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 - 2018 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    else
 104       ip->m_charset = n_charsetalias_expand(ip->m_charset);
 105
 106    if ((ip->m_ct_enc = hfield1("content-transfer-encoding",
 107          (struct message*)ip)) == NULL)
 108       ip->m_ct_enc = mime_enc_from_conversion(CONV_7BIT);
 109    ip->m_mime_enc = mime_enc_from_ctehead(ip->m_ct_enc);
 110
 111    if (((cp = hfield1("content-disposition", (struct message*)ip)) == NULL ||
 112          (ip->m_filename = mime_param_get("filename", cp)) == NULL) &&
 113          ip->m_ct_type != NULL)
 114       ip->m_filename = mime_param_get("name", ip->m_ct_type);
 115
 116    if ((cp = hfield1("content-description", (struct message*)ip)) != NULL)
 117       ip->m_content_description = cp;
 118
 119    if ((ip->m_mimecontent = n_mimetype_classify_part(ip,
 120          ((mpf & MIME_PARSE_FOR_USER_CONTEXT) != 0))) == MIME_822) {
 121       /* TODO (v15) HACK: message/rfc822 is treated special, that this one is
 122        * TODO too stupid to apply content-decoding when (falsely) applied */
 123       if (ip->m_mime_enc != MIMEE_8B && ip->m_mime_enc != MIMEE_7B) {
 124          n_err(_("Pre-v15 %s cannot handle (falsely) encoded message/rfc822\n"
 125             "  (not 7bit or 8bit)!  Interpreting as text/plain!\n"),
 126             n_uagent);
 127          ip->m_mimecontent = MIME_TEXT_PLAIN;
 128       }
 129    }
 130
 131    assert(ip->m_external_body_url == NULL);
 132    if(!asccasecmp(ip->m_ct_type_plain, "message/external-body") &&
 133          (cp = mime_param_get("access-type", ip->m_ct_type)) != NULL &&
 134          !asccasecmp(cp, "URL"))
 135       ip->m_external_body_url = mime_param_get("url", ip->m_ct_type);
 136
 137    if (mpf & MIME_PARSE_PARTS) {
 138       if (level > 9999) { /* TODO MAGIC */
 139          n_err(_("MIME content too deeply nested\n"));
 140          goto jleave;
 141       }
 142       switch (ip->m_mimecontent) {
 143       case MIME_PKCS7:
 144          if (mpf & MIME_PARSE_DECRYPT) {
 145 #ifdef HAVE_SSL
 146             _mime_parse_pkcs7(zmp, ip, mpf, level);
 147             if (ip->m_content_info & CI_ENCRYPTED_OK)
 148                ip->m_content_info |= CI_EXPANDED;
 149             break;
 150 #else
 151             n_err(_("No SSL / S/MIME support compiled in\n"));
 152             goto jleave;
 153 #endif
 154          }
 155          break;
 156       default:
 157          break;
 158       case MIME_ALTERNATIVE:
 159       case MIME_RELATED: /* TODO /related yet handled like /alternative */
 160       case MIME_DIGEST:
 161       case MIME_SIGNED:
 162       case MIME_ENCRYPTED:
 163       case MIME_MULTI:
 164          if (!_mime_parse_multipart(zmp, ip, mpf, level))
 165             goto jleave;
 166          break;
 167       case MIME_822:
 168          _mime_parse_rfc822(zmp, ip, mpf, level);
 169          break;
 170       }
 171    }
 172
 173    rv = TRU1;
 174 jleave:
 175    NYD_LEAVE;
 176    return rv;
 177 }
 178
 179 static void
 180 _mime_parse_rfc822(struct message *zmp, struct mimepart *ip,
 181    enum mime_parse_flags mpf, int level)
 182 {
 183    int c, lastc = '\n';
 184    size_t cnt;
 185    FILE *ibuf;
 186    off_t offs;
 187    struct mimepart *np;
 188    long lines;
 189    NYD_ENTER;
 190
 191    if ((ibuf = setinput(&mb, (struct message*)ip, NEED_BODY)) == NULL)
 192       goto jleave;
 193
 194    cnt = ip->m_size;
 195    lines = ip->m_lines;
 196    while (cnt && ((c = getc(ibuf)) != EOF)) {
 197       --cnt;
 198       if (c == '\n') {
 199          --lines;
 200          if (lastc == '\n')
 201             break;
 202       }
 203       lastc = c;
 204    }
 205    offs = ftell(ibuf);
 206
 207    np = csalloc(1, sizeof *np);
 208    np->m_flag = MNOFROM;
 209    np->m_content_info = CI_HAVE_HEADER | CI_HAVE_BODY;
 210    np->m_block = mailx_blockof(offs);
 211    np->m_offset = mailx_offsetof(offs);
 212    np->m_size = np->m_xsize = cnt;
 213    np->m_lines = np->m_xlines = lines;
 214    np->m_partstring = ip->m_partstring;
 215    np->m_parent = ip;
 216    ip->m_multipart = np;
 217
 218    if (!(mpf & MIME_PARSE_SHALLOW) && ok_blook(rfc822_body_from_)) {
 219       substdate((struct message*)np);
 220       np->m_from = fakefrom((struct message*)np);/* TODO strip MNOFROM flag? */
 221    }
 222
 223    _mime_parse_part(zmp, np, mpf, level + 1);
 224 jleave:
 225    NYD_LEAVE;
 226 }
 227
 228 #ifdef HAVE_SSL
 229 static void
 230 _mime_parse_pkcs7(struct message *zmp, struct mimepart *ip,
 231    enum mime_parse_flags mpf, int level)
 232 {
 233    struct message m, *xmp;
 234    struct mimepart *np;
 235    char *to, *cc;
 236    NYD_ENTER;
 237
 238    memcpy(&m, ip, sizeof m);
 239    to = hfield1("to", zmp);
 240    cc = hfield1("cc", zmp);
 241
 242    if ((xmp = smime_decrypt(&m, to, cc, 0)) != NULL) {
 243       np = csalloc(1, sizeof *np);
 244       np->m_flag = xmp->m_flag;
 245       np->m_content_info = xmp->m_content_info | CI_ENCRYPTED | CI_ENCRYPTED_OK;
 246       np->m_block = xmp->m_block;
 247       np->m_offset = xmp->m_offset;
 248       np->m_size = xmp->m_size;
 249       np->m_xsize = xmp->m_xsize;
 250       np->m_lines = xmp->m_lines;
 251       np->m_xlines = xmp->m_xlines;
 252
 253       /* TODO using part "1" for decrypted content is a hack */
 254       if ((np->m_partstring = ip->m_partstring) == NULL)
 255          ip->m_partstring = np->m_partstring = n_UNCONST(n_1);
 256
 257       if (_mime_parse_part(zmp, np, mpf, level + 1) == OKAY) {
 258          ip->m_content_info |= CI_ENCRYPTED | CI_ENCRYPTED_OK;
 259          np->m_parent = ip;
 260          ip->m_multipart = np;
 261       }
 262    } else
 263       ip->m_content_info |= CI_ENCRYPTED | CI_ENCRYPTED_BAD;
 264    NYD_LEAVE;
 265 }
 266 #endif /* HAVE_SSL */
 267
 268 static bool_t
 269 _mime_parse_multipart(struct message *zmp, struct mimepart *ip,
 270    enum mime_parse_flags mpf, int level)
 271 {
 272    struct mimepart *np = NULL;
 273    char *boundary, *line = NULL;
 274    size_t linesize = 0, linelen, cnt, boundlen;
 275    FILE *ibuf;
 276    off_t offs;
 277    int part = 0;
 278    long lines = 0;
 279    NYD_ENTER;
 280
 281    if ((boundary = mime_param_boundary_get(ip->m_ct_type, &linelen)) == NULL)
 282       goto jleave;
 283
 284    boundlen = linelen;
 285    if ((ibuf = setinput(&mb, (struct message*)ip, NEED_BODY)) == NULL)
 286       goto jleave;
 287
 288    cnt = ip->m_size;
 289    while (fgetline(&line, &linesize, &cnt, &linelen, ibuf, 0))
 290       if (line[0] == '\n')
 291          break;
 292    offs = ftell(ibuf);
 293
 294    /* TODO using part "1" for decrypted content is a hack */
 295    if (ip->m_partstring == NULL)
 296       ip->m_partstring = n_UNCONST("1");
 297    __mime_parse_new(ip, &np, offs, NULL);
 298
 299    while (fgetline(&line, &linesize, &cnt, &linelen, ibuf, 0)) {
 300       /* XXX linelen includes LF */
 301       if (!((lines > 0 || part == 0) && linelen > boundlen &&
 302             !strncmp(line, boundary, boundlen))) {
 303          ++lines;
 304          continue;
 305       }
 306
 307       /* Subpart boundary? */
 308       if (line[boundlen] == '\n') {
 309          offs = ftell(ibuf);
 310          if (part > 0) {
 311             __mime_parse_end(&np, offs - boundlen - 2, lines);
 312             __mime_parse_new(ip, &np, offs - boundlen - 2, NULL);
 313          }
 314          __mime_parse_end(&np, offs, 2);
 315          __mime_parse_new(ip, &np, offs, &part);
 316          lines = 0;
 317          continue;
 318       }
 319
 320       /* Final boundary?  Be aware of cases where there is no separating
 321        * newline in between boundaries, as has been seen in a message with
 322        * "Content-Type: multipart/appledouble;" */
 323       if (linelen < boundlen + 2)
 324          continue;
 325       linelen -= boundlen + 2;
 326       if (line[boundlen] != '-' || line[boundlen + 1] != '-' ||
 327             (linelen > 0 && line[boundlen + 2] != '\n'))
 328          continue;
 329       offs = ftell(ibuf);
 330       if (part != 0) {
 331          __mime_parse_end(&np, offs - boundlen - 4, lines);
 332          __mime_parse_new(ip, &np, offs - boundlen - 4, NULL);
 333       }
 334       __mime_parse_end(&np, offs + cnt, 2);
 335       break;
 336    }
 337    if (np) {
 338       offs = ftell(ibuf);
 339       __mime_parse_end(&np, offs, lines);
 340    }
 341
 342    for (np = ip->m_multipart; np != NULL; np = np->m_nextpart)
 343       if (np->m_mimecontent != MIME_DISCARD)
 344          _mime_parse_part(zmp, np, mpf, level + 1);
 345
 346 jleave:
 347    if (line != NULL)
 348       free(line);
 349    NYD_LEAVE;
 350    return TRU1;
 351 }
 352
 353 static void
 354 __mime_parse_new(struct mimepart *ip, struct mimepart **np, off_t offs,
 355    int *part)
 356 {
 357    struct mimepart *pp;
 358    size_t sz;
 359    NYD_ENTER;
 360
 361    *np = csalloc(1, sizeof **np);
 362    (*np)->m_flag = MNOFROM;
 363    (*np)->m_content_info = CI_HAVE_HEADER | CI_HAVE_BODY;
 364    (*np)->m_block = mailx_blockof(offs);
 365    (*np)->m_offset = mailx_offsetof(offs);
 366
 367    if (part) {
 368       ++(*part);
 369       sz = (ip->m_partstring != NULL) ? strlen(ip->m_partstring) : 0;
 370       sz += 20;
 371       (*np)->m_partstring = salloc(sz);
 372       if (ip->m_partstring)
 373          snprintf((*np)->m_partstring, sz, "%s.%u", ip->m_partstring, *part);
 374       else
 375          snprintf((*np)->m_partstring, sz, "%u", *part);
 376    } else
 377       (*np)->m_mimecontent = MIME_DISCARD;
 378    (*np)->m_parent = ip;
 379
 380    if (ip->m_multipart) {
 381       for (pp = ip->m_multipart; pp->m_nextpart != NULL; pp = pp->m_nextpart)
 382          ;
 383       pp->m_nextpart = *np;
 384    } else
 385       ip->m_multipart = *np;
 386    NYD_LEAVE;
 387 }
 388
 389 static void
 390 __mime_parse_end(struct mimepart **np, off_t xoffs, long lines)
 391 {
 392    off_t offs;
 393    NYD_ENTER;
 394
 395    offs = mailx_positionof((*np)->m_block, (*np)->m_offset);
 396    (*np)->m_size = (*np)->m_xsize = xoffs - offs;
 397    (*np)->m_lines = (*np)->m_xlines = lines;
 398    *np = NULL;
 399    NYD_LEAVE;
 400 }
 401
 402 FL struct mimepart *
 403 mime_parse_msg(struct message *mp, enum mime_parse_flags mpf)
 404 {
 405    struct mimepart *ip;
 406    NYD_ENTER;
 407
 408    ip = csalloc(1, sizeof *ip);
 409    ip->m_flag = mp->m_flag;
 410    ip->m_content_info = mp->m_content_info;
 411    ip->m_block = mp->m_block;
 412    ip->m_offset = mp->m_offset;
 413    ip->m_size = mp->m_size;
 414    ip->m_xsize = mp->m_xsize;
 415    ip->m_lines = mp->m_lines;
 416    ip->m_xlines = mp->m_lines;
 417    if (!_mime_parse_part(mp, ip, mpf, 0))
 418       ip = NULL;
 419    NYD_LEAVE;
 420    return ip;
 421 }
 422
 423 /* s-it-mode */