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