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 - 2015 Steffen (Daode) Nurpmeso <sdaoden@users.sf.net>.
   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 static bool_t  _mime_parse_part(struct message *zmp, struct mimepart *ip,
  43                   enum mime_parse_flags mpf, int level);
  44
  45 static void    _mime_parse_rfc822(struct message *zmp, struct mimepart *ip,
  46                   enum mime_parse_flags mpf, int level);
  47
  48 #ifdef HAVE_SSL
  49 static void    _mime_parse_pkcs7(struct message *zmp, struct mimepart *ip,
  50                   enum mime_parse_flags mpf, int level);
  51 #endif
  52
  53 static void    _mime_parse_multipart(struct message *zmp,
  54                   struct mimepart *ip, enum mime_parse_flags mpf, int level);
  55 static void    __mime_parse_new(struct mimepart *ip, struct mimepart **np,
  56                   off_t offs, int *part);
  57 static void    __mime_parse_end(struct mimepart **np, off_t xoffs,
  58                   long lines);
  59
  60 static bool_t
  61 _mime_parse_part(struct message *zmp, struct mimepart *ip,
  62    enum mime_parse_flags mpf, int level)
  63 {
  64    char *cp_b, *cp;
  65    bool_t rv = FAL0;
  66    NYD_ENTER;
  67
  68    ip->m_ct_type = hfield1("content-type", (struct message*)ip);
  69    if (ip->m_ct_type != NULL) {
  70       ip->m_ct_type_plain = cp_b = savestr(ip->m_ct_type);
  71       if ((cp = strchr(cp_b, ';')) != NULL)
  72          *cp = '\0';
  73       cp = cp_b + strlen(cp_b);
  74       while (cp > cp_b && blankchar(cp[-1]))
  75          --cp;
  76       *cp = '\0';
  77    } else if (ip->m_parent != NULL &&
  78          ip->m_parent->m_mimecontent == MIME_DIGEST)
  79       ip->m_ct_type_plain = "message/rfc822";
  80    else
  81       ip->m_ct_type_plain = "text/plain";
  82    ip->m_ct_type_usr_ovwr = NULL;
  83
  84    if (ip->m_ct_type != NULL)
  85       ip->m_charset = mime_param_get("charset", ip->m_ct_type);
  86    if (ip->m_charset == NULL)
  87       ip->m_charset = charset_get_7bit();
  88
  89    if ((ip->m_ct_enc = hfield1("content-transfer-encoding",
  90          (struct message*)ip)) == NULL)
  91       ip->m_ct_enc = mime_enc_from_conversion(CONV_7BIT);
  92    ip->m_mime_enc = mime_enc_from_ctehead(ip->m_ct_enc);
  93
  94    if (((cp = hfield1("content-disposition", (struct message*)ip)) == NULL ||
  95          (ip->m_filename = mime_param_get("filename", cp)) == NULL) &&
  96          ip->m_ct_type != NULL)
  97       ip->m_filename = mime_param_get("name", ip->m_ct_type);
  98
  99    ip->m_mimecontent = mime_type_classify_part(ip);
 100
 101    if (mpf & MIME_PARSE_PARTS) {
 102       if (level > 9999) { /* TODO MAGIC */
 103          n_err(_("MIME content too deeply nested\n"));
 104          goto jleave;
 105       }
 106       switch (ip->m_mimecontent) {
 107       case MIME_PKCS7:
 108          if (mpf & MIME_PARSE_DECRYPT) {
 109 #ifdef HAVE_SSL
 110             _mime_parse_pkcs7(zmp, ip, mpf, level);
 111             break;
 112 #else
 113             n_err(_("No SSL support compiled in\n"));
 114             goto jleave;
 115 #endif
 116          }
 117          /* FALLTHRU */
 118       default:
 119          break;
 120       case MIME_MULTI:
 121       case MIME_ALTERNATIVE:
 122       case MIME_RELATED: /* TODO /related yet handled like /alternative */
 123       case MIME_DIGEST:
 124          _mime_parse_multipart(zmp, ip, mpf, level);
 125          break;
 126       case MIME_822:
 127          _mime_parse_rfc822(zmp, ip, mpf, level);
 128          break;
 129       }
 130    }
 131
 132    rv = TRU1;
 133 jleave:
 134    NYD_LEAVE;
 135    return rv;
 136 }
 137
 138 static void
 139 _mime_parse_rfc822(struct message *zmp, struct mimepart *ip,
 140    enum mime_parse_flags mpf, int level)
 141 {
 142    int c, lastc = '\n';
 143    size_t cnt;
 144    FILE *ibuf;
 145    off_t offs;
 146    struct mimepart *np;
 147    long lines;
 148    NYD_ENTER;
 149
 150    if ((ibuf = setinput(&mb, (struct message*)ip, NEED_BODY)) == NULL)
 151       goto jleave;
 152
 153    cnt = ip->m_size;
 154    lines = ip->m_lines;
 155    while (cnt && ((c = getc(ibuf)) != EOF)) {
 156       --cnt;
 157       if (c == '\n') {
 158          --lines;
 159          if (lastc == '\n')
 160             break;
 161       }
 162       lastc = c;
 163    }
 164    offs = ftell(ibuf);
 165
 166    np = csalloc(1, sizeof *np);
 167    np->m_flag = MNOFROM;
 168    np->m_have = HAVE_HEADER | HAVE_BODY;
 169    np->m_block = mailx_blockof(offs);
 170    np->m_offset = mailx_offsetof(offs);
 171    np->m_size = np->m_xsize = cnt;
 172    np->m_lines = np->m_xlines = lines;
 173    np->m_partstring = ip->m_partstring;
 174    np->m_parent = ip;
 175    ip->m_multipart = np;
 176
 177    if (ok_blook(rfc822_body_from_)) {
 178       substdate((struct message*)np);
 179       np->m_from = fakefrom((struct message*)np);/* TODO strip MNOFROM flag? */
 180    }
 181
 182    _mime_parse_part(zmp, np, mpf, level + 1);
 183 jleave:
 184    NYD_LEAVE;
 185 }
 186
 187 #ifdef HAVE_SSL
 188 static void
 189 _mime_parse_pkcs7(struct message *zmp, struct mimepart *ip,
 190    enum mime_parse_flags mpf, int level)
 191 {
 192    struct message m, *xmp;
 193    struct mimepart *np;
 194    char *to, *cc;
 195    NYD_ENTER;
 196
 197    memcpy(&m, ip, sizeof m);
 198    to = hfield1("to", zmp);
 199    cc = hfield1("cc", zmp);
 200
 201    if ((xmp = smime_decrypt(&m, to, cc, 0)) != NULL) {
 202       np = csalloc(1, sizeof *np);
 203       np->m_flag = xmp->m_flag;
 204       np->m_have = xmp->m_have;
 205       np->m_block = xmp->m_block;
 206       np->m_offset = xmp->m_offset;
 207       np->m_size = xmp->m_size;
 208       np->m_xsize = xmp->m_xsize;
 209       np->m_lines = xmp->m_lines;
 210       np->m_xlines = xmp->m_xlines;
 211       np->m_partstring = ip->m_partstring;
 212
 213       if (_mime_parse_part(zmp, np, mpf, level + 1) == OKAY) {
 214          np->m_parent = ip;
 215          ip->m_multipart = np;
 216       }
 217    }
 218    NYD_LEAVE;
 219 }
 220 #endif /* HAVE_SSL */
 221
 222 static void
 223 _mime_parse_multipart(struct message *zmp, struct mimepart *ip,
 224    enum mime_parse_flags mpf, int level)
 225 {
 226    /* TODO Instead of the recursive multiple run parse we have today,
 227     * TODO the send/MIME layer rewrite must create a "tree" of parts with
 228     * TODO a single-pass parse, then address each part directly as
 229     * TODO necessary; since boundaries start with -- and the content
 230     * TODO rather forms a stack this is pretty cheap indeed! */
 231    struct mimepart *np = NULL;
 232    char *boundary, *line = NULL;
 233    size_t linesize = 0, linelen, cnt, boundlen;
 234    FILE *ibuf;
 235    off_t offs;
 236    int part = 0;
 237    long lines = 0;
 238    NYD_ENTER;
 239
 240    if ((boundary = mime_param_boundary_get(ip->m_ct_type, &linelen)) == NULL)
 241       goto jleave;
 242
 243    boundlen = linelen;
 244    if ((ibuf = setinput(&mb, (struct message*)ip, NEED_BODY)) == NULL)
 245       goto jleave;
 246
 247    cnt = ip->m_size;
 248    while (fgetline(&line, &linesize, &cnt, &linelen, ibuf, 0))
 249       if (line[0] == '\n')
 250          break;
 251    offs = ftell(ibuf);
 252
 253    __mime_parse_new(ip, &np, offs, NULL);
 254    while (fgetline(&line, &linesize, &cnt, &linelen, ibuf, 0)) {
 255       /* XXX linelen includes LF */
 256       if (!((lines > 0 || part == 0) && linelen > boundlen &&
 257             !strncmp(line, boundary, boundlen))) {
 258          ++lines;
 259          continue;
 260       }
 261
 262       /* Subpart boundary? */
 263       if (line[boundlen] == '\n') {
 264          offs = ftell(ibuf);
 265          if (part > 0) {
 266             __mime_parse_end(&np, offs - boundlen - 2, lines);
 267             __mime_parse_new(ip, &np, offs - boundlen - 2, NULL);
 268          }
 269          __mime_parse_end(&np, offs, 2);
 270          __mime_parse_new(ip, &np, offs, &part);
 271          lines = 0;
 272          continue;
 273       }
 274
 275       /* Final boundary?  Be aware of cases where there is no separating
 276        * newline in between boundaries, as has been seen in a message with
 277        * "Content-Type: multipart/appledouble;" */
 278       if (linelen < boundlen + 2)
 279          continue;
 280       linelen -= boundlen + 2;
 281       if (line[boundlen] != '-' || line[boundlen + 1] != '-' ||
 282             (linelen > 0 && line[boundlen + 2] != '\n'))
 283          continue;
 284       offs = ftell(ibuf);
 285       if (part != 0) {
 286          __mime_parse_end(&np, offs - boundlen - 4, lines);
 287          __mime_parse_new(ip, &np, offs - boundlen - 4, NULL);
 288       }
 289       __mime_parse_end(&np, offs + cnt, 2);
 290       break;
 291    }
 292    if (np) {
 293       offs = ftell(ibuf);
 294       __mime_parse_end(&np, offs, lines);
 295    }
 296
 297    for (np = ip->m_multipart; np != NULL; np = np->m_nextpart)
 298       if (np->m_mimecontent != MIME_DISCARD)
 299          _mime_parse_part(zmp, np, mpf, level + 1);
 300    free(line);
 301 jleave:
 302    NYD_LEAVE;
 303 }
 304
 305 static void
 306 __mime_parse_new(struct mimepart *ip, struct mimepart **np, off_t offs,
 307    int *part)
 308 {
 309    struct mimepart *pp;
 310    size_t sz;
 311    NYD_ENTER;
 312
 313    *np = csalloc(1, sizeof **np);
 314    (*np)->m_flag = MNOFROM;
 315    (*np)->m_have = HAVE_HEADER | HAVE_BODY;
 316    (*np)->m_block = mailx_blockof(offs);
 317    (*np)->m_offset = mailx_offsetof(offs);
 318
 319    if (part) {
 320       ++(*part);
 321       sz = (ip->m_partstring != NULL) ? strlen(ip->m_partstring) : 0;
 322       sz += 20;
 323       (*np)->m_partstring = salloc(sz);
 324       if (ip->m_partstring)
 325          snprintf((*np)->m_partstring, sz, "%s.%u", ip->m_partstring, *part);
 326       else
 327          snprintf((*np)->m_partstring, sz, "%u", *part);
 328    } else
 329       (*np)->m_mimecontent = MIME_DISCARD;
 330    (*np)->m_parent = ip;
 331
 332    if (ip->m_multipart) {
 333       for (pp = ip->m_multipart; pp->m_nextpart != NULL; pp = pp->m_nextpart)
 334          ;
 335       pp->m_nextpart = *np;
 336    } else
 337       ip->m_multipart = *np;
 338    NYD_LEAVE;
 339 }
 340
 341 static void
 342 __mime_parse_end(struct mimepart **np, off_t xoffs, long lines)
 343 {
 344    off_t offs;
 345    NYD_ENTER;
 346
 347    offs = mailx_positionof((*np)->m_block, (*np)->m_offset);
 348    (*np)->m_size = (*np)->m_xsize = xoffs - offs;
 349    (*np)->m_lines = (*np)->m_xlines = lines;
 350    *np = NULL;
 351    NYD_LEAVE;
 352 }
 353
 354 FL struct mimepart *
 355 mime_parse_msg(struct message *mp, enum mime_parse_flags mpf)
 356 {
 357    struct mimepart *ip;
 358    NYD_ENTER;
 359
 360    ip = csalloc(1, sizeof *ip);
 361    ip->m_flag = mp->m_flag;
 362    ip->m_have = mp->m_have;
 363    ip->m_block = mp->m_block;
 364    ip->m_offset = mp->m_offset;
 365    ip->m_size = mp->m_size;
 366    ip->m_xsize = mp->m_xsize;
 367    ip->m_lines = mp->m_lines;
 368    ip->m_xlines = mp->m_lines;
 369    if (!_mime_parse_part(mp, ip, mpf, 0))
 370       ip = NULL;
 371    NYD_LEAVE;
 372    return ip;
 373 }
 374
 375 /* s-it-mode */