @len = String length
474: * @end = End of parsed part
475: * return: -1 error or 0 ok
476: */
477: int
478: mime_parseHeader(struct tagMIME * __restrict m, const char *str, size_t len, const char **end)
479: {
480: const char *e, *colon, *eoh;
481: struct tagCGI *c, *old = NULL;
482:
483: if (!m || !str) {
484: www_SetErr(EINVAL, "Mime part or string is NULL");
485: return -1;
486: } else
487: e = str + len;
488:
489: while (str < e) {
490: if (!memcmp(str, CRLF, strlen(CRLF))) {
491: str += 2;
492: break;
493: }
494:
495: colon = memchr(str, ':', e - str);
496: eoh = findtextpos(str, e - str, CRLF, strlen(CRLF));
497: if (!colon || !eoh || colon > eoh) {
498: www_SetErr(EFAULT, "Bad MIME format message");
499: freeHeader(m);
500: return -1;
501: }
502:
503: c = malloc(sizeof(struct tagCGI));
504: if (!c) {
505: LOGERR;
506: freeHeader(m);
507: return -1;
508: }
509: /* get name */
510: c->cgi_name = malloc(colon - str + 1);
511: if (!c->cgi_name) {
512: LOGERR;
513: free(c);
514: freeHeader(m);
515: return -1;
516: } else {
517: memcpy(c->cgi_name, str, colon - str);
518: c->cgi_name[colon - str] = 0;
519: }
520: /* get value */
521: c->cgi_value = hdrValue(colon + 1, e - colon - 1, &str);
522: if (!c->cgi_value) {
523: free(c->cgi_name);
524: free(c);
525: freeHeader(m);
526: return -1;
527: }
528:
529: if (!old)
530: SLIST_INSERT_HEAD(&m->mime_header, c, cgi_node);
531: else
532: SLIST_INSERT_AFTER(old, c, cgi_node);
533: old = c;
534: }
535:
536: if (end)
537: *end = str;
538: return 0;
539: }
540:
541: /*
542: * mime_getValue() - Get value from MIME header
543: *
544: * @m = Mime part
545: * @name = Header name
546: * return: NULL not found or !=NULL value
547: */
548: inline const char *
549: mime_getValue(struct tagMIME * __restrict m, const char *name)
550: {
551: struct tagCGI *c;
552: const char *v = NULL;
553:
554: SLIST_FOREACH(c, &m->mime_header, cgi_node)
555: if (!strcasecmp(c->cgi_name, name)) {
556: v = c->cgi_value;
557: break;
558: }
559: return v;
560: }
561:
562: /*
563: * mime_readPart() Read and parse MIME part
564: *
565: * @m = Mime part
566: * @str = String
567: * @len = String length
568: * return: -1 error or 0 ok
569: */
570: int
571: mime_readPart(struct tagMIME * __restrict m, const char *str, size_t len)
572: {
573: const char *eoh, *ct, *eb;
574: cgi_t *attr;
575: struct iovec bd;
576:
577: if (!m || !str || (ssize_t) len < 0) {
578: www_SetErr(EINVAL, "Mime part, string is NULL or length is less 0");
579: return -1;
580: }
581:
582: if (mime_parseHeader(m, str, len, &eoh))
583: return -1;
584:
585: ct = mime_getValue(m, "content-type");
586: if (!ct || www_cmptype(ct, "multipart")) {
587: /* not multi part, assign like body element */
588: m->mime_body.iov_base = malloc(len - (eoh - str) + 1);
589: if (!m->mime_body.iov_base) {
590: LOGERR;
591: freeHeader(m);
592: return -1;
593: }
594: memcpy(m->mime_body.iov_base, eoh, len - (eoh - str));
595: ((char*) m->mime_body.iov_base)[len - (eoh - str)] = 0;
596: m->mime_body.iov_len = len - (eoh - str) + 1;
597: } else {
598: /* multi part */
599: attr = www_parseAttributes(&ct);
600: if (!attr) {
601: freeHeader(m);
602: return -1;
603: }
604: bd.iov_base = bd_begin(www_getAttribute(attr, "boundary"));
605: bd.iov_len = strlen(bd.iov_base);
606: eb = findtextpos(eoh, len - (eoh - str), bd.iov_base, bd.iov_len);
607: free(bd.iov_base);
608:
609: /* set prolog if exists */
610: if (eb != eoh) {
611: m->mime_prolog.iov_base = malloc(eb - eoh + 1);
612: if (!m->mime_prolog.iov_base) {
613: LOGERR;
614: www_freeAttributes(&attr);
615: freeHeader(m);
616: return -1;
617: }
618: memcpy(m->mime_prolog.iov_base, eoh, eb - eoh);
619: ((char*) m->mime_prolog.iov_base)[eb - eoh] = 0;
620: m->mime_prolog.iov_len = eb - eoh + 1;
621: }
622:
623: m->mime_attach = mime_parseMultiPart(eb + 1, len - (eb + 1 - str),
624: www_getAttribute(attr, "boundary"), &eoh);
625:
626: /* set epilog if exists */
627: if (eoh - str < len) {
628: m->mime_epilog.iov_base = malloc(len - (eoh - str) + 1);
629: if (!m->mime_epilog.iov_base) {
630: LOGERR;
631: www_freeAttributes(&attr);
632: freeHeader(m);
633: return -1;
634: }
635: memcpy(m->mime_epilog.iov_base, str, len - (eoh - str));
636: ((char*) m->mime_epilog.iov_base)[len - (eoh - str)] = 0;
637: m->mime_epilog.iov_len = len - (eoh - str) + 1;
638:
639: }
640:
641: www_freeAttributes(&attr);
642: }
643:
644: return 0;
645: }
646:
647: /*
648: * mime_calcRawSize() - Calculate estimated memory for data from parsed MIME part
649: *
650: * @m = Mime part
651: * return: -1 error or >-1 data size in mime part
652: */
653: int
654: mime_calcRawSize(struct tagMIME * __restrict m)
655: {
656: const char *s;
657: char *t;
658: int len;
659: register int i;
660:
661: if (!m) {
662: www_SetErr(EINVAL, "Mime part is NULL");
663: return -1;
664: }
665:
666: /* no body */
667: if (m->mime_body.iov_len < 1)
668: return 0;
669:
670: s = mime_getValue(m, "content-transfer-encoding");
671: if (!s)
672: return m->mime_body.iov_len;
673: /* strip whitespaces */
674: while (isspace(*s))
675: s++;
676: t = strchr(s, ';');
677: len = t ? strlen(s) : t - s;
678:
679: /* find proper encoding */
680: for (i = 0; i < sizeof n_encode / sizeof *n_encode; i++)
681: if (len == strlen(n_encode[i]) && !strncasecmp(s, n_encode[i], len))
682: return m->mime_body.iov_len;
683:
684: for (i = 0; i < sizeof encode / sizeof *encode; i++)
685: if (len == strlen(encode[i].name) && !strncasecmp(s, encode[i].name, len))
686: return m->mime_body.iov_len * encode[i].mul;
687:
688: /* fail */
689: return -1;
690: }
691:
692: /*
693: * mime_getRawData() - Get ready parsed data from MIME part body
694: *
695: * @m = Mime part
696: * @str = output data buffer
697: * @len = output data buffer length
698: * return: -1 error or >-1 data length in output buffer
699: */
700: int
701: mime_getRawData(struct tagMIME * __restrict m, char * __restrict str, int slen)
702: {
703: const char *s;
704: char *t;
705: int len;
706: register int i;
707:
708: if (!m || !str) {
709: www_SetErr(EINVAL, "Mime part or string is NULL");
710: return -1;
711: }
712:
713: /* no body */
714: if (m->mime_body.iov_len < 1)
715: return 0;
716:
717: s = mime_getValue(m, "content-transfer-encoding");
718: if (!s) {
719: memcpy(str, m->mime_body.iov_base, m->mime_body.iov_len > (slen - 1) ?
720: slen - 1 : m->mime_body.iov_len);
721: return m->mime_body.iov_len;
722: }
723:
724: /* strip whitespaces */
725: while (isspace(*s))
726: s++;
727: t = strchr(s, ';');
728: len = t ? strlen(s) : t - s;
729:
730: /* decoding body */
731: for (i = 0; i < sizeof encode / sizeof *encode; i++)
732: if (len == strlen(encode[i].name) && !strncasecmp(s, encode[i].name, len))
733: return encode[i].decode(m->mime_body.iov_base,
734: m->mime_body.iov_len, str);
735:
736: /* fail */
737: return -1;
738: }
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to mime.c CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [ELWIX - Embedded LightWeight unIX -] / libaitwww / src