src/unmime.c

   1 /*
   2  * Sylpheed -- a GTK+ based, lightweight, and fast e-mail client
   3  * Copyright (C) 1999-2011 Hiroyuki Yamamoto and the Claws Mail team
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 3 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License
  16  * along with this program. If not, see <http://www.gnu.org/licenses/>.
  17  *
  18  */
  19
  20 #ifdef HAVE_CONFIG_H
  21 #  include "config.h"
  22 #endif
  23
  24 #include <glib.h>
  25 #include <string.h>
  26 #include <ctype.h>
  27
  28 #include "codeconv.h"
  29 #include "base64.h"
  30 #include "quoted-printable.h"
  31
  32 #define ENCODED_WORD_BEGIN      "=?"
  33 #define ENCODED_WORD_END        "?="
  34
  35 /* Decodes headers based on RFC2045 and RFC2047. */
  36
  37 gchar *unmime_header(const gchar *encoded_str, gboolean addr_field)
  38 {
  39         const gchar *p = encoded_str;
  40         const gchar *eword_begin_p, *encoding_begin_p, *text_begin_p,
  41                     *eword_end_p;
  42         gchar charset[32];
  43         gchar encoding;
  44         gchar *conv_str;
  45         GString *outbuf;
  46         gchar *out_str;
  47         gsize out_len;
  48         int in_quote = FALSE;
  49
  50         outbuf = g_string_sized_new(strlen(encoded_str) * 2);
  51
  52         while (*p != '\0') {
  53                 gchar *decoded_text = NULL;
  54                 const gchar *quote_p;
  55                 gint len;
  56
  57                 eword_begin_p = strstr(p, ENCODED_WORD_BEGIN);
  58                 if (!eword_begin_p) {
  59                         g_string_append(outbuf, p);
  60                         break;
  61                 }
  62
  63                 quote_p = p;
  64                 while ((quote_p = strchr(quote_p, '"')) != NULL) {
  65                         if (quote_p && quote_p < eword_begin_p) {
  66                                 /* Found a quote before the encoded word. */
  67                                 in_quote = !in_quote;
  68                                 quote_p++;
  69                         }
  70                         if (quote_p >= eword_begin_p)
  71                                 break;
  72                 }
  73
  74                 encoding_begin_p = strchr(eword_begin_p + 2, '?');
  75                 if (!encoding_begin_p) {
  76                         g_string_append(outbuf, p);
  77                         break;
  78                 }
  79                 text_begin_p = strchr(encoding_begin_p + 1, '?');
  80                 if (!text_begin_p) {
  81                         g_string_append(outbuf, p);
  82                         break;
  83                 }
  84                 eword_end_p = strstr(text_begin_p + 1, ENCODED_WORD_END);
  85                 if (!eword_end_p) {
  86                         g_string_append(outbuf, p);
  87                         break;
  88                 }
  89
  90                 if (p == encoded_str) {
  91                         g_string_append_len(outbuf, p, eword_begin_p - p);
  92                         p = eword_begin_p;
  93                 } else {
  94                         /* ignore spaces between encoded words */
  95                         const gchar *sp;
  96
  97                         for (sp = p; sp < eword_begin_p; sp++) {
  98                                 if (!g_ascii_isspace(*sp)) {
  99                                         g_string_append_len
 100                                                 (outbuf, p, eword_begin_p - p);
 101                                         p = eword_begin_p;
 102                                         break;
 103                                 }
 104                         }
 105                 }
 106
 107                 len = MIN(sizeof(charset) - 1,
 108                           encoding_begin_p - (eword_begin_p + 2));
 109                 memcpy(charset, eword_begin_p + 2, len);
 110                 charset[len] = '\0';
 111                 encoding = g_ascii_toupper(*(encoding_begin_p + 1));
 112
 113                 if (encoding == 'B') {
 114                         decoded_text = g_malloc
 115                                 (eword_end_p - (text_begin_p + 1) + 1);
 116                         len = base64_decode(decoded_text, text_begin_p + 1,
 117                                             eword_end_p - (text_begin_p + 1));
 118                         decoded_text[len] = '\0';
 119                 } else if (encoding == 'Q') {
 120                         decoded_text = g_malloc
 121                                 (eword_end_p - (text_begin_p + 1) + 1);
 122                         len = qp_decode_q_encoding
 123                                 (decoded_text, text_begin_p + 1,
 124                                  eword_end_p - (text_begin_p + 1));
 125                 } else {
 126                         g_string_append_len(outbuf, p, eword_end_p + 2 - p);
 127                         p = eword_end_p + 2;
 128                         continue;
 129                 }
 130
 131                 /* An encoded word MUST not appear within a quoted string,
 132                  * so quoting that word after decoding should be safe.
 133                  * We check there are no quotes just to be sure. If there
 134                  * are, well, the comma won't pose a problem, probably.
 135                  */
 136                 if (addr_field && strchr(decoded_text, ',') && !in_quote) {
 137                         gchar *tmp = g_strdup_printf("\"%s\"", decoded_text);
 138                         g_free(decoded_text);
 139                         decoded_text = tmp;
 140                 }
 141
 142                 /* convert to UTF-8 */
 143                 conv_str = conv_codeset_strdup(decoded_text, charset, NULL);
 144                 if (!conv_str || !g_utf8_validate(conv_str, -1, NULL)) {
 145                         g_free(conv_str);
 146                         conv_str = g_malloc(len + 1);
 147                         conv_utf8todisp(conv_str, len + 1, decoded_text);
 148                 }
 149                 g_string_append(outbuf, conv_str);
 150                 g_free(conv_str);
 151
 152                 g_free(decoded_text);
 153
 154                 p = eword_end_p + 2;
 155         }
 156
 157         out_str = outbuf->str;
 158         out_len = outbuf->len;
 159         g_string_free(outbuf, FALSE);
 160
 161         return g_realloc(out_str, out_len + 1);
 162 }