mime_parse.c

   1 /*@ S-nail - a mail user agent derived from Berkeley Mail.
   2  *@ Parse a message into a tree of struct mimepart objects.
   3  *
   4  * Copyright (c) 2000-2004 Gunnar Ritter, Freiburg i. Br., Germany.
   5  * Copyright (c) 2012 - 2016 Steffen (Daode) Nurpmeso <steffen@sdaoden.eu>.
   6  */
   7 /*
   8  * Copyright (c) 1980, 1993
   9  *      The Regents of the University of California.  All rights reserved.
  10  *
  11  * Redistribution and use in source and binary forms, with or without
  12  * modification, are permitted provided that the following conditions
  13  * are met:
  14  * 1. Redistributions of source code must retain the above copyright
  15  *    notice, this list of conditions and the following disclaimer.
  16  * 2. Redistributions in binary form must reproduce the above copyright
  17  *    notice, this list of conditions and the following disclaimer in the
  18  *    documentation and/or other materials provided with the distribution.
  19  * 3. Neither the name of the University nor the names of its contributors
  20  *    may be used to endorse or promote products derived from this software
  21  *    without specific prior written permission.
  22  *
  23  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  24  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  25  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  26  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  27  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  28  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  29  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  30  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  31  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  32  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  33  * SUCH DAMAGE.
  34  */
  35 #undef n_FILE
  36 #define n_FILE mime_parse
  37
  38 #ifndef HAVE_AMALGAMATION
  39 # include "nail.h"
  40 #endif
  41
  42 /* Fetch plain */
  43 static char *  _mime_parse_ct_plain_from_ct(char const *cth);
  44
  45 static bool_t  _mime_parse_part(struct message *zmp, struct mimepart *ip,
  46                   enum mime_parse_flags mpf, int level);
  47
  48 static void    _mime_parse_rfc822(struct message *zmp, struct mimepart *ip,
  49                   enum mime_parse_flags mpf, int level);
  50
  51 #ifdef HAVE_SSL
  52 static void    _mime_parse_pkcs7(struct message *zmp, struct mimepart *ip,
  53                   enum mime_parse_flags mpf, int level);
  54 #endif
  55
  56 static bool_t  _mime_parse_multipart(struct message *zmp,
  57                   struct mimepart *ip, enum mime_parse_flags mpf, int level);
  58 static void    __mime_parse_new(struct mimepart *ip, struct mimepart **np,
  59                   off_t offs, int *part);
  60 static void    __mime_parse_end(struct mimepart **np, off_t xoffs,
  61                   long lines);
  62
  63 static char *
  64 _mime_parse_ct_plain_from_ct(char const *cth)
  65 {
  66    char *rv_b, *rv;
  67    NYD2_ENTER;
  68
  69    rv_b = savestr(cth);
  70
  71    if ((rv = strchr(rv_b, ';')) != NULL)
  72       *rv = '\0';
  73
  74    rv = rv_b + strlen(rv_b);
  75    while (rv > rv_b && blankchar(rv[-1]))
  76       --rv;
  77    *rv = '\0';
  78    NYD2_LEAVE;
  79    return rv_b;
  80 }
  81
  82 static bool_t
  83 _mime_parse_part(struct message *zmp, struct mimepart *ip,
  84    enum mime_parse_flags mpf, int level)
  85 {
  86    char *cp;
  87    bool_t rv = FAL0;
  88    NYD_ENTER;
  89
  90    ip->m_ct_type = hfield1("content-type", (struct message*)ip);
  91    if (ip->m_ct_type != NULL)
  92       ip->m_ct_type_plain = _mime_parse_ct_plain_from_ct(ip->m_ct_type);
  93    else if (ip->m_parent != NULL && ip->m_parent->m_mimecontent == MIME_DIGEST)
  94       ip->m_ct_type_plain = "message/rfc822";
  95    else
  96       ip->m_ct_type_plain = "text/plain";
  97    ip->m_ct_type_usr_ovwr = NULL;
  98
  99    if (ip->m_ct_type != NULL)
 100       ip->m_charset = mime_param_get("charset", ip->m_ct_type);
 101    if (ip->m_charset == NULL)
 102       ip->m_charset = charset_get_7bit();
 103
 104    if ((ip->m_ct_enc = hfield1("content-transfer-encoding",
 105          (struct message*)ip)) == NULL)
 106       ip->m_ct_enc = mime_enc_from_conversion(CONV_7BIT);
 107    ip->m_mime_enc = mime_enc_from_ctehead(ip->m_ct_enc);
 108
 109    if (((cp = hfield1("content-disposition", (struct message*)ip)) == NULL ||
 110          (ip->m_filename = mime_param_get("filename", cp)) == NULL) &&
 111          ip->m_ct_type != NULL)
 112       ip->m_filename = mime_param_get("name", ip->m_ct_type);
 113
 114    if ((ip->m_mimecontent = mime_type_classify_part(ip)) == MIME_822) {
 115       /* TODO (v15) HACK: message/rfc822 is treated special, that this one is
 116        * TODO too stupid to apply content-decoding when (falsely) applied */
 117       if (ip->m_mime_enc != MIMEE_8B && ip->m_mime_enc != MIMEE_7B) {
 118          n_err(_("Pre-v15 %s cannot handle (falsely) encoded message/rfc822\n"
 119             "  (not 7bit or 8bit)!  Interpreting as text/plain!\n"),
 120             uagent);
 121          ip->m_mimecontent = MIME_TEXT_PLAIN;
 122       }
 123    }
 124
 125    if (mpf & MIME_PARSE_PARTS) {
 126       if (level > 9999) { /* TODO MAGIC */
 127          n_err(_("MIME content too deeply nested\n"));
 128          goto jleave;
 129       }
 130       switch (ip->m_mimecontent) {
 131       case MIME_PKCS7:
 132          if (mpf & MIME_PARSE_DECRYPT) {
 133 #ifdef HAVE_SSL
 134             _mime_parse_pkcs7(zmp, ip, mpf, level);
 135             if (ip->m_content_info & CI_ENCRYPTED_OK)
 136                ip->m_content_info |= CI_EXPANDED;
 137             break;
 138 #else
 139             n_err(_("No SSL / S/MIME support compiled in\n"));
 140             goto jleave;
 141 #endif
 142          }
 143          break;
 144       default:
 145          break;
 146       case MIME_ALTERNATIVE:
 147       case MIME_RELATED: /* TODO /related yet handled like /alternative */
 148       case MIME_DIGEST:
 149       case MIME_SIGNED:
 150       case MIME_ENCRYPTED:
 151       case MIME_MULTI:
 152          if (!_mime_parse_multipart(zmp, ip, mpf, level))
 153             goto jleave;
 154          break;
 155       case MIME_822:
 156          _mime_parse_rfc822(zmp, ip, mpf, level);
 157          break;
 158       }
 159    }
 160
 161    rv = TRU1;
 162 jleave:
 163    NYD_LEAVE;
 164    return rv;
 165 }
 166
 167 static void
 168 _mime_parse_rfc822(struct message *zmp, struct mimepart *ip,
 169    enum mime_parse_flags mpf, int level)
 170 {
 171    int c, lastc = '\n';
 172    size_t cnt;
 173    FILE *ibuf;
 174    off_t offs;
 175    struct mimepart *np;
 176    long lines;
 177    NYD_ENTER;
 178
 179    if ((ibuf = setinput(&mb, (struct message*)ip, NEED_BODY)) == NULL)
 180       goto jleave;
 181
 182    cnt = ip->m_size;
 183    lines = ip->m_lines;
 184    while (cnt && ((c = getc(ibuf)) != EOF)) {
 185       --cnt;
 186       if (c == '\n') {
 187          --lines;
 188          if (lastc == '\n')
 189             break;
 190       }
 191       lastc = c;
 192    }
 193    offs = ftell(ibuf);
 194
 195    np = csalloc(1, sizeof *np);
 196    np->m_flag = MNOFROM;
 197    np->m_content_info = CI_HAVE_HEADER | CI_HAVE_BODY;
 198    np->m_block = mailx_blockof(offs);
 199    np->m_offset = mailx_offsetof(offs);
 200    np->m_size = np->m_xsize = cnt;
 201    np->m_lines = np->m_xlines = lines;
 202    np->m_partstring = ip->m_partstring;
 203    np->m_parent = ip;
 204    ip->m_multipart = np;
 205
 206    if (!(mpf & MIME_PARSE_SHALLOW) && ok_blook(rfc822_body_from_)) {
 207       substdate((struct message*)np);
 208       np->m_from = fakefrom((struct message*)np);/* TODO strip MNOFROM flag? */
 209    }
 210
 211    _mime_parse_part(zmp, np, mpf, level + 1);
 212 jleave:
 213    NYD_LEAVE;
 214 }
 215
 216 #ifdef HAVE_SSL
 217 static void
 218 _mime_parse_pkcs7(struct message *zmp, struct mimepart *ip,
 219    enum mime_parse_flags mpf, int level)
 220 {
 221    struct message m, *xmp;
 222    struct mimepart *np;
 223    char *to, *cc;
 224    NYD_ENTER;
 225
 226    memcpy(&m, ip, sizeof m);
 227    to = hfield1("to", zmp);
 228    cc = hfield1("cc", zmp);
 229
 230    if ((xmp = smime_decrypt(&m, to, cc, 0)) != NULL) {
 231       np = csalloc(1, sizeof *np);
 232       np->m_flag = xmp->m_flag;
 233       np->m_content_info = xmp->m_content_info | CI_ENCRYPTED | CI_ENCRYPTED_OK;
 234       np->m_block = xmp->m_block;
 235       np->m_offset = xmp->m_offset;
 236       np->m_size = xmp->m_size;
 237       np->m_xsize = xmp->m_xsize;
 238       np->m_lines = xmp->m_lines;
 239       np->m_xlines = xmp->m_xlines;
 240
 241       /* TODO using part "1" for decrypted content is a hack */
 242       if ((np->m_partstring = ip->m_partstring) == NULL)
 243          ip->m_partstring = np->m_partstring = n_UNCONST("1");
 244
 245       if (_mime_parse_part(zmp, np, mpf, level + 1) == OKAY) {
 246          ip->m_content_info |= CI_ENCRYPTED | CI_ENCRYPTED_OK;
 247          np->m_parent = ip;
 248          ip->m_multipart = np;
 249       }
 250    } else
 251       ip->m_content_info |= CI_ENCRYPTED | CI_ENCRYPTED_BAD;
 252    NYD_LEAVE;
 253 }
 254 #endif /* HAVE_SSL */
 255
 256 static bool_t
 257 _mime_parse_multipart(struct message *zmp, struct mimepart *ip,
 258    enum mime_parse_flags mpf, int level)
 259 {
 260    struct mimepart *np = NULL;
 261    char *boundary, *line = NULL;
 262    size_t linesize = 0, linelen, cnt, boundlen;
 263    FILE *ibuf;
 264    off_t offs;
 265    int part = 0;
 266    long lines = 0;
 267    NYD_ENTER;
 268
 269    if ((boundary = mime_param_boundary_get(ip->m_ct_type, &linelen)) == NULL)
 270       goto jleave;
 271
 272    boundlen = linelen;
 273    if ((ibuf = setinput(&mb, (struct message*)ip, NEED_BODY)) == NULL)
 274       goto jleave;
 275
 276    cnt = ip->m_size;
 277    while (fgetline(&line, &linesize, &cnt, &linelen, ibuf, 0))
 278       if (line[0] == '\n')
 279          break;
 280    offs = ftell(ibuf);
 281
 282    /* TODO using part "1" for decrypted content is a hack */
 283    if (ip->m_partstring == NULL)
 284       ip->m_partstring = n_UNCONST("1");
 285    __mime_parse_new(ip, &np, offs, NULL);
 286
 287    while (fgetline(&line, &linesize, &cnt, &linelen, ibuf, 0)) {
 288       /* XXX linelen includes LF */
 289       if (!((lines > 0 || part == 0) && linelen > boundlen &&
 290             !strncmp(line, boundary, boundlen))) {
 291          ++lines;
 292          continue;
 293       }
 294
 295       /* Subpart boundary? */
 296       if (line[boundlen] == '\n') {
 297          offs = ftell(ibuf);
 298          if (part > 0) {
 299             __mime_parse_end(&np, offs - boundlen - 2, lines);
 300             __mime_parse_new(ip, &np, offs - boundlen - 2, NULL);
 301          }
 302          __mime_parse_end(&np, offs, 2);
 303          __mime_parse_new(ip, &np, offs, &part);
 304          lines = 0;
 305          continue;
 306       }
 307
 308       /* Final boundary?  Be aware of cases where there is no separating
 309        * newline in between boundaries, as has been seen in a message with
 310        * "Content-Type: multipart/appledouble;" */
 311       if (linelen < boundlen + 2)
 312          continue;
 313       linelen -= boundlen + 2;
 314       if (line[boundlen] != '-' || line[boundlen + 1] != '-' ||
 315             (linelen > 0 && line[boundlen + 2] != '\n'))
 316          continue;
 317       offs = ftell(ibuf);
 318       if (part != 0) {
 319          __mime_parse_end(&np, offs - boundlen - 4, lines);
 320          __mime_parse_new(ip, &np, offs - boundlen - 4, NULL);
 321       }
 322       __mime_parse_end(&np, offs + cnt, 2);
 323       break;
 324    }
 325    if (np) {
 326       offs = ftell(ibuf);
 327       __mime_parse_end(&np, offs, lines);
 328    }
 329
 330    for (np = ip->m_multipart; np != NULL; np = np->m_nextpart)
 331       if (np->m_mimecontent != MIME_DISCARD)
 332          _mime_parse_part(zmp, np, mpf, level + 1);
 333
 334 jleave:
 335    if (line != NULL)
 336       free(line);
 337    NYD_LEAVE;
 338    return (ip != NULL);
 339 }
 340
 341 static void
 342 __mime_parse_new(struct mimepart *ip, struct mimepart **np, off_t offs,
 343    int *part)
 344 {
 345    struct mimepart *pp;
 346    size_t sz;
 347    NYD_ENTER;
 348
 349    *np = csalloc(1, sizeof **np);
 350    (*np)->m_flag = MNOFROM;
 351    (*np)->m_content_info = CI_HAVE_HEADER | CI_HAVE_BODY;
 352    (*np)->m_block = mailx_blockof(offs);
 353    (*np)->m_offset = mailx_offsetof(offs);
 354
 355    if (part) {
 356       ++(*part);
 357       sz = (ip->m_partstring != NULL) ? strlen(ip->m_partstring) : 0;
 358       sz += 20;
 359       (*np)->m_partstring = salloc(sz);
 360       if (ip->m_partstring)
 361          snprintf((*np)->m_partstring, sz, "%s.%u", ip->m_partstring, *part);
 362       else
 363          snprintf((*np)->m_partstring, sz, "%u", *part);
 364    } else
 365       (*np)->m_mimecontent = MIME_DISCARD;
 366    (*np)->m_parent = ip;
 367
 368    if (ip->m_multipart) {
 369       for (pp = ip->m_multipart; pp->m_nextpart != NULL; pp = pp->m_nextpart)
 370          ;
 371       pp->m_nextpart = *np;
 372    } else
 373       ip->m_multipart = *np;
 374    NYD_LEAVE;
 375 }
 376
 377 static void
 378 __mime_parse_end(struct mimepart **np, off_t xoffs, long lines)
 379 {
 380    off_t offs;
 381    NYD_ENTER;
 382
 383    offs = mailx_positionof((*np)->m_block, (*np)->m_offset);
 384    (*np)->m_size = (*np)->m_xsize = xoffs - offs;
 385    (*np)->m_lines = (*np)->m_xlines = lines;
 386    *np = NULL;
 387    NYD_LEAVE;
 388 }
 389
 390 FL struct mimepart *
 391 mime_parse_msg(struct message *mp, enum mime_parse_flags mpf)
 392 {
 393    struct mimepart *ip;
 394    NYD_ENTER;
 395
 396    ip = csalloc(1, sizeof *ip);
 397    ip->m_flag = mp->m_flag;
 398    ip->m_content_info = mp->m_content_info;
 399    ip->m_block = mp->m_block;
 400    ip->m_offset = mp->m_offset;
 401    ip->m_size = mp->m_size;
 402    ip->m_xsize = mp->m_xsize;
 403    ip->m_lines = mp->m_lines;
 404    ip->m_xlines = mp->m_lines;
 405    if (!_mime_parse_part(mp, ip, mpf, 0))
 406       ip = NULL;
 407    NYD_LEAVE;
 408    return ip;
 409 }
 410
 411 /* s-it-mode */