git.maemo.org Git - modest/blob - src/modest-text-utils.c

   1 /* Copyright (c) 2006, Nokia Corporation
   2  * All rights reserved.
   3  *
   4  * Redistribution and use in source and binary forms, with or without
   5  * modification, are permitted provided that the following conditions are
   6  * met:
   7  *
   8  * * Redistributions of source code must retain the above copyright
   9  *   notice, this list of conditions and the following disclaimer.
  10  * * Redistributions in binary form must reproduce the above copyright
  11  *   notice, this list of conditions and the following disclaimer in the
  12  *   documentation and/or other materials provided with the distribution.
  13  * * Neither the name of the Nokia Corporation nor the names of its
  14  *   contributors may be used to endorse or promote products derived from
  15  *   this software without specific prior written permission.
  16  *
  17  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS
  18  * IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
  19  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
  20  * PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
  21  * OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
  22  * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
  23  * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
  24  * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
  25  * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
  26  * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
  27  * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  28  */
  29
  30
  31
  32 #ifndef _GNU_SOURCE
  33 #define _GNU_SOURCE
  34 #endif /*_GNU_SOURCE*/
  35 #include <string.h> /* for strcasestr */
  36
  37
  38 #include <glib.h>
  39 #include <stdlib.h>
  40 #include <glib/gi18n.h>
  41 #include <regex.h>
  42 #include <modest-tny-platform-factory.h>
  43 #include <modest-text-utils.h>
  44 #include <modest-runtime.h>
  45 #include <ctype.h>
  46
  47 #ifdef HAVE_CONFIG_H
  48 #include <config.h>
  49 #endif /*HAVE_CONFIG_H */
  50
  51 /* defines */
  52 #define FORWARD_STRING _("mcen_ia_editor_original_message")
  53 #define FROM_STRING _("mail_va_from")
  54 #define SENT_STRING _("mcen_fi_message_properties_sent")
  55 #define TO_STRING _("mail_va_to")
  56 #define SUBJECT_STRING _("mail_va_subject")
  57 #define EMPTY_STRING ""
  58
  59 /*
  60  * do the hyperlinkification only for texts < 50 Kb,
  61  * as it's quite slow. Without this, e.g. mail with
  62  * an uuencoded part (which is not recognized as attachment,
  63  * will hang modest
  64  */
  65 #define HYPERLINKIFY_MAX_LENGTH (1024*50)
  66
  67 /*
  68  * we need these regexps to find URLs in plain text e-mails
  69  */
  70 typedef struct _url_match_pattern_t url_match_pattern_t;
  71 struct _url_match_pattern_t {
  72         gchar   *regex;
  73         regex_t *preg;
  74         gchar   *prefix;
  75 };
  76
  77 typedef struct _url_match_t url_match_t;
  78 struct _url_match_t {
  79         guint offset;
  80         guint len;
  81         const gchar* prefix;
  82 };
  83
  84
  85 /*
  86  * we mark the ampersand with \007 when converting text->html
  87  * because after text->html we do hyperlink detecting, which
  88  * could be screwed up by the ampersand.
  89  * ie. 1<3 ==> 1\007lt;3
  90  */
  91 #define MARK_AMP '\007'
  92 #define MARK_AMP_STR "\007"
  93
  94 /* mark &amp; separately, because they are parts of urls.
  95  * ie. a&b => a\006amp;b, but a>b => a\007gt;b
  96  *
  97  * we need to handle '&' separately, because it can be part of URIs
  98  * (as in href="http://foo.bar?a=1&b=1"), so inside those URIs
  99  * we need to re-replace \006amp; with '&' again, while outside uri's
 100  * it will be '&amp;'
 101  *
 102  * yes, it's messy, but a consequence of doing text->html first, then hyperlinkify
 103  */
 104 #define MARK_AMP_URI '\006'
 105 #define MARK_AMP_URI_STR "\006"
 106
 107
 108 /* note: match MARK_AMP_URI_STR as well, because after txt->html, a '&' will look like $(MARK_AMP_URI_STR)"amp;" */
 109 #define MAIL_VIEWER_URL_MATCH_PATTERNS  {                               \
 110         { "(feed:|)(file|rtsp|http|ftp|https|mms|mmsh|webcal|feed|rtsp|rdp|lastfm|sip)://[-a-z0-9_$.+!*(),;:@%=\?/~#&" MARK_AMP_URI_STR \
 111                         "]+[-a-z0-9_$%&" MARK_AMP_URI_STR "=?/~#]",     \
 112           NULL, NULL },\
 113         { "www\\.[-a-z0-9_$.+!*(),;:@%=?/~#" MARK_AMP_URI_STR "]+[-a-z0-9_$%" MARK_AMP_URI_STR "=?/~#]",\
 114                         NULL, "http://" },                              \
 115         { "ftp\\.[-a-z0-9_$.+!*(),;:@%=?/~#" MARK_AMP_URI_STR "]+[-a-z0-9_$%" MARK_AMP_URI_STR "=?/~#]",\
 116           NULL, "ftp://" },\
 117         { "(jabberto|voipto|sipto|sip|chatto|skype|xmpp):[-_a-z@0-9.+]+", \
 118            NULL, NULL},                                             \
 119         { "mailto:[-_a-z0-9.\\+]+@[-_a-z0-9.]+",                    \
 120           NULL, NULL},\
 121         { "[-_a-z0-9.\\+]+@[-_a-z0-9.]+",\
 122           NULL, "mailto:"}\
 123         }
 124
 125 const gchar account_title_forbidden_chars[] = {
 126         '\\', '/', ':', '*', '?', '\'', '<', '>', '|', '^'
 127 };
 128 const gchar folder_name_forbidden_chars[] = {
 129         '<', '>', ':', '\'', '/', '\\', '|', '?', '*', '^', '%', '$', '#', '&'
 130 };
 131 const gchar user_name_forbidden_chars[] = {
 132         '<', '>'
 133 };
 134 const guint ACCOUNT_TITLE_FORBIDDEN_CHARS_LENGTH = G_N_ELEMENTS (account_title_forbidden_chars);
 135 const guint FOLDER_NAME_FORBIDDEN_CHARS_LENGTH = G_N_ELEMENTS (folder_name_forbidden_chars);
 136 const guint USER_NAME_FORBIDDEN_CHARS_LENGTH = G_N_ELEMENTS (user_name_forbidden_chars);
 137
 138 /* private */
 139 static gchar*   cite                    (const time_t sent_date, const gchar *from);
 140 static void     hyperlinkify_plain_text (GString *txt, gint offset);
 141 static gint     cmp_offsets_reverse     (const url_match_t *match1, const url_match_t *match2);
 142 static GSList*  get_url_matches         (GString *txt, gint offset);
 143
 144 static GString* get_next_line           (const char *b, const gsize blen, const gchar * iter);
 145 static int      get_indent_level        (const char *l);
 146 static void     unquote_line            (GString * l, const gchar *quote_symbol);
 147 static void     append_quoted           (GString * buf, const gchar *quote_symbol,
 148                                          const int indent, const GString * str,
 149                                          const int cutpoint);
 150 static int      get_breakpoint_utf8     (const gchar * s, const gint indent, const gint limit);
 151 static int      get_breakpoint_ascii    (const gchar * s, const gint indent, const gint limit);
 152 static int      get_breakpoint          (const gchar * s, const gint indent, const gint limit);
 153
 154 static gchar*   modest_text_utils_quote_plain_text (const gchar *text,
 155                                                     const gchar *cite,
 156                                                     const gchar *signature,
 157                                                     GList *attachments,
 158                                                     int limit);
 159
 160 static gchar*   modest_text_utils_quote_html       (const gchar *text,
 161                                                     const gchar *cite,
 162                                                     const gchar *signature,
 163                                                     GList *attachments,
 164                                                     int limit);
 165 static gchar*   get_email_from_address (const gchar *address);
 166
 167
 168 /* ******************************************************************* */
 169 /* ************************* PUBLIC FUNCTIONS ************************ */
 170 /* ******************************************************************* */
 171
 172 gchar *
 173 modest_text_utils_quote (const gchar *text,
 174                          const gchar *content_type,
 175                          const gchar *signature,
 176                          const gchar *from,
 177                          const time_t sent_date,
 178                          GList *attachments,
 179                          int limit)
 180 {
 181         gchar *retval, *cited;
 182
 183         g_return_val_if_fail (text, NULL);
 184         g_return_val_if_fail (content_type, NULL);
 185
 186         cited = cite (sent_date, from);
 187
 188         if (content_type && strcmp (content_type, "text/html") == 0)
 189                 /* TODO: extract the <body> of the HTML and pass it to
 190                    the function */
 191                 retval = modest_text_utils_quote_html (text, cited, signature, attachments, limit);
 192         else
 193                 retval = modest_text_utils_quote_plain_text (text, cited, signature, attachments, limit);
 194
 195         g_free (cited);
 196
 197         return retval;
 198 }
 199
 200
 201 gchar *
 202 modest_text_utils_cite (const gchar *text,
 203                         const gchar *content_type,
 204                         const gchar *signature,
 205                         const gchar *from,
 206                         time_t sent_date)
 207 {
 208         gchar *retval;
 209         gchar *tmp_sig;
 210
 211         g_return_val_if_fail (text, NULL);
 212         g_return_val_if_fail (content_type, NULL);
 213
 214         if (!signature) {
 215                 tmp_sig = g_strdup (text);
 216         } else {
 217                 tmp_sig = g_strconcat (text, "\n", MODEST_TEXT_UTILS_SIGNATURE_MARKER, "\n", signature, NULL);
 218         }
 219
 220         if (strcmp (content_type, "text/html") == 0) {
 221                 retval = modest_text_utils_convert_to_html_body (tmp_sig, -1, TRUE);
 222                 g_free (tmp_sig);
 223         } else {
 224                 retval = tmp_sig;
 225         }
 226
 227         return retval;
 228 }
 229
 230 static gchar *
 231 forward_cite (const gchar *from,
 232               const gchar *sent,
 233               const gchar *to,
 234               const gchar *subject)
 235 {
 236         g_return_val_if_fail (sent, NULL);
 237
 238         return g_strdup_printf ("%s\n%s %s\n%s %s\n%s %s\n%s %s\n",
 239                                 FORWARD_STRING,
 240                                 FROM_STRING, (from)?from:"",
 241                                 SENT_STRING, sent,
 242                                 TO_STRING, (to)?to:"",
 243                                 SUBJECT_STRING, (subject)?subject:"");
 244 }
 245
 246 gchar *
 247 modest_text_utils_inline (const gchar *text,
 248                           const gchar *content_type,
 249                           const gchar *signature,
 250                           const gchar *from,
 251                           time_t sent_date,
 252                           const gchar *to,
 253                           const gchar *subject)
 254 {
 255         gchar sent_str[101];
 256         gchar *cited;
 257         gchar *retval;
 258
 259         g_return_val_if_fail (text, NULL);
 260         g_return_val_if_fail (content_type, NULL);
 261
 262         modest_text_utils_strftime (sent_str, 100, "%c", sent_date);
 263
 264         cited = forward_cite (from, sent_str, to, subject);
 265
 266         if (content_type && strcmp (content_type, "text/html") == 0)
 267                 retval = modest_text_utils_quote_html (text, cited, signature, NULL, 80);
 268         else
 269                 retval = modest_text_utils_quote_plain_text (text, cited, signature, NULL, 80);
 270
 271         g_free (cited);
 272         return retval;
 273 }
 274
 275 /* just to prevent warnings:
 276  * warning: `%x' yields only last 2 digits of year in some locales
 277  */
 278 gsize
 279 modest_text_utils_strftime(char *s, gsize max, const char *fmt, time_t timet)
 280 {
 281         struct tm tm;
 282
 283         /* To prevent possible problems in strftime that could leave
 284            garbage in the s variable */
 285         if (s)
 286                 s[0] = '\0';
 287         else
 288                 return 0;
 289
 290         /* does not work on old maemo glib:
 291          *   g_date_set_time_t (&date, timet);
 292          */
 293         localtime_r (&timet, &tm);
 294         return strftime(s, max, fmt, &tm);
 295 }
 296
 297 gchar *
 298 modest_text_utils_derived_subject (const gchar *subject, const gchar *prefix)
 299 {
 300         gchar *tmp, *subject_dup, *retval;
 301         gint prefix_len;
 302
 303         g_return_val_if_fail (prefix, NULL);
 304
 305         if (!subject || subject[0] == '\0')
 306                 subject = _("mail_va_no_subject");
 307
 308         subject_dup = g_strdup (subject);
 309         tmp = g_strchug (subject_dup);
 310
 311         /* We do not want things like "Re: Re: Re:" or "Fw: Fw:" so
 312            delete the previous ones */
 313         prefix_len = strlen (prefix);
 314         do {
 315                 if (g_str_has_prefix (tmp, prefix)) {
 316                         tmp += prefix_len;
 317                         tmp = g_strchug (tmp);
 318                 } else {
 319                         break;
 320                 }
 321         } while (tmp);
 322
 323         retval = g_strdup_printf ("%s %s", prefix, tmp);
 324         g_free (subject_dup);
 325
 326         return retval;
 327 }
 328
 329 gchar*
 330 modest_text_utils_remove_address (const gchar *address_list, const gchar *address)
 331 {
 332         gchar *dup, *token, *ptr = NULL, *result;
 333         GString *filtered_emails;
 334         gchar *email_address;
 335
 336         g_return_val_if_fail (address_list, NULL);
 337
 338         if (!address)
 339                 return g_strdup (address_list);
 340
 341         email_address = get_email_from_address (address);
 342
 343         /* search for substring */
 344         if (!strstr ((const char *) address_list, (const char *) email_address)) {
 345                 g_free (email_address);
 346                 return g_strdup (address_list);
 347         }
 348
 349         dup = g_strdup (address_list);
 350         filtered_emails = g_string_new (NULL);
 351
 352         token = strtok_r (dup, ",", &ptr);
 353
 354         while (token != NULL) {
 355                 /* Add to list if not found */
 356                 if (!strstr ((const char *) token, (const char *) email_address)) {
 357                         if (filtered_emails->len == 0)
 358                                 g_string_append_printf (filtered_emails, "%s", g_strstrip (token));
 359                         else
 360                                 g_string_append_printf (filtered_emails, ",%s", g_strstrip (token));
 361                 }
 362                 token = strtok_r (NULL, ",", &ptr);
 363         }
 364         result = filtered_emails->str;
 365
 366         /* Clean */
 367         g_free (email_address);
 368         g_free (dup);
 369         g_string_free (filtered_emails, FALSE);
 370
 371         return result;
 372 }
 373
 374
 375 gchar*
 376 modest_text_utils_remove_duplicate_addresses (const gchar *address_list)
 377 {
 378         GSList *addresses, *cursor;
 379         GHashTable *table;
 380         gchar *new_list = NULL;
 381
 382         g_return_val_if_fail (address_list, NULL);
 383
 384         table = g_hash_table_new_full (g_str_hash, g_str_equal, g_free, NULL);
 385         addresses = modest_text_utils_split_addresses_list (address_list);
 386
 387         cursor = addresses;
 388         while (cursor) {
 389                 const gchar* address = (const gchar*)cursor->data;
 390
 391                 /* We need only the email to just compare it and not
 392                    the full address which would make "a <a@a.com>"
 393                    different from "a@a.com" */
 394                 const gchar *email = get_email_from_address (address);
 395
 396                 /* ignore the address if already seen */
 397                 if (g_hash_table_lookup (table, email) == 0) {
 398                         gchar *tmp;
 399
 400                         /* Include the full address and not only the
 401                            email in the returned list */
 402                         if (!new_list) {
 403                                 tmp = g_strdup (address);
 404                         } else {
 405                                 tmp = g_strjoin (",", new_list, address, NULL);
 406                                 g_free (new_list);
 407                         }
 408                         new_list = tmp;
 409
 410                         g_hash_table_insert (table, (gchar*)email, GINT_TO_POINTER(1));
 411                 }
 412                 cursor = g_slist_next (cursor);
 413         }
 414
 415         g_hash_table_unref (table);
 416         g_slist_foreach (addresses, (GFunc)g_free, NULL);
 417         g_slist_free (addresses);
 418
 419         return new_list;
 420 }
 421
 422
 423 static void
 424 modest_text_utils_convert_buffer_to_html_start (GString *html, const gchar *data, gssize n)
 425 {
 426         guint           i;
 427         gboolean        space_seen = FALSE;
 428         guint           break_dist = 0; /* distance since last break point */
 429
 430         if (n == -1)
 431                 n = strlen (data);
 432
 433         /* replace with special html chars where needed*/
 434         for (i = 0; i != n; ++i)  {
 435                 guchar kar = data[i];
 436
 437                 if (space_seen && kar != ' ') {
 438                         g_string_append (html, "&nbsp;");
 439                         space_seen = FALSE;
 440                 }
 441
 442                 /* we artificially insert a breakpoint (newline)
 443                  * after 256, to make sure our lines are not so long
 444                  * they will DOS the regexping later
 445                  * Also, check that kar is ASCII to make sure that we
 446                  * don't break a UTF8 char in two
 447                  */
 448                 if (++break_dist >= 256 && kar < 127) {
 449                         g_string_append_c (html, '\n');
 450                         break_dist = 0;
 451                 }
 452
 453                 switch (kar) {
 454                 case 0:
 455                 case MARK_AMP:
 456                 case MARK_AMP_URI:
 457                         /* this is a temp place holder for '&'; we can only
 458                                 * set the real '&' after hyperlink translation, otherwise
 459                                 * we might screw that up */
 460                         break; /* ignore embedded \0s and MARK_AMP */
 461                 case '<'  : g_string_append (html, MARK_AMP_STR "lt;");   break;
 462                 case '>'  : g_string_append (html, MARK_AMP_STR "gt;");   break;
 463                 case '&'  : g_string_append (html, MARK_AMP_URI_STR "amp;");  break; /* special case */
 464                 case '"'  : g_string_append (html, MARK_AMP_STR "quot;");  break;
 465
 466                 /* don't convert &apos; --> wpeditor will try to re-convert it... */
 467                 //case '\'' : g_string_append (html, "&apos;"); break;
 468                 case '\n' : g_string_append (html, "<br>\n");break_dist= 0; break;
 469                 case '\t' : g_string_append (html, MARK_AMP_STR "nbsp;" MARK_AMP_STR "nbsp;" MARK_AMP_STR "nbsp; ");
 470                         break_dist=0; break; /* note the space at the end*/
 471                 case ' ':
 472                         break_dist = 0;
 473                         if (space_seen) { /* second space in a row */
 474                                 g_string_append (html, "&nbsp; ");
 475                                 space_seen = FALSE;
 476                         } else
 477                                 space_seen = TRUE;
 478                         break;
 479                 default:
 480                         g_string_append_c (html, kar);
 481                 }
 482         }
 483 }
 484
 485
 486 static void
 487 modest_text_utils_convert_buffer_to_html_finish (GString *html)
 488 {
 489         int i;
 490         /* replace all our MARK_AMPs with real ones */
 491         for (i = 0; i != html->len; ++i)
 492                 if ((html->str)[i] == MARK_AMP || (html->str)[i] == MARK_AMP_URI)
 493                         (html->str)[i] = '&';
 494 }
 495
 496
 497 gchar*
 498 modest_text_utils_convert_to_html (const gchar *data)
 499 {
 500         GString         *html;
 501         gsize           len;
 502
 503         g_return_val_if_fail (data, NULL);
 504
 505         if (!data)
 506                 return NULL;
 507
 508         len = strlen (data);
 509         html = g_string_sized_new (1.5 * len);  /* just a  guess... */
 510
 511         g_string_append_printf (html,
 512                                 "<html><head>"
 513                                 "<meta http-equiv=\"content-type\" content=\"text/html; charset=utf8\">"
 514                                 "</head>"
 515                                 "<body>");
 516
 517         modest_text_utils_convert_buffer_to_html_start (html, data, -1);
 518
 519         g_string_append (html, "</body></html>");
 520
 521         if (len <= HYPERLINKIFY_MAX_LENGTH)
 522                 hyperlinkify_plain_text (html, 0);
 523
 524         modest_text_utils_convert_buffer_to_html_finish (html);
 525
 526         return g_string_free (html, FALSE);
 527 }
 528
 529 gchar *
 530 modest_text_utils_convert_to_html_body (const gchar *data, gssize n, gboolean hyperlinkify)
 531 {
 532         GString         *html;
 533
 534         g_return_val_if_fail (data, NULL);
 535
 536         if (!data)
 537                 return NULL;
 538
 539         if (n == -1)
 540                 n = strlen (data);
 541         html = g_string_sized_new (1.5 * n);    /* just a  guess... */
 542
 543         modest_text_utils_convert_buffer_to_html_start (html, data, n);
 544
 545         if (hyperlinkify && (n < HYPERLINKIFY_MAX_LENGTH))
 546                 hyperlinkify_plain_text (html, 0);
 547
 548         modest_text_utils_convert_buffer_to_html_finish (html);
 549
 550         return g_string_free (html, FALSE);
 551 }
 552
 553 void
 554 modest_text_utils_get_addresses_indexes (const gchar *addresses, GSList **start_indexes, GSList **end_indexes)
 555 {
 556         gchar *current, *start, *last_blank;
 557         gint start_offset = 0, current_offset = 0;
 558
 559         g_return_if_fail (start_indexes != NULL);
 560         g_return_if_fail (end_indexes != NULL);
 561
 562         start = (gchar *) addresses;
 563         current = start;
 564         last_blank = start;
 565
 566         while (*current != '\0') {
 567                 if ((start == current)&&((*current == ' ')||(*current == ',')||(*current == ';'))) {
 568                         start = g_utf8_next_char (start);
 569                         start_offset++;
 570                         last_blank = current;
 571                 } else if ((*current == ',')||(*current == ';')) {
 572                         gint *start_index, *end_index;
 573                         start_index = g_new0(gint, 1);
 574                         end_index = g_new0(gint, 1);
 575                         *start_index = start_offset;
 576                         *end_index = current_offset;
 577                         *start_indexes = g_slist_prepend (*start_indexes, start_index);
 578                         *end_indexes = g_slist_prepend (*end_indexes, end_index);
 579                         start = g_utf8_next_char (current);
 580                         start_offset = current_offset + 1;
 581                         last_blank = start;
 582                 } else if (*current == '"') {
 583                         current = g_utf8_next_char (current);
 584                         current_offset ++;
 585                         while ((*current != '"')&&(*current != '\0')) {
 586                                 current = g_utf8_next_char (current);
 587                                 current_offset ++;
 588                         }
 589                 }
 590
 591                 current = g_utf8_next_char (current);
 592                 current_offset ++;
 593         }
 594
 595         if (start != current) {
 596                         gint *start_index, *end_index;
 597                         start_index = g_new0(gint, 1);
 598                         end_index = g_new0(gint, 1);
 599                         *start_index = start_offset;
 600                         *end_index = current_offset;
 601                         *start_indexes = g_slist_prepend (*start_indexes, start_index);
 602                         *end_indexes = g_slist_prepend (*end_indexes, end_index);
 603         }
 604
 605         *start_indexes = g_slist_reverse (*start_indexes);
 606         *end_indexes = g_slist_reverse (*end_indexes);
 607
 608         return;
 609 }
 610
 611
 612 GSList *
 613 modest_text_utils_split_addresses_list (const gchar *addresses)
 614 {
 615         GSList *head;
 616         const gchar *my_addrs = addresses;
 617         const gchar *end;
 618         gchar *addr;
 619         gboolean after_at = FALSE;
 620
 621         g_return_val_if_fail (addresses, NULL);
 622
 623         /* skip any space, ',', ';' at the start */
 624         while (my_addrs && (my_addrs[0] == ' ' || my_addrs[0] == ',' || my_addrs[0] == ';'))
 625                ++my_addrs;
 626
 627         /* are we at the end of addresses list? */
 628         if (!my_addrs[0])
 629                 return NULL;
 630
 631         /* nope, we are at the start of some address
 632          * now, let's find the end of the address */
 633         end = my_addrs + 1;
 634         while (end[0] && end[0] != ';' && !(after_at && end[0] == ',')) {
 635                 if (end[0] == '\"') {
 636                         while (end[0] && end[0] != '\"')
 637                                 ++end;
 638                 }
 639                 if (end[0] == '@') {
 640                         after_at = TRUE;
 641                 }
 642                 if ((end[0] && end[0] == '>')&&(end[1] && end[1] == ',')) {
 643                         ++end;
 644                         break;
 645                 }
 646                 ++end;
 647         }
 648
 649         /* we got the address; copy it and remove trailing whitespace */
 650         addr = g_strndup (my_addrs, end - my_addrs);
 651         g_strchomp (addr);
 652
 653         head = g_slist_append (NULL, addr);
 654         head->next = modest_text_utils_split_addresses_list (end); /* recurse */
 655
 656         return head;
 657 }
 658
 659
 660 void
 661 modest_text_utils_address_range_at_position (const gchar *recipients_list,
 662                                              guint position,
 663                                              guint *start,
 664                                              guint *end)
 665 {
 666         gchar *current = NULL;
 667         gint range_start = 0;
 668         gint range_end = 0;
 669         gint index;
 670         gboolean is_quoted = FALSE;
 671
 672         g_return_if_fail (recipients_list);
 673         g_return_if_fail (position < g_utf8_strlen(recipients_list, -1));
 674
 675         index = 0;
 676         for (current = (gchar *) recipients_list; *current != '\0';
 677              current = g_utf8_find_next_char (current, NULL)) {
 678                 gunichar c = g_utf8_get_char (current);
 679
 680                 if ((c == ',') && (!is_quoted)) {
 681                         if (index < position) {
 682                                 range_start = index + 1;
 683                         } else {
 684                                 break;
 685                         }
 686                 } else if (c == '\"') {
 687                         is_quoted = !is_quoted;
 688                 } else if ((c == ' ') &&(range_start == index)) {
 689                         range_start ++;
 690                 }
 691                 index ++;
 692                 range_end = index;
 693         }
 694
 695         if (start)
 696                 *start = range_start;
 697         if (end)
 698                 *end = range_end;
 699 }
 700
 701 gchar *
 702 modest_text_utils_address_with_standard_length (const gchar *recipients_list)
 703 {
 704         gchar ** splitted;
 705         gchar ** current;
 706         GString *buffer = g_string_new ("");
 707
 708         splitted = g_strsplit (recipients_list, "\n", 0);
 709         current = splitted;
 710         while (*current) {
 711                 gchar *line;
 712                 if (current != splitted)
 713                         buffer = g_string_append_c (buffer, '\n');
 714                 line = g_strndup (*splitted, 1000);
 715                 buffer = g_string_append (buffer, line);
 716                 g_free (line);
 717                 current++;
 718         }
 719
 720         g_strfreev (splitted);
 721
 722         return g_string_free (buffer, FALSE);
 723 }
 724
 725
 726 /* ******************************************************************* */
 727 /* ************************* UTILIY FUNCTIONS ************************ */
 728 /* ******************************************************************* */
 729
 730 static GString *
 731 get_next_line (const gchar * b, const gsize blen, const gchar * iter)
 732 {
 733         GString *gs;
 734         const gchar *i0;
 735
 736         if (iter > b + blen)
 737                 return g_string_new("");
 738
 739         i0 = iter;
 740         while (iter[0]) {
 741                 if (iter[0] == '\n')
 742                         break;
 743                 iter++;
 744         }
 745         gs = g_string_new_len (i0, iter - i0);
 746         return gs;
 747 }
 748 static int
 749 get_indent_level (const char *l)
 750 {
 751         int indent = 0;
 752
 753         while (l[0]) {
 754                 if (l[0] == '>') {
 755                         indent++;
 756                         if (l[1] == ' ') {
 757                                 l++;
 758                         }
 759                 } else {
 760                         break;
 761                 }
 762                 l++;
 763
 764         }
 765
 766         /*      if we hit the signature marker "-- ", we return -(indent + 1). This
 767          *      stops reformatting.
 768          */
 769         if (strcmp (l, MODEST_TEXT_UTILS_SIGNATURE_MARKER) == 0) {
 770                 return -1 - indent;
 771         } else {
 772                 return indent;
 773         }
 774 }
 775
 776 static void
 777 unquote_line (GString * l, const gchar *quote_symbol)
 778 {
 779         gchar *p;
 780         gint quote_len;
 781
 782         p = l->str;
 783         quote_len = strlen (quote_symbol);
 784         while (p[0]) {
 785                 if (g_str_has_prefix (p, quote_symbol)) {
 786                         if (p[quote_len] == ' ') {
 787                                 p += quote_len;
 788                         }
 789                 } else {
 790                         break;
 791                 }
 792                 p++;
 793         }
 794         g_string_erase (l, 0, p - l->str);
 795 }
 796
 797 static void
 798 append_quoted (GString * buf, const gchar *quote_symbol,
 799                int indent, const GString * str,
 800                const int cutpoint)
 801 {
 802         int i;
 803         gchar *quote_concat;
 804
 805         indent = indent < 0 ? abs (indent) - 1 : indent;
 806         quote_concat = g_strconcat (quote_symbol, " ", NULL);
 807         for (i = 0; i <= indent; i++) {
 808                 g_string_append (buf, quote_concat);
 809         }
 810         g_free (quote_concat);
 811         if (cutpoint > 0) {
 812                 g_string_append_len (buf, str->str, cutpoint);
 813         } else {
 814                 g_string_append (buf, str->str);
 815         }
 816         g_string_append (buf, "\n");
 817 }
 818
 819 static int
 820 get_breakpoint_utf8 (const gchar * s, gint indent, const gint limit)
 821 {
 822         gint index = 0;
 823         const gchar *pos, *last;
 824         gunichar *uni;
 825
 826         indent = indent < 0 ? abs (indent) - 1 : indent;
 827
 828         last = NULL;
 829         pos = s;
 830         uni = g_utf8_to_ucs4_fast (s, -1, NULL);
 831         while (pos[0]) {
 832                 if ((index + 2 * indent > limit) && last) {
 833                         g_free (uni);
 834                         return last - s;
 835                 }
 836                 if (g_unichar_isspace (uni[index])) {
 837                         last = pos;
 838                 }
 839                 pos = g_utf8_next_char (pos);
 840                 index++;
 841         }
 842         g_free (uni);
 843         return strlen (s);
 844 }
 845
 846 static int
 847 get_breakpoint_ascii (const gchar * s, const gint indent, const gint limit)
 848 {
 849         gint i, last;
 850
 851         last = strlen (s);
 852         if (last + 2 * indent < limit)
 853                 return last;
 854
 855         for (i = strlen (s); i > 0; i--) {
 856                 if (s[i] == ' ') {
 857                         if (i + 2 * indent <= limit) {
 858                                 return i;
 859                         } else {
 860                                 last = i;
 861                         }
 862                 }
 863         }
 864         return last;
 865 }
 866
 867 static int
 868 get_breakpoint (const gchar * s, const gint indent, const gint limit)
 869 {
 870
 871         if (g_utf8_validate (s, -1, NULL)) {
 872                 return get_breakpoint_utf8 (s, indent, limit);
 873         } else {                /* assume ASCII */
 874                 //g_warning("invalid UTF-8 in msg");
 875                 return get_breakpoint_ascii (s, indent, limit);
 876         }
 877 }
 878
 879 static gchar *
 880 cite (const time_t sent_date, const gchar *from)
 881 {
 882         return g_strdup (_("mcen_ia_editor_original_message"));
 883 }
 884
 885 static gchar *
 886 quoted_attachments (GList *attachments)
 887 {
 888         GList *node = NULL;
 889         GString *result = g_string_new ("");
 890         for (node = attachments; node != NULL; node = g_list_next (node)) {
 891                 gchar *filename = (gchar *) node->data;
 892                 g_string_append_printf ( result, "%s %s\n", _("mcen_ia_editor_attach_filename"), filename);
 893         }
 894
 895         return g_string_free (result, FALSE);
 896
 897 }
 898
 899 static GString *
 900 modest_text_utils_quote_body (GString *output, const gchar *text,
 901                               const gchar *quote_symbol,
 902                               int limit)
 903 {
 904
 905         const gchar *iter;
 906         gsize len;
 907         gint indent, breakpoint, rem_indent = 0;
 908         GString *l, *remaining;
 909
 910         iter = text;
 911         len = strlen(text);
 912         remaining = g_string_new ("");
 913         do {
 914                 l = get_next_line (text, len, iter);
 915                 iter = iter + l->len + 1;
 916                 indent = get_indent_level (l->str);
 917                 unquote_line (l, quote_symbol);
 918
 919                 if (remaining->len) {
 920                         if (l->len && indent == rem_indent) {
 921                                 g_string_prepend (l, " ");
 922                                 g_string_prepend (l, remaining->str);
 923                         } else {
 924                                 do {
 925                                         breakpoint =
 926                                                 get_breakpoint (remaining->str,
 927                                                                 rem_indent,
 928                                                                 limit);
 929                                         append_quoted (output, quote_symbol, rem_indent,
 930                                                        remaining, breakpoint);
 931                                         g_string_erase (remaining, 0,
 932                                                         breakpoint);
 933                                         if (remaining->str[0] == ' ') {
 934                                                 g_string_erase (remaining, 0,
 935                                                                 1);
 936                                         }
 937                                 } while (remaining->len);
 938                         }
 939                 }
 940                 g_string_free (remaining, TRUE);
 941                 breakpoint = get_breakpoint (l->str, indent, limit);
 942                 remaining = g_string_new (l->str + breakpoint);
 943                 if (remaining->str[0] == ' ') {
 944                         g_string_erase (remaining, 0, 1);
 945                 }
 946                 rem_indent = indent;
 947                 append_quoted (output, quote_symbol, indent, l, breakpoint);
 948                 g_string_free (l, TRUE);
 949         } while ((iter < text + len) || (remaining->str[0]));
 950
 951         return output;
 952 }
 953
 954 static gchar *
 955 modest_text_utils_quote_plain_text (const gchar *text,
 956                                     const gchar *cite,
 957                                     const gchar *signature,
 958                                     GList *attachments,
 959                                     int limit)
 960 {
 961         GString *q;
 962         gchar *attachments_string = NULL;
 963
 964         q = g_string_new ("");
 965
 966         if (signature != NULL) {
 967                 g_string_append_printf (q, "\n%s\n", MODEST_TEXT_UTILS_SIGNATURE_MARKER);
 968                 q = g_string_append (q, signature);
 969         }
 970
 971         q = g_string_append (q, "\n");
 972         q = g_string_append (q, cite);
 973         q = g_string_append_c (q, '\n');
 974
 975         q = modest_text_utils_quote_body (q, text, ">", limit);
 976
 977         attachments_string = quoted_attachments (attachments);
 978         q = g_string_append (q, attachments_string);
 979         g_free (attachments_string);
 980
 981         return g_string_free (q, FALSE);
 982 }
 983
 984 static void
 985 quote_html_add_to_gstring (GString *string,
 986                            const gchar *text)
 987 {
 988         if (text && strcmp (text, "")) {
 989                 gchar *html_text = modest_text_utils_convert_to_html_body (text, -1, TRUE);
 990                 g_string_append_printf (string, "%s<br/>", html_text);
 991                 g_free (html_text);
 992         }
 993 }
 994
 995 static gchar*
 996 modest_text_utils_quote_html (const gchar *text,
 997                               const gchar *cite,
 998                               const gchar *signature,
 999                               GList *attachments,
1000                               int limit)
1001 {
1002         GString *result_string;
1003
1004         result_string =
1005                 g_string_new ( \
1006                               "<!DOCTYPE html PUBLIC \"-//W3C//DTD HTML 4.01 Transitional//EN\">\n" \
1007                               "<html>\n"                                \
1008                               "<body>\n<br/>\n");
1009
1010         if (text || cite || signature) {
1011                 GString *quoted_text;
1012                 g_string_append (result_string, "<pre>\n");
1013                 if (signature) {
1014                         quote_html_add_to_gstring (result_string, MODEST_TEXT_UTILS_SIGNATURE_MARKER);
1015                         quote_html_add_to_gstring (result_string, signature);
1016                 }
1017                 quote_html_add_to_gstring (result_string, cite);
1018                 quoted_text = g_string_new ("");
1019                 quoted_text = modest_text_utils_quote_body (quoted_text, (text) ? text : "", ">", limit);
1020                 quote_html_add_to_gstring (result_string, quoted_text->str);
1021                 g_string_free (quoted_text, TRUE);
1022                 if (attachments) {
1023                         gchar *attachments_string = quoted_attachments (attachments);
1024                         quote_html_add_to_gstring (result_string, attachments_string);
1025                         g_free (attachments_string);
1026                 }
1027                 g_string_append (result_string, "</pre>");
1028         }
1029         g_string_append (result_string, "</body>");
1030         g_string_append (result_string, "</html>");
1031
1032         return g_string_free (result_string, FALSE);
1033 }
1034
1035 static gint
1036 cmp_offsets_reverse (const url_match_t *match1, const url_match_t *match2)
1037 {
1038         return match2->offset - match1->offset;
1039 }
1040
1041 static gint url_matches_block = 0;
1042 static url_match_pattern_t patterns[] = MAIL_VIEWER_URL_MATCH_PATTERNS;
1043 static GMutex *url_patterns_mutex = NULL;
1044
1045
1046 static gboolean
1047 compile_patterns ()
1048 {
1049         guint i;
1050         const size_t pattern_num = sizeof(patterns)/sizeof(url_match_pattern_t);
1051         for (i = 0; i != pattern_num; ++i) {
1052                 patterns[i].preg = g_slice_new0 (regex_t);
1053
1054                 /* this should not happen */
1055                 if (regcomp (patterns[i].preg, patterns[i].regex,
1056                              REG_ICASE|REG_EXTENDED|REG_NEWLINE) != 0) {
1057                         g_warning ("%s: error in regexp:\n%s\n", __FUNCTION__, patterns[i].regex);
1058                         return FALSE;
1059                 }
1060         }
1061         return TRUE;
1062 }
1063
1064 static void
1065 free_patterns ()
1066 {
1067         guint i;
1068         const size_t pattern_num = sizeof(patterns)/sizeof(url_match_pattern_t);
1069         for (i = 0; i != pattern_num; ++i) {
1070                 regfree (patterns[i].preg);
1071                 g_slice_free  (regex_t, patterns[i].preg);
1072         } /* don't free patterns itself -- it's static */
1073 }
1074
1075 void
1076 modest_text_utils_hyperlinkify_begin (void)
1077 {
1078
1079         if (url_patterns_mutex == NULL) {
1080                 url_patterns_mutex = g_mutex_new ();
1081         }
1082         g_mutex_lock (url_patterns_mutex);
1083         if (url_matches_block == 0)
1084                 compile_patterns ();
1085         url_matches_block ++;
1086         g_mutex_unlock (url_patterns_mutex);
1087 }
1088
1089 void
1090 modest_text_utils_hyperlinkify_end (void)
1091 {
1092         g_mutex_lock (url_patterns_mutex);
1093         url_matches_block--;
1094         if (url_matches_block <= 0)
1095                 free_patterns ();
1096         g_mutex_unlock (url_patterns_mutex);
1097 }
1098
1099
1100 static GSList*
1101 get_url_matches (GString *txt, gint offset)
1102 {
1103         regmatch_t rm;
1104         guint rv, i, tmp_offset = 0;
1105         GSList *match_list = NULL;
1106
1107         const size_t pattern_num = sizeof(patterns)/sizeof(url_match_pattern_t);
1108
1109         /* initalize the regexps */
1110         modest_text_utils_hyperlinkify_begin ();
1111
1112         /* find all the matches */
1113         for (i = 0; i != pattern_num; ++i) {
1114                 tmp_offset     = offset;
1115                 while (1) {
1116                         url_match_t *match;
1117                         gboolean is_submatch;
1118                         GSList *cursor;
1119
1120                         if ((rv = regexec (patterns[i].preg, txt->str + tmp_offset, 1, &rm, 0)) != 0) {
1121                                 g_return_val_if_fail (rv == REG_NOMATCH, NULL); /* this should not happen */
1122                                 break; /* try next regexp */
1123                         }
1124                         if (rm.rm_so == -1)
1125                                 break;
1126
1127                         is_submatch = FALSE;
1128                         /* check  old matches to see if this has already been matched */
1129                         cursor = match_list;
1130                         while (cursor && !is_submatch) {
1131                                 const url_match_t *old_match =
1132                                         (const url_match_t *) cursor->data;
1133                                 guint new_offset = tmp_offset + rm.rm_so;
1134                                 is_submatch = (new_offset >  old_match->offset &&
1135                                                new_offset <  old_match->offset + old_match->len);
1136                                 cursor = g_slist_next (cursor);
1137                         }
1138
1139                         if (!is_submatch) {
1140                                 /* make a list of our matches (<offset, len, prefix> tupels)*/
1141                                 match = g_slice_new (url_match_t);
1142                                 match->offset = tmp_offset + rm.rm_so;
1143                                 match->len    = rm.rm_eo - rm.rm_so;
1144                                 match->prefix = patterns[i].prefix;
1145                                 match_list = g_slist_prepend (match_list, match);
1146                         }
1147                         tmp_offset += rm.rm_eo;
1148                 }
1149         }
1150
1151         modest_text_utils_hyperlinkify_end ();
1152
1153         /* now sort the list, so the matches are in reverse order of occurence.
1154          * that way, we can do the replacements starting from the end, so we don't need
1155          * to recalculate the offsets
1156          */
1157         match_list = g_slist_sort (match_list,
1158                                    (GCompareFunc)cmp_offsets_reverse);
1159         return match_list;
1160 }
1161
1162
1163
1164 /* replace all occurences of needle in haystack with repl*/
1165 static gchar*
1166 replace_string (const gchar *haystack, const gchar *needle, gchar repl)
1167 {
1168         gchar *str, *cursor;
1169
1170         if (!haystack || !needle || strlen(needle) == 0)
1171                 return haystack ? g_strdup(haystack) : NULL;
1172
1173         str = g_strdup (haystack);
1174
1175         for (cursor = str; cursor && *cursor; ++cursor) {
1176                 if (g_str_has_prefix (cursor, needle)) {
1177                         cursor[0] = repl;
1178                         memmove (cursor + 1,
1179                                  cursor + strlen (needle),
1180                                  strlen (cursor + strlen (needle)) + 1);
1181                 }
1182         }
1183
1184         return str;
1185 }
1186
1187 static void
1188 hyperlinkify_plain_text (GString *txt, gint offset)
1189 {
1190         GSList *cursor;
1191         GSList *match_list = get_url_matches (txt, offset);
1192
1193         /* we will work backwards, so the offsets stay valid */
1194         for (cursor = match_list; cursor; cursor = cursor->next) {
1195
1196                 url_match_t *match = (url_match_t*) cursor->data;
1197                 gchar *url  = g_strndup (txt->str + match->offset, match->len);
1198                 gchar *repl = NULL; /* replacement  */
1199
1200                 /* the string still contains $(MARK_AMP_URI_STR)"amp;" for each
1201                  * '&' in the original, because of the text->html conversion.
1202                  * in the href-URL (and only there), we must convert that back to
1203                  * '&'
1204                  */
1205                 gchar *href_url = replace_string (url, MARK_AMP_URI_STR "amp;", '&');
1206
1207                 /* the prefix is NULL: use the one that is already there */
1208                 repl = g_strdup_printf ("<a href=\"%s%s\">%s</a>",
1209                                         match->prefix ? match->prefix : EMPTY_STRING,
1210                                         href_url, url);
1211
1212                 /* replace the old thing with our hyperlink
1213                  * replacement thing */
1214                 g_string_erase  (txt, match->offset, match->len);
1215                 g_string_insert (txt, match->offset, repl);
1216
1217                 g_free (url);
1218                 g_free (repl);
1219                 g_free (href_url);
1220
1221                 g_slice_free (url_match_t, match);
1222         }
1223
1224         g_slist_free (match_list);
1225 }
1226
1227 void
1228 modest_text_utils_hyperlinkify (GString *string_buffer)
1229 {
1230         gchar *after_body;
1231         gint offset = 0;
1232
1233         after_body = strstr (string_buffer->str, "<body>");
1234         if (after_body != NULL)
1235                 offset = after_body - string_buffer->str;
1236         hyperlinkify_plain_text (string_buffer, offset);
1237 }
1238
1239
1240 /* for optimization reasons, we change the string in-place */
1241 void
1242 modest_text_utils_get_display_address (gchar *address)
1243 {
1244         int i;
1245
1246         g_return_if_fail (address);
1247
1248         if (!address)
1249                 return;
1250
1251         /* should not be needed, and otherwise, we probably won't screw up the address
1252          * more than it already is :)
1253          * g_return_val_if_fail (g_utf8_validate (address, -1, NULL), NULL);
1254          * */
1255
1256         /* remove leading whitespace */
1257         if (address[0] == ' ')
1258                 g_strchug (address);
1259
1260         for (i = 0; address[i]; ++i) {
1261                 if (address[i] == '<') {
1262                         if (G_UNLIKELY(i == 0)) {
1263                                 break; /* there's nothing else, leave it */
1264                         }else {
1265                                 address[i] = '\0'; /* terminate the string here */
1266                                 break;
1267                         }
1268                 }
1269         }
1270
1271         g_strchomp (address);
1272 }
1273
1274
1275 gchar *
1276 modest_text_utils_get_display_addresses (const gchar *recipients)
1277 {
1278         gchar *addresses;
1279         GSList *recipient_list;
1280
1281         addresses = NULL;
1282         recipient_list = modest_text_utils_split_addresses_list (recipients);
1283         if (recipient_list) {
1284                 GString *add_string = g_string_sized_new (strlen (recipients));
1285                 GSList *iter = recipient_list;
1286                 gboolean first = TRUE;
1287
1288                 while (iter) {
1289                         /* Strings are changed in place */
1290                         modest_text_utils_get_display_address ((gchar *) iter->data);
1291                         if (G_UNLIKELY (first)) {
1292                                 g_string_append_printf (add_string, "%s", (gchar *) iter->data);
1293                                 first = FALSE;
1294                         } else {
1295                                 g_string_append_printf (add_string, ", %s", (gchar *) iter->data);
1296                         }
1297                         iter = g_slist_next (iter);
1298                 }
1299                 g_slist_foreach (recipient_list, (GFunc) g_free, NULL);
1300                 g_slist_free (recipient_list);
1301                 addresses = g_string_free (add_string, FALSE);
1302         }
1303
1304         return addresses;
1305 }
1306
1307
1308 gchar *
1309 modest_text_utils_get_email_address (const gchar *full_address)
1310 {
1311         const gchar *left, *right;
1312
1313         g_return_val_if_fail (full_address, NULL);
1314
1315         if (!full_address)
1316                 return NULL;
1317
1318         g_return_val_if_fail (g_utf8_validate (full_address, -1, NULL), NULL);
1319
1320         left = g_strrstr_len (full_address, strlen(full_address), "<");
1321         if (left == NULL)
1322                 return g_strdup (full_address);
1323
1324         right = g_strstr_len (left, strlen(left), ">");
1325         if (right == NULL)
1326                 return g_strdup (full_address);
1327
1328         return g_strndup (left + 1, right - left - 1);
1329 }
1330
1331 gint
1332 modest_text_utils_get_subject_prefix_len (const gchar *sub)
1333 {
1334         gint prefix_len = 0;
1335
1336         g_return_val_if_fail (sub, 0);
1337
1338         if (!sub)
1339                 return 0;
1340
1341         /* optimization: "Re", "RE", "re","Fwd", "FWD", "fwd","FW","Fw", "fw" */
1342         if (sub[0] != 'R' && sub[0] != 'F' && sub[0] != 'r' && sub[0] != 'f')
1343                 return 0;
1344         else if (sub[0] && sub[1] != 'e' && sub[1] != 'E' && sub[1] != 'w' && sub[1] != 'W')
1345                 return 0;
1346
1347         prefix_len = 2;
1348         if (sub[2] == 'd')
1349                 ++prefix_len;
1350
1351         /* skip over a [...] block */
1352         if (sub[prefix_len] == '[') {
1353                 int c = prefix_len + 1;
1354                 while (sub[c] && sub[c] != ']')
1355                         ++c;
1356                 if (!sub[c])
1357                         return 0; /* no end to the ']' found */
1358                 else
1359                         prefix_len = c + 1;
1360         }
1361
1362         /* did we find the ':' ? */
1363         if (sub[prefix_len] == ':') {
1364                 ++prefix_len;
1365                 if (sub[prefix_len] == ' ')
1366                         ++prefix_len;
1367                 prefix_len += modest_text_utils_get_subject_prefix_len (sub + prefix_len);
1368 /*              g_warning ("['%s','%s']", sub, (char*) sub + prefix_len); */
1369                 return prefix_len;
1370         } else
1371                 return 0;
1372 }
1373
1374
1375 gint
1376 modest_text_utils_utf8_strcmp (const gchar* s1, const gchar *s2, gboolean insensitive)
1377 {
1378
1379 /* work even when s1 and/or s2 == NULL */
1380         if (G_UNLIKELY(s1 == s2))
1381                 return 0;
1382         if (G_UNLIKELY(!s1))
1383                 return -1;
1384         if (G_UNLIKELY(!s2))
1385                 return 1;
1386
1387         /* if it's not case sensitive */
1388         if (!insensitive) {
1389
1390                 /* optimization: shortcut if first char is ascii */
1391                 if (((s1[0] & 0x80)== 0) && ((s2[0] & 0x80) == 0) &&
1392                     (s1[0] != s2[0]))
1393                         return s1[0] - s2[0];
1394
1395                 return g_utf8_collate (s1, s2);
1396
1397         } else {
1398                 gint result;
1399                 gchar *n1, *n2;
1400
1401                 /* optimization: shortcut if first char is ascii */
1402                 if (((s1[0] & 0x80) == 0) && ((s2[0] & 0x80) == 0) &&
1403                     (tolower(s1[0]) != tolower (s2[0])))
1404                         return tolower(s1[0]) - tolower(s2[0]);
1405
1406                 n1 = g_utf8_strdown (s1, -1);
1407                 n2 = g_utf8_strdown (s2, -1);
1408
1409                 result = g_utf8_collate (n1, n2);
1410
1411                 g_free (n1);
1412                 g_free (n2);
1413
1414                 return result;
1415         }
1416 }
1417
1418
1419 const gchar*
1420 modest_text_utils_get_display_date (time_t date)
1421 {
1422 #define DATE_BUF_SIZE 64
1423         static gchar date_buf[DATE_BUF_SIZE];
1424
1425         /* calculate the # of days since epoch for
1426          * for today and for the date provided
1427          * based on idea from pvanhoof */
1428         int day      = time(NULL) / (24 * 60 * 60);
1429         int date_day = date       / (24 * 60 * 60);
1430
1431         /* if it's today, show the time, if it's not today, show the date instead */
1432
1433         /* TODO: take into account the system config for 24/12h */
1434 #ifdef MODEST_TOOLKIT_HILDON2
1435         if (day == date_day) /* is the date today? */
1436                 modest_text_utils_strftime (date_buf, DATE_BUF_SIZE, _HL("wdgt_va_24h_time"), date);
1437         else
1438                 modest_text_utils_strftime (date_buf, DATE_BUF_SIZE, _HL("wdgt_va_date"), date);
1439 #else
1440         if (day == date_day) /* is the date today? */
1441                 modest_text_utils_strftime (date_buf, DATE_BUF_SIZE, "%X", date);
1442         else
1443                 modest_text_utils_strftime (date_buf, DATE_BUF_SIZE, "%x", date);
1444 #endif
1445
1446         return date_buf; /* this is a static buffer, don't free! */
1447 }
1448
1449
1450
1451 gboolean
1452 modest_text_utils_validate_folder_name (const gchar *folder_name)
1453 {
1454         /* based on http://msdn2.microsoft.com/en-us/library/aa365247.aspx,
1455          * with some extras */
1456
1457         guint len;
1458         gint i;
1459         const gchar **cursor = NULL;
1460         const gchar *forbidden_names[] = { /* windows does not like these */
1461                 "CON", "PRN", "AUX", "NUL", ".", "..", "cur", "tmp", "new",
1462                 NULL /* cur, tmp, new are reserved for Maildir */
1463         };
1464
1465         /* cannot be NULL */
1466         if (!folder_name)
1467                 return FALSE;
1468
1469         /* cannot be empty */
1470         len = strlen(folder_name);
1471         if (len == 0)
1472                 return FALSE;
1473
1474         /* cannot start with a dot, vfat does not seem to like that */
1475         if (folder_name[0] == '.')
1476                 return FALSE;
1477
1478         /* cannot start or end with a space */
1479         if (g_ascii_isspace(folder_name[0]) || g_ascii_isspace(folder_name[len - 1]))
1480                 return FALSE;
1481
1482         /* cannot contain a forbidden char */
1483         for (i = 0; i < len; i++)
1484                 if (modest_text_utils_is_forbidden_char (folder_name[i], FOLDER_NAME_FORBIDDEN_CHARS))
1485                         return FALSE;
1486
1487         /* Cannot contain Windows port numbers. I'd like to use GRegex
1488            but it's still not available in Maemo. sergio */
1489         if (!g_ascii_strncasecmp (folder_name, "LPT", 3) ||
1490             !g_ascii_strncasecmp (folder_name, "COM", 3)) {
1491                 glong val;
1492                 gchar *endptr;
1493
1494                 /* We skip the first 3 characters for the
1495                    comparison */
1496                 val = strtol(folder_name+3, &endptr, 10);
1497
1498                 /* If the conversion to long succeeded then the string
1499                    is not valid for us */
1500                 if (*endptr == '\0')
1501                         return FALSE;
1502                 else
1503                         return TRUE;
1504         }
1505
1506         /* cannot contain a forbidden word */
1507         if (len <= 4) {
1508                 for (cursor = forbidden_names; cursor && *cursor; ++cursor) {
1509                         if (g_ascii_strcasecmp (folder_name, *cursor) == 0)
1510                                 return FALSE;
1511                 }
1512         }
1513
1514         return TRUE; /* it's valid! */
1515 }
1516
1517
1518
1519 gboolean
1520 modest_text_utils_validate_domain_name (const gchar *domain)
1521 {
1522         gboolean valid = FALSE;
1523         regex_t rx;
1524         const gchar* domain_regex = "^([a-z0-9-]*[a-z0-9]\\.)+[a-z0-9-]*[a-z0-9]$";
1525
1526         g_return_val_if_fail (domain, FALSE);
1527
1528         if (!domain)
1529                 return FALSE;
1530
1531         memset (&rx, 0, sizeof(regex_t)); /* coverity wants this... */
1532
1533         /* domain name: all alphanum or '-' or '.',
1534          * but beginning/ending in alphanum */
1535         if (regcomp (&rx, domain_regex, REG_ICASE|REG_EXTENDED|REG_NOSUB)) {
1536                 g_warning ("BUG: error in regexp");
1537                 return FALSE;
1538         }
1539
1540         valid = (regexec (&rx, domain, 1, NULL, 0) == 0);
1541         regfree (&rx);
1542
1543         return valid;
1544 }
1545
1546
1547
1548 gboolean
1549 modest_text_utils_validate_email_address (const gchar *email_address,
1550                                           const gchar **invalid_char_position)
1551 {
1552         int count = 0;
1553         const gchar *c = NULL, *domain = NULL;
1554         static gchar *rfc822_specials = "()<>@,;:\\\"[]&";
1555
1556         if (invalid_char_position)
1557                 *invalid_char_position = NULL;
1558
1559         g_return_val_if_fail (email_address, FALSE);
1560
1561         /* check that the email adress contains exactly one @ */
1562         if (!strstr(email_address, "@") ||
1563                         (strstr(email_address, "@") != g_strrstr(email_address, "@")))
1564                 return FALSE;
1565
1566         /* first we validate the name portion (name@domain) */
1567         for (c = email_address;  *c;  c++) {
1568                 if (*c == '\"' &&
1569                     (c == email_address ||
1570                      *(c - 1) == '.' ||
1571                      *(c - 1) == '\"')) {
1572                         while (*++c) {
1573                                 if (*c == '\"')
1574                                         break;
1575                                 if (*c == '\\' && (*++c == ' '))
1576                                         continue;
1577                                 if (*c <= ' ' || *c >= 127)
1578                                         return FALSE;
1579                         }
1580                         if (!*c++)
1581                                 return FALSE;
1582                         if (*c == '@')
1583                                 break;
1584                         if (*c != '.')
1585                                 return FALSE;
1586                         continue;
1587                 }
1588                 if (*c == '@')
1589                         break;
1590                 if (*c <= ' ' || *c >= 127)
1591                         return FALSE;
1592                 if (strchr(rfc822_specials, *c)) {
1593                         if (invalid_char_position)
1594                                 *invalid_char_position = c;
1595                         return FALSE;
1596                 }
1597         }
1598         if (c == email_address || *(c - 1) == '.')
1599                 return FALSE;
1600
1601         /* next we validate the domain portion (name@domain) */
1602         if (!*(domain = ++c))
1603                 return FALSE;
1604         do {
1605                 if (*c == '.') {
1606                         if (c == domain || *(c - 1) == '.' || *(c + 1) == '\0')
1607                                 return FALSE;
1608                         count++;
1609                 }
1610                 if (*c <= ' ' || *c >= 127)
1611                         return FALSE;
1612                 if (strchr(rfc822_specials, *c)) {
1613                         if (invalid_char_position)
1614                                 *invalid_char_position = c;
1615                         return FALSE;
1616                 }
1617         } while (*++c);
1618
1619         return (count >= 1) ? TRUE : FALSE;
1620 }
1621
1622 gboolean
1623 modest_text_utils_validate_recipient (const gchar *recipient, const gchar **invalid_char_position)
1624 {
1625         gchar *stripped, *current;
1626         gchar *right_part;
1627         gboolean has_error = FALSE;
1628
1629         if (invalid_char_position)
1630                 *invalid_char_position = NULL;
1631
1632         g_return_val_if_fail (recipient, FALSE);
1633
1634         if (modest_text_utils_validate_email_address (recipient, invalid_char_position))
1635                 return TRUE;
1636
1637         stripped = g_strdup (recipient);
1638         stripped = g_strstrip (stripped);
1639         current = stripped;
1640
1641         if (*current == '\0') {
1642                 g_free (stripped);
1643                 return FALSE;
1644         }
1645
1646         /* quoted string */
1647         if (*current == '\"') {
1648                 current = g_utf8_next_char (current);
1649                 has_error = TRUE;
1650                 for (; *current != '\0'; current = g_utf8_next_char (current)) {
1651                         if (*current == '\\') {
1652                                 /* TODO: This causes a warning, which breaks the build,
1653                                  * because a gchar cannot be < 0.
1654                                  * murrayc.
1655                                 if (current[1] <0) {
1656                                         has_error = TRUE;
1657                                         break;
1658                                 }
1659                                 */
1660                         } else if (*current == '\"') {
1661                                 has_error = FALSE;
1662                                 current = g_utf8_next_char (current);
1663                                 break;
1664                         }
1665                 }
1666         } else {
1667                 has_error = TRUE;
1668                 for (current = stripped ; *current != '\0'; current = g_utf8_next_char (current)) {
1669                         if (*current == '<') {
1670                                 has_error = FALSE;
1671                                 break;
1672                         }
1673                 }
1674         }
1675
1676         if (has_error) {
1677                 g_free (stripped);
1678                 return FALSE;
1679         }
1680
1681         right_part = g_strdup (current);
1682         g_free (stripped);
1683         right_part = g_strstrip (right_part);
1684
1685         if (g_str_has_prefix (right_part, "<") &&
1686             g_str_has_suffix (right_part, ">")) {
1687                 gchar *address;
1688                 gboolean valid;
1689
1690                 address = g_strndup (right_part+1, strlen (right_part) - 2);
1691                 g_free (right_part);
1692                 valid = modest_text_utils_validate_email_address (address, invalid_char_position);
1693                 g_free (address);
1694                 return valid;
1695         } else {
1696                 g_free (right_part);
1697                 return FALSE;
1698         }
1699 }
1700
1701
1702 gchar *
1703 modest_text_utils_get_display_size (guint64 size)
1704 {
1705         const guint KB=1024;
1706         const guint MB=1024 * KB;
1707         const guint GB=1024 * MB;
1708
1709         if (size == 0)
1710                 return g_strdup_printf (_FM("sfil_li_size_kb"), (int) 0);
1711         if (0 <= size && size < KB)
1712                 return g_strdup_printf (_FM("sfil_li_size_1kb_99kb"), (int) 1);
1713         else if (KB <= size && size < 100 * KB)
1714                 return g_strdup_printf (_FM("sfil_li_size_1kb_99kb"), (int) size / KB);
1715         else if (100*KB <= size && size < MB)
1716                 return g_strdup_printf (_FM("sfil_li_size_100kb_1mb"), (int) size / KB);
1717         else if (MB <= size && size < 10*MB)
1718                 return g_strdup_printf (_FM("sfil_li_size_1mb_10mb"), (float) size / MB);
1719         else if (10*MB <= size && size < GB)
1720                 return g_strdup_printf (_FM("sfil_li_size_10mb_1gb"), (float) size / MB);
1721         else
1722                 return g_strdup_printf (_FM("sfil_li_size_1gb_or_greater"), (float) size / GB);
1723 }
1724
1725 static gchar *
1726 get_email_from_address (const gchar * address)
1727 {
1728         gchar *left_limit, *right_limit;
1729
1730         left_limit = strstr (address, "<");
1731         right_limit = g_strrstr (address, ">");
1732
1733         if ((left_limit == NULL)||(right_limit == NULL)|| (left_limit > right_limit))
1734                 return g_strdup (address);
1735         else
1736                 return g_strndup (left_limit + 1, (right_limit - left_limit) - 1);
1737 }
1738
1739 gchar *
1740 modest_text_utils_get_color_string (GdkColor *color)
1741 {
1742         g_return_val_if_fail (color, NULL);
1743
1744         return g_strdup_printf ("#%x%x%x%x%x%x%x%x%x%x%x%x",
1745                                 (color->red >> 12)   & 0xf, (color->red >> 8)   & 0xf,
1746                                 (color->red >>  4)   & 0xf, (color->red)        & 0xf,
1747                                 (color->green >> 12) & 0xf, (color->green >> 8) & 0xf,
1748                                 (color->green >>  4) & 0xf, (color->green)      & 0xf,
1749                                 (color->blue >> 12)  & 0xf, (color->blue >> 8)  & 0xf,
1750                                 (color->blue >>  4)  & 0xf, (color->blue)       & 0xf);
1751 }
1752
1753 gchar *
1754 modest_text_utils_text_buffer_get_text (GtkTextBuffer *buffer)
1755 {
1756         GtkTextIter start, end;
1757         gchar *slice, *current;
1758         GString *result = g_string_new ("");
1759
1760         g_return_val_if_fail (buffer && GTK_IS_TEXT_BUFFER (buffer), NULL);
1761
1762         gtk_text_buffer_get_start_iter (buffer, &start);
1763         gtk_text_buffer_get_end_iter (buffer, &end);
1764
1765         slice = gtk_text_buffer_get_slice (buffer, &start, &end, FALSE);
1766         current = slice;
1767
1768         while (current && current != '\0') {
1769                 if (g_utf8_get_char (current) == 0xFFFC) {
1770                         result = g_string_append_c (result, ' ');
1771                         current = g_utf8_next_char (current);
1772                 } else {
1773                         gchar *next = g_utf8_strchr (current, -1, 0xFFFC);
1774                         if (next == NULL) {
1775                                 result = g_string_append (result, current);
1776                         } else {
1777                                 result = g_string_append_len (result, current, next - current);
1778                         }
1779                         current = next;
1780                 }
1781         }
1782         g_free (slice);
1783
1784         return g_string_free (result, FALSE);
1785
1786 }
1787
1788 gboolean
1789 modest_text_utils_is_forbidden_char (const gchar character,
1790                                      ModestTextUtilsForbiddenCharType type)
1791 {
1792         gint i, len;
1793         const gchar *forbidden_chars = NULL;
1794
1795         /* We need to get the length in the switch because the
1796            compiler needs to know the size at compile time */
1797         switch (type) {
1798         case ACCOUNT_TITLE_FORBIDDEN_CHARS:
1799                 forbidden_chars = account_title_forbidden_chars;
1800                 len = G_N_ELEMENTS (account_title_forbidden_chars);
1801                 break;
1802         case FOLDER_NAME_FORBIDDEN_CHARS:
1803                 forbidden_chars = folder_name_forbidden_chars;
1804                 len = G_N_ELEMENTS (folder_name_forbidden_chars);
1805                 break;
1806         case USER_NAME_FORBIDDEN_NAMES:
1807                 forbidden_chars = user_name_forbidden_chars;
1808                 len = G_N_ELEMENTS (user_name_forbidden_chars);
1809                 break;
1810         default:
1811                 g_return_val_if_reached (TRUE);
1812         }
1813
1814         for (i = 0; i < len ; i++)
1815                 if (forbidden_chars[i] == character)
1816                         return TRUE;
1817
1818         return FALSE; /* it's valid! */
1819 }
1820
1821 gchar *
1822 modest_text_utils_label_get_selection (GtkLabel *label)
1823 {
1824         gint start, end;
1825         gchar *selection;
1826
1827         if (gtk_label_get_selection_bounds (GTK_LABEL (label), &start, &end)) {
1828                 const gchar *start_offset;
1829                 const gchar *end_offset;
1830                 start_offset = gtk_label_get_text (GTK_LABEL (label));
1831                 start_offset = g_utf8_offset_to_pointer (start_offset, start);
1832                 end_offset = gtk_label_get_text (GTK_LABEL (label));
1833                 end_offset = g_utf8_offset_to_pointer (end_offset, end);
1834                 selection = g_strndup (start_offset, end_offset - start_offset);
1835                 return selection;
1836         } else {
1837                 return g_strdup ("");
1838         }
1839 }
1840
1841 static gboolean
1842 _forward_search_image_char (gunichar ch,
1843                             gpointer userdata)
1844 {
1845         return (ch == 0xFFFC);
1846 }
1847
1848 gboolean
1849 modest_text_utils_buffer_selection_is_valid (GtkTextBuffer *buffer)
1850 {
1851         gboolean result;
1852         GtkTextIter start, end;
1853
1854         g_return_val_if_fail (GTK_IS_TEXT_BUFFER (buffer), FALSE);
1855
1856         result = gtk_text_buffer_get_has_selection (GTK_TEXT_BUFFER (buffer));
1857
1858         /* check there are no images in selection */
1859         if (result) {
1860                 gtk_text_buffer_get_selection_bounds (buffer, &start, &end);
1861                 if (gtk_text_iter_get_char (&start)== 0xFFFC)
1862                         result = FALSE;
1863                 else {
1864                         gtk_text_iter_backward_char (&end);
1865                         if (gtk_text_iter_forward_find_char (&start, _forward_search_image_char,
1866                                                              NULL, &end))
1867                                 result = FALSE;
1868                 }
1869
1870         }
1871
1872         return result;
1873 }
1874
1875 static void
1876 remove_quotes (gchar **quotes)
1877 {
1878         if (g_str_has_prefix (*quotes, "\"") && g_str_has_suffix (*quotes, "\"")) {
1879                 gchar *result;
1880                 result = g_strndup ((*quotes)+1, strlen (*quotes) - 2);
1881                 g_free (*quotes);
1882                 *quotes = result;
1883         }
1884 }
1885
1886 gchar *
1887 modest_text_utils_escape_mnemonics (const gchar *text)
1888 {
1889         const gchar *p;
1890         GString *result = NULL;
1891
1892         if (text == NULL)
1893                 return NULL;
1894
1895         result = g_string_new ("");
1896         for (p = text; *p != '\0'; p++) {
1897                 if (*p == '_')
1898                         result = g_string_append (result, "__");
1899                 else
1900                         result = g_string_append_c (result, *p);
1901         }
1902
1903         return g_string_free (result, FALSE);
1904 }
1905
1906 gchar *
1907 modest_text_utils_simplify_recipients (const gchar *recipients)
1908 {
1909         GSList *addresses, *node;
1910         GString *result;
1911         gboolean is_first = TRUE;
1912
1913         if (recipients == NULL)
1914                 return g_strdup ("");
1915
1916         addresses = modest_text_utils_split_addresses_list (recipients);
1917         result = g_string_new ("");
1918
1919         for (node = addresses; node != NULL; node = g_slist_next (node)) {
1920                 const gchar *address = (const gchar *) node->data;
1921                 gchar *left_limit, *right_limit;
1922                 left_limit = strstr (address, "<");
1923                 right_limit = g_strrstr (address, ">");
1924
1925                 if (is_first)
1926                         is_first = FALSE;
1927                 else
1928                         result = g_string_append (result, ", ");
1929
1930                 if ((left_limit == NULL)||(right_limit == NULL)|| (left_limit > right_limit)) {
1931                         result = g_string_append (result, address);
1932                 } else {
1933                         gchar *name_side;
1934                         gchar *email_side;
1935                         name_side = g_strndup (address, left_limit - address);
1936                         name_side = g_strstrip (name_side);
1937                         remove_quotes (&name_side);
1938                         email_side = get_email_from_address (address);
1939                         if (name_side && email_side && !strcmp (name_side, email_side)) {
1940                                 result = g_string_append (result, email_side);
1941                         } else {
1942                                 result = g_string_append (result, address);
1943                         }
1944                         g_free (name_side);
1945                         g_free (email_side);
1946                 }
1947
1948         }
1949         g_slist_foreach (addresses, (GFunc)g_free, NULL);
1950         g_slist_free (addresses);
1951
1952         return g_string_free (result, FALSE);
1953
1954 }