parse.c

mutt-1.5.12 源代码。linux 下邮件接受的工具。

    mutt_str_replace (&b->subtype, "plain");
  }
}

/* parse a MESSAGE/RFC822 body
 *
 * args:
 *	fp		stream to read from
 *
 *	parent		structure which contains info about the message/rfc822
 *			body part
 *
 * NOTE: this assumes that `parent->length' has been set!
 */

BODY *mutt_parse_messageRFC822 (FILE *fp, BODY *parent)
{
  BODY *msg;

  parent->hdr = mutt_new_header ();
  parent->hdr->offset = ftello (fp);
  parent->hdr->env = mutt_read_rfc822_header (fp, parent->hdr, 0, 0);
  msg = parent->hdr->content;

  /* ignore the length given in the content-length since it could be wrong
     and we already have the info to calculate the correct length */
  /* if (msg->length == -1) */
  msg->length = parent->length - (msg->offset - parent->offset);

  /* if body of this message is empty, we can end up with a negative length */
  if (msg->length < 0)
    msg->length = 0;

  mutt_parse_part(fp, msg);
  return (msg);
}

/* parse a multipart structure
 *
 * args:
 *	fp		stream to read from
 *
 *	boundary	body separator
 *
 *	end_off		length of the multipart body (used when the final
 *			boundary is missing to avoid reading too far)
 *
 *	digest		1 if reading a multipart/digest, 0 otherwise
 */

BODY *mutt_parse_multipart (FILE *fp, const char *boundary, LOFF_T end_off, int digest)
{
#ifdef SUN_ATTACHMENT
  int lines;
#endif
  int blen, len, crlf = 0;
  char buffer[LONG_STRING];
  BODY *head = 0, *last = 0, *new = 0;
  int i;
  int final = 0; /* did we see the ending boundary? */

  if (!boundary)
  {
    mutt_error _("multipart message has no boundary parameter!");
    return (NULL);
  }

  blen = mutt_strlen (boundary);
  while (ftello (fp) < end_off && fgets (buffer, LONG_STRING, fp) != NULL)
  {
    len = mutt_strlen (buffer);

    crlf =  (len > 1 && buffer[len - 2] == '\r') ? 1 : 0;

    if (buffer[0] == '-' && buffer[1] == '-' &&
	mutt_strncmp (buffer + 2, boundary, blen) == 0)
    {
      if (last)
      {
	last->length = ftello (fp) - last->offset - len - 1 - crlf;
	if (last->parts && last->parts->length == 0)
	  last->parts->length = ftello (fp) - last->parts->offset - len - 1 - crlf;
	/* if the body is empty, we can end up with a -1 length */
	if (last->length < 0)
	  last->length = 0;
      }

      /* Remove any trailing whitespace, up to the length of the boundary */
      for (i = len - 1; ISSPACE (buffer[i]) && i >= blen + 2; i--)
        buffer[i] = 0;

      /* Check for the end boundary */
      if (mutt_strcmp (buffer + blen + 2, "--") == 0)
      {
	final = 1;
	break; /* done parsing */
      }
      else if (buffer[2 + blen] == 0)
      {
	new = mutt_read_mime_header (fp, digest);

#ifdef SUN_ATTACHMENT
        if (mutt_get_parameter ("content-lines", new->parameter)) {
	  for (lines = atoi(mutt_get_parameter ("content-lines", new->parameter));
	       lines; lines-- )
	     if (ftello (fp) >= end_off || fgets (buffer, LONG_STRING, fp) == NULL)
	       break;
	}
#endif
	
	/*
	 * Consistency checking - catch
	 * bad attachment end boundaries
	 */
	
	if(new->offset > end_off)
	{
	  mutt_free_body(&new);
	  break;
	}
	if (head)
	{
	  last->next = new;
	  last = new;
	}
	else
	  last = head = new;
      }
    }
  }

  /* in case of missing end boundary, set the length to something reasonable */
  if (last && last->length == 0 && !final)
    last->length = end_off - last->offset;

  /* parse recursive MIME parts */
  for(last = head; last; last = last->next)
    mutt_parse_part(fp, last);
  
  return (head);
}

static const char *uncomment_timezone (char *buf, size_t buflen, const char *tz)
{
  char *p;
  size_t len;

  if (*tz != '(')
    return tz; /* no need to do anything */
  tz++;
  SKIPWS (tz);
  if ((p = strpbrk (tz, " )")) == NULL)
    return tz;
  len = p - tz;
  if (len > buflen - 1)
    len = buflen - 1;
  memcpy (buf, tz, len);
  buf[len] = 0;
  return buf;
}

static struct tz_t
{
  char tzname[5];
  unsigned char zhours;
  unsigned char zminutes;
  unsigned char zoccident; /* west of UTC? */
}
TimeZones[] =
{
  { "aat",   1,  0, 1 }, /* Atlantic Africa Time */
  { "adt",   4,  0, 0 }, /* Arabia DST */
  { "ast",   3,  0, 0 }, /* Arabia */
/*{ "ast",   4,  0, 1 },*/ /* Atlantic */
  { "bst",   1,  0, 0 }, /* British DST */
  { "cat",   1,  0, 0 }, /* Central Africa */
  { "cdt",   5,  0, 1 },
  { "cest",  2,  0, 0 }, /* Central Europe DST */
  { "cet",   1,  0, 0 }, /* Central Europe */
  { "cst",   6,  0, 1 },
/*{ "cst",   8,  0, 0 },*/ /* China */
/*{ "cst",   9, 30, 0 },*/ /* Australian Central Standard Time */
  { "eat",   3,  0, 0 }, /* East Africa */
  { "edt",   4,  0, 1 },
  { "eest",  3,  0, 0 }, /* Eastern Europe DST */
  { "eet",   2,  0, 0 }, /* Eastern Europe */
  { "egst",  0,  0, 0 }, /* Eastern Greenland DST */
  { "egt",   1,  0, 1 }, /* Eastern Greenland */
  { "est",   5,  0, 1 },
  { "gmt",   0,  0, 0 },
  { "gst",   4,  0, 0 }, /* Presian Gulf */
  { "hkt",   8,  0, 0 }, /* Hong Kong */
  { "ict",   7,  0, 0 }, /* Indochina */
  { "idt",   3,  0, 0 }, /* Israel DST */
  { "ist",   2,  0, 0 }, /* Israel */
/*{ "ist",   5, 30, 0 },*/ /* India */
  { "jst",   9,  0, 0 }, /* Japan */
  { "kst",   9,  0, 0 }, /* Korea */
  { "mdt",   6,  0, 1 },
  { "met",   1,  0, 0 }, /* this is now officially CET */
  { "msd",   4,  0, 0 }, /* Moscow DST */
  { "msk",   3,  0, 0 }, /* Moscow */
  { "mst",   7,  0, 1 },
  { "nzdt", 13,  0, 0 }, /* New Zealand DST */
  { "nzst", 12,  0, 0 }, /* New Zealand */
  { "pdt",   7,  0, 1 },
  { "pst",   8,  0, 1 },
  { "sat",   2,  0, 0 }, /* South Africa */
  { "smt",   4,  0, 0 }, /* Seychelles */
  { "sst",  11,  0, 1 }, /* Samoa */
/*{ "sst",   8,  0, 0 },*/ /* Singapore */
  { "utc",   0,  0, 0 },
  { "wat",   0,  0, 0 }, /* West Africa */
  { "west",  1,  0, 0 }, /* Western Europe DST */
  { "wet",   0,  0, 0 }, /* Western Europe */
  { "wgst",  2,  0, 1 }, /* Western Greenland DST */
  { "wgt",   3,  0, 1 }, /* Western Greenland */
  { "wst",   8,  0, 0 }, /* Western Australia */
};

/* parses a date string in RFC822 format:
 *
 * Date: [ weekday , ] day-of-month month year hour:minute:second timezone
 *
 * This routine assumes that `h' has been initialized to 0.  the `timezone'
 * field is optional, defaulting to +0000 if missing.
 */
time_t mutt_parse_date (const char *s, HEADER *h)
{
  int count = 0;
  char *t;
  int hour, min, sec;
  struct tm tm;
  int i;
  int tz_offset = 0;
  int zhours = 0;
  int zminutes = 0;
  int zoccident = 0;
  const char *ptz;
  char tzstr[SHORT_STRING];
  char scratch[SHORT_STRING];

  /* Don't modify our argument. Fixed-size buffer is ok here since
   * the date format imposes a natural limit. 
   */

  strfcpy (scratch, s, sizeof (scratch));
  
  /* kill the day of the week, if it exists. */
  if ((t = strchr (scratch, ',')))
    t++;
  else
    t = scratch;
  SKIPWS (t);

  memset (&tm, 0, sizeof (tm));

  while ((t = strtok (t, " \t")) != NULL)
  {
    switch (count)
    {
      case 0: /* day of the month */
	if (!isdigit ((unsigned char) *t))
	  return (-1);
	tm.tm_mday = atoi (t);
	if (tm.tm_mday > 31)
	  return (-1);
	break;

      case 1: /* month of the year */
	if ((i = mutt_check_month (t)) < 0)
	  return (-1);
	tm.tm_mon = i;
	break;

      case 2: /* year */
	tm.tm_year = atoi (t);
        if (tm.tm_year < 50)
	  tm.tm_year += 100;
        else if (tm.tm_year >= 1900)
	  tm.tm_year -= 1900;
	break;

      case 3: /* time of day */
	if (sscanf (t, "%d:%d:%d", &hour, &min, &sec) == 3)
	  ;
	else if (sscanf (t, "%d:%d", &hour, &min) == 2)
	  sec = 0;
	else
	{
	  dprint(1, (debugfile, "parse_date: could not process time format: %s\n", t));
	  return(-1);
	}
	tm.tm_hour = hour;
	tm.tm_min = min;
	tm.tm_sec = sec;
	break;

      case 4: /* timezone */
	/* sometimes we see things like (MST) or (-0700) so attempt to
	 * compensate by uncommenting the string if non-RFC822 compliant
	 */
	ptz = uncomment_timezone (tzstr, sizeof (tzstr), t);

	if (*ptz == '+' || *ptz == '-')
	{
	  if (ptz[1] && ptz[2] && ptz[3] && ptz[4]
	      && isdigit ((unsigned char) ptz[1]) && isdigit ((unsigned char) ptz[2])
	      && isdigit ((unsigned char) ptz[3]) && isdigit ((unsigned char) ptz[4]))
	  {
	    zhours = (ptz[1] - '0') * 10 + (ptz[2] - '0');
	    zminutes = (ptz[3] - '0') * 10 + (ptz[4] - '0');

	    if (ptz[0] == '-')
	      zoccident = 1;
	  }
	}
	else
	{
	  struct tz_t *tz;

	  tz = bsearch (ptz, TimeZones, sizeof TimeZones/sizeof (struct tz_t),
			sizeof (struct tz_t),
			(int (*)(const void *, const void *)) ascii_strcasecmp
			/* This is safe to do: A pointer to a struct equals
			 * a pointer to its first element*/);

	  if (tz)
	  {
	    zhours = tz->zhours;
	    zminutes = tz->zminutes;
	    zoccident = tz->zoccident;
	  }

	  /* ad hoc support for the European MET (now officially CET) TZ */
	  if (ascii_strcasecmp (t, "MET") == 0)
	  {
	    if ((t = strtok (NULL, " \t")) != NULL)
	    {
	      if (!ascii_strcasecmp (t, "DST"))
		zhours++;
	    }
	  }
	}
	tz_offset = zhours * 3600 + zminutes * 60;
	if (!zoccident)
	  tz_offset = -tz_offset;
	break;
    }
    count++;
    t = 0;
  }

  if (count < 4) /* don't check for missing timezone */
  {
    dprint(1,(debugfile, "parse_date(): error parsing date format, using received time\n"));
    return (-1);
  }

  if (h)
  {
    h->zhours = zhours;
    h->zminutes = zminutes;
    h->zoccident = zoccident;
  }

  return (mutt_mktime (&tm, 0) + tz_offset);
}

/* extract the first substring that looks like a message-id */
static char *extract_message_id (const char *s)
{
  const char *p;
  char *r;
  size_t l;

  if ((s = strchr (s, '<')) == NULL || (p = strchr (s, '>')) == NULL)
    return (NULL);
  l = (size_t)(p - s) + 1;
  r = safe_malloc (l + 1);
  memcpy (r, s, l);
  r[l] = 0;
  return (r);
}

void mutt_parse_mime_message (CONTEXT *ctx, HEADER *cur)
{
  MESSAGE *msg;

  do {
    if (cur->content->type != TYPEMESSAGE &&
        cur->content->type != TYPEMULTIPART)
      break; /* nothing to do */

    if (cur->content->parts)
      break; /* The message was parsed earlier. */

    if ((msg = mx_open_message (ctx, cur->msgno)))
    {
      mutt_parse_part (msg->fp, cur->content);

      if (WithCrypto)
        cur->security = crypt_query (cur->content);

      mx_close_message (&msg);
    }
  } while (0);

  cur->attach_valid = 0;
}

int mutt_parse_rfc822_line (ENVELOPE *e, HEADER *hdr, char *line, char *p, short user_hdrs, short weed,
			    short do_2047, LIST **lastp)
{
  int matched = 0;
  LIST *last = NULL;
  
  if (lastp)
    last = *lastp;
  
  switch (ascii_tolower (line[0]))
  {
    case 'a':
    if (ascii_strcasecmp (line+1, "pparently-to") == 0)
    {
      e->to = rfc822_parse_adrlist (e->to, p);
      matched = 1;
    }
    else if (ascii_strcasecmp (line+1, "pparently-from") == 0)
    {
      e->from = rfc822_parse_adrlist (e->from, p);
      matched = 1;
    }
    break;
    
    case 'b':
    if (ascii_strcasecmp (line+1, "cc") == 0)
    {
      e->bcc = rfc822_parse_adrlist (e->bcc, p);
      matched = 1;
    }
    break;
    
    case 'c':
    if (ascii_strcasecmp (line+1, "c") == 0)
    {
      e->cc = rfc822_parse_adrlist (e->cc, p);
      matched = 1;
    }
    else if (ascii_strncasecmp (line + 1, "ontent-", 7) == 0)
    {
      if (ascii_strcasecmp (line+8, "type") == 0)
      {
	if (hdr)
	  mutt_parse_content_type (p, hdr->content);
	matched = 1;
      }
      else if (ascii_strcasecmp (line+8, "transfer-encoding") == 0)
      {
	if (hdr)
	  hdr->content->encoding = mutt_check_encoding (p);
	matched = 1;
      }
      else if (ascii_strcasecmp (line+8, "length") == 0)
      {
	if (hdr)
	{
	  if ((hdr->content->length = atoi (p)) < 0)
	    hdr->content->length = -1;
	}
	matched = 1;
      }
      else if (ascii_strcasecmp (line+8, "description") == 0)
      {
	if (hdr)
	{
	  mutt_str_replace (&hdr->content->description, p);
	  rfc2047_decode (&hdr->content->description);
	}
	matched = 1;
      }
      else if (ascii_strcasecmp (line+8, "disposition") == 0)
      {
	if (hdr)
	  parse_content_disposition (p, hdr->content);
	matched = 1;
      }
    }
    break;
    
    case 'd':
    if (!ascii_strcasecmp ("ate", line + 1))
    {
      mutt_str_replace (&e->date, p);
      if (hdr)
	hdr->date_sent = mutt_parse_date (p, hdr);
      matched = 1;
    }
    break;
    
    case 'e':
    if (!ascii_strcasecmp ("xpires", line + 1) &&
	hdr && mutt_parse_date (p, NULL) < time (NULL))
      hdr->expired = 1;
    break;
    
    case 'f':
    if (!ascii_strcasecmp ("rom", line + 1))
    {
      e->from = rfc822_parse_adrlist (e->from, p);
      matched = 1;
    }
    break;
    
    case 'i':
    if (!ascii_strcasecmp (line+1, "n-reply-to"))
    {
      mutt_free_list (&e->in_reply_to);
      e->in_reply_to = mutt_parse_references (p, 1);
      matched = 1;
    }
    break;
    
    case 'l':
    if (!ascii_strcasecmp (line + 1, "ines"))
    {
      if (hdr)
      {
	hdr->lines = atoi (p);

	/* 
	 * HACK - mutt has, for a very short time, produced negative
	 * Lines header values.  Ignore them. 
	 */
	if (hdr->lines < 0)
	  hdr->lines = 0;
      }

      matched = 1;
    }
    else if (!ascii_strcasecmp (line + 1, "ist-Post"))
    {