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 - 2017 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 = ok_vlook(charset_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 ((cp = hfield1("content-description", (struct message*)ip)) != NULL)
 115       ip->m_content_description = cp;
 116
 117    if ((ip->m_mimecontent = mime_type_classify_part(ip)) == MIME_822) {
 118       /* TODO (v15) HACK: message/rfc822 is treated special, that this one is
 119        * TODO too stupid to apply content-decoding when (falsely) applied */
 120       if (ip->m_mime_enc != MIMEE_8B && ip->m_mime_enc != MIMEE_7B) {
 121          n_err(_("Pre-v15 %s cannot handle (falsely) encoded message/rfc822\n"
 122             "  (not 7bit or 8bit)!  Interpreting as text/plain!\n"),
 123             uagent);
 124          ip->m_mimecontent = MIME_TEXT_PLAIN;
 125       }
 126    }
 127
 128    if (mpf & MIME_PARSE_PARTS) {
 129       if (level > 9999) { /* TODO MAGIC */
 130          n_err(_("MIME content too deeply nested\n"));
 131          goto jleave;
 132       }
 133       switch (ip->m_mimecontent) {
 134       case MIME_PKCS7:
 135          if (mpf & MIME_PARSE_DECRYPT) {
 136 #ifdef HAVE_SSL
 137             _mime_parse_pkcs7(zmp, ip, mpf, level);
 138             if (ip->m_content_info & CI_ENCRYPTED_OK)
 139                ip->m_content_info |= CI_EXPANDED;
 140             break;
 141 #else
 142             n_err(_("No SSL / S/MIME support compiled in\n"));
 143             goto jleave;
 144 #endif
 145          }
 146          break;
 147       default:
 148          break;
 149       case MIME_ALTERNATIVE:
 150       case MIME_RELATED: /* TODO /related yet handled like /alternative */
 151       case MIME_DIGEST:
 152       case MIME_SIGNED:
 153       case MIME_ENCRYPTED:
 154       case MIME_MULTI:
 155          if (!_mime_parse_multipart(zmp, ip, mpf, level))
 156             goto jleave;
 157          break;
 158       case MIME_822:
 159          _mime_parse_rfc822(zmp, ip, mpf, level);
 160          break;
 161       }
 162    }
 163
 164    rv = TRU1;
 165 jleave:
 166    NYD_LEAVE;
 167    return rv;
 168 }
 169
 170 static void
 171 _mime_parse_rfc822(struct message *zmp, struct mimepart *ip,
 172    enum mime_parse_flags mpf, int level)
 173 {
 174    int c, lastc = '\n';
 175    size_t cnt;
 176    FILE *ibuf;
 177    off_t offs;
 178    struct mimepart *np;
 179    long lines;
 180    NYD_ENTER;
 181
 182    if ((ibuf = setinput(&mb, (struct message*)ip, NEED_BODY)) == NULL)
 183       goto jleave;
 184
 185    cnt = ip->m_size;
 186    lines = ip->m_lines;
 187    while (cnt && ((c = getc(ibuf)) != EOF)) {
 188       --cnt;
 189       if (c == '\n') {
 190          --lines;
 191          if (lastc == '\n')
 192             break;
 193       }
 194       lastc = c;
 195    }
 196    offs = ftell(ibuf);
 197
 198    np = csalloc(1, sizeof *np);
 199    np->m_flag = MNOFROM;
 200    np->m_content_info = CI_HAVE_HEADER | CI_HAVE_BODY;
 201    np->m_block = mailx_blockof(offs);
 202    np->m_offset = mailx_offsetof(offs);
 203    np->m_size = np->m_xsize = cnt;
 204    np->m_lines = np->m_xlines = lines;
 205    np->m_partstring = ip->m_partstring;
 206    np->m_parent = ip;
 207    ip->m_multipart = np;
 208
 209    if (!(mpf & MIME_PARSE_SHALLOW) && ok_blook(rfc822_body_from_)) {
 210       substdate((struct message*)np);
 211       np->m_from = fakefrom((struct message*)np);/* TODO strip MNOFROM flag? */
 212    }
 213
 214    _mime_parse_part(zmp, np, mpf, level + 1);
 215 jleave:
 216    NYD_LEAVE;
 217 }
 218
 219 #ifdef HAVE_SSL
 220 static void
 221 _mime_parse_pkcs7(struct message *zmp, struct mimepart *ip,
 222    enum mime_parse_flags mpf, int level)
 223 {
 224    struct message m, *xmp;
 225    struct mimepart *np;
 226    char *to, *cc;
 227    NYD_ENTER;
 228
 229    memcpy(&m, ip, sizeof m);
 230    to = hfield1("to", zmp);
 231    cc = hfield1("cc", zmp);
 232
 233    if ((xmp = smime_decrypt(&m, to, cc, 0)) != NULL) {
 234       np = csalloc(1, sizeof *np);
 235       np->m_flag = xmp->m_flag;
 236       np->m_content_info = xmp->m_content_info | CI_ENCRYPTED | CI_ENCRYPTED_OK;
 237       np->m_block = xmp->m_block;
 238       np->m_offset = xmp->m_offset;
 239       np->m_size = xmp->m_size;
 240       np->m_xsize = xmp->m_xsize;
 241       np->m_lines = xmp->m_lines;
 242       np->m_xlines = xmp->m_xlines;
 243
 244       /* TODO using part "1" for decrypted content is a hack */
 245       if ((np->m_partstring = ip->m_partstring) == NULL)
 246          ip->m_partstring = np->m_partstring = n_UNCONST("1");
 247
 248       if (_mime_parse_part(zmp, np, mpf, level + 1) == OKAY) {
 249          ip->m_content_info |= CI_ENCRYPTED | CI_ENCRYPTED_OK;
 250          np->m_parent = ip;
 251          ip->m_multipart = np;
 252       }
 253    } else
 254       ip->m_content_info |= CI_ENCRYPTED | CI_ENCRYPTED_BAD;
 255    NYD_LEAVE;
 256 }
 257 #endif /* HAVE_SSL */
 258
 259 static bool_t
 260 _mime_parse_multipart(struct message *zmp, struct mimepart *ip,
 261    enum mime_parse_flags mpf, int level)
 262 {
 263    struct mimepart *np = NULL;
 264    char *boundary, *line = NULL;
 265    size_t linesize = 0, linelen, cnt, boundlen;
 266    FILE *ibuf;
 267    off_t offs;
 268    int part = 0;
 269    long lines = 0;
 270    NYD_ENTER;
 271
 272    if ((boundary = mime_param_boundary_get(ip->m_ct_type, &linelen)) == NULL)
 273       goto jleave;
 274
 275    boundlen = linelen;
 276    if ((ibuf = setinput(&mb, (struct message*)ip, NEED_BODY)) == NULL)
 277       goto jleave;
 278
 279    cnt = ip->m_size;
 280    while (fgetline(&line, &linesize, &cnt, &linelen, ibuf, 0))
 281       if (line[0] == '\n')
 282          break;
 283    offs = ftell(ibuf);
 284
 285    /* TODO using part "1" for decrypted content is a hack */
 286    if (ip->m_partstring == NULL)
 287       ip->m_partstring = n_UNCONST("1");
 288    __mime_parse_new(ip, &np, offs, NULL);
 289
 290    while (fgetline(&line, &linesize, &cnt, &linelen, ibuf, 0)) {
 291       /* XXX linelen includes LF */
 292       if (!((lines > 0 || part == 0) && linelen > boundlen &&
 293             !strncmp(line, boundary, boundlen))) {
 294          ++lines;
 295          continue;
 296       }
 297
 298       /* Subpart boundary? */
 299       if (line[boundlen] == '\n') {
 300          offs = ftell(ibuf);
 301          if (part > 0) {
 302             __mime_parse_end(&np, offs - boundlen - 2, lines);
 303             __mime_parse_new(ip, &np, offs - boundlen - 2, NULL);
 304          }
 305          __mime_parse_end(&np, offs, 2);
 306          __mime_parse_new(ip, &np, offs, &part);
 307          lines = 0;
 308          continue;
 309       }
 310
 311       /* Final boundary?  Be aware of cases where there is no separating
 312        * newline in between boundaries, as has been seen in a message with
 313        * "Content-Type: multipart/appledouble;" */
 314       if (linelen < boundlen + 2)
 315          continue;
 316       linelen -= boundlen + 2;
 317       if (line[boundlen] != '-' || line[boundlen + 1] != '-' ||
 318             (linelen > 0 && line[boundlen + 2] != '\n'))
 319          continue;
 320       offs = ftell(ibuf);
 321       if (part != 0) {
 322          __mime_parse_end(&np, offs - boundlen - 4, lines);
 323          __mime_parse_new(ip, &np, offs - boundlen - 4, NULL);
 324       }
 325       __mime_parse_end(&np, offs + cnt, 2);
 326       break;
 327    }
 328    if (np) {
 329       offs = ftell(ibuf);
 330       __mime_parse_end(&np, offs, lines);
 331    }
 332
 333    for (np = ip->m_multipart; np != NULL; np = np->m_nextpart)
 334       if (np->m_mimecontent != MIME_DISCARD)
 335          _mime_parse_part(zmp, np, mpf, level + 1);
 336
 337 jleave:
 338    if (line != NULL)
 339       free(line);
 340    NYD_LEAVE;
 341    return (ip != NULL);
 342 }
 343
 344 static void
 345 __mime_parse_new(struct mimepart *ip, struct mimepart **np, off_t offs,
 346    int *part)
 347 {
 348    struct mimepart *pp;
 349    size_t sz;
 350    NYD_ENTER;
 351
 352    *np = csalloc(1, sizeof **np);
 353    (*np)->m_flag = MNOFROM;
 354    (*np)->m_content_info = CI_HAVE_HEADER | CI_HAVE_BODY;
 355    (*np)->m_block = mailx_blockof(offs);
 356    (*np)->m_offset = mailx_offsetof(offs);
 357
 358    if (part) {
 359       ++(*part);
 360       sz = (ip->m_partstring != NULL) ? strlen(ip->m_partstring) : 0;
 361       sz += 20;
 362       (*np)->m_partstring = salloc(sz);
 363       if (ip->m_partstring)
 364          snprintf((*np)->m_partstring, sz, "%s.%u", ip->m_partstring, *part);
 365       else
 366          snprintf((*np)->m_partstring, sz, "%u", *part);
 367    } else
 368       (*np)->m_mimecontent = MIME_DISCARD;
 369    (*np)->m_parent = ip;
 370
 371    if (ip->m_multipart) {
 372       for (pp = ip->m_multipart; pp->m_nextpart != NULL; pp = pp->m_nextpart)
 373          ;
 374       pp->m_nextpart = *np;
 375    } else
 376       ip->m_multipart = *np;
 377    NYD_LEAVE;
 378 }
 379
 380 static void
 381 __mime_parse_end(struct mimepart **np, off_t xoffs, long lines)
 382 {
 383    off_t offs;
 384    NYD_ENTER;
 385
 386    offs = mailx_positionof((*np)->m_block, (*np)->m_offset);
 387    (*np)->m_size = (*np)->m_xsize = xoffs - offs;
 388    (*np)->m_lines = (*np)->m_xlines = lines;
 389    *np = NULL;
 390    NYD_LEAVE;
 391 }
 392
 393 FL struct mimepart *
 394 mime_parse_msg(struct message *mp, enum mime_parse_flags mpf)
 395 {
 396    struct mimepart *ip;
 397    NYD_ENTER;
 398
 399    ip = csalloc(1, sizeof *ip);
 400    ip->m_flag = mp->m_flag;
 401    ip->m_content_info = mp->m_content_info;
 402    ip->m_block = mp->m_block;
 403    ip->m_offset = mp->m_offset;
 404    ip->m_size = mp->m_size;
 405    ip->m_xsize = mp->m_xsize;
 406    ip->m_lines = mp->m_lines;
 407    ip->m_xlines = mp->m_lines;
 408    if (!_mime_parse_part(mp, ip, mpf, 0))
 409       ip = NULL;
 410    NYD_LEAVE;
 411    return ip;
 412 }
 413
 414 /* s-it-mode */