libempathy-gtk/empathy-string-parser.c

   1 /*
   2  * Copyright (C) 2010 Collabora Ltd.
   3  *
   4  * This library is free software; you can redistribute it and/or
   5  * modify it under the terms of the GNU Lesser General Public
   6  * License as published by the Free Software Foundation; either
   7  * version 2.1 of the License, or (at your option) any later version.
   8  *
   9  * This library is distributed in the hope that it will be useful,
  10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  12  * Lesser General Public License for more details.
  13  *
  14  * You should have received a copy of the GNU Lesser General Public
  15  * License along with this library; if not, write to the Free Software
  16  * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
  17  *
  18  * Authors: Xavier Claessens <xclaesse@gmail.com>
  19  */
  20
  21 #include <config.h>
  22
  23 #include <string.h>
  24
  25 #include "empathy-string-parser.h"
  26 #include "empathy-smiley-manager.h"
  27 #include "empathy-ui-utils.h"
  28
  29 #define SCHEMES           "([a-zA-Z\\+]+)"
  30 #define INVALID_CHARS     "\\s\"'"
  31 #define INVALID_CHARS_EXT INVALID_CHARS "\\[\\]<>(){},;:?"
  32 #define BODY              "([^"INVALID_CHARS"]+)"
  33 #define BODY_END          "([^"INVALID_CHARS"]*)[^"INVALID_CHARS_EXT".]"
  34 #define BODY_STRICT       "([^"INVALID_CHARS_EXT"]+)"
  35 #define URI_REGEX         "("SCHEMES"://"BODY_END")" \
  36                           "|((www|ftp)\\."BODY_END")" \
  37                           "|((mailto:)?"BODY_STRICT"@"BODY"\\."BODY_END")"
  38
  39 static GRegex *
  40 uri_regex_dup_singleton (void)
  41 {
  42         static GRegex *uri_regex = NULL;
  43
  44         /* We intentionally leak the regex so it's not recomputed */
  45         if (!uri_regex) {
  46                 uri_regex = g_regex_new (URI_REGEX, 0, 0, NULL);
  47         }
  48
  49         return g_regex_ref (uri_regex);
  50 }
  51
  52 void
  53 empathy_string_parser_substr (const gchar *text,
  54                               gssize len,
  55                               EmpathyStringParser *parsers,
  56                               gpointer user_data)
  57 {
  58         if (parsers != NULL && parsers[0].match_func != NULL) {
  59                 parsers[0].match_func (text, len,
  60                                        parsers[0].replace_func, parsers + 1,
  61                                        user_data);
  62         }
  63 }
  64
  65 void
  66 empathy_string_match_link (const gchar *text,
  67                            gssize len,
  68                            EmpathyStringReplace replace_func,
  69                            EmpathyStringParser *sub_parsers,
  70                            gpointer user_data)
  71 {
  72         GRegex     *uri_regex;
  73         GMatchInfo *match_info;
  74         gboolean    match;
  75         gint        last = 0;
  76
  77         uri_regex = uri_regex_dup_singleton ();
  78         match = g_regex_match_full (uri_regex, text, len, 0, 0, &match_info, NULL);
  79         if (match) {
  80                 gint s = 0, e = 0;
  81
  82                 do {
  83                         g_match_info_fetch_pos (match_info, 0, &s, &e);
  84
  85                         if (s > last) {
  86                                 /* Append the text between last link (or the
  87                                  * start of the message) and this link */
  88                                 empathy_string_parser_substr (text + last,
  89                                                               s - last,
  90                                                               sub_parsers,
  91                                                               user_data);
  92                         }
  93
  94                         replace_func (text + s, e - s, NULL, user_data);
  95
  96                         last = e;
  97                 } while (g_match_info_next (match_info, NULL));
  98         }
  99
 100         empathy_string_parser_substr (text + last, len - last,
 101                                       sub_parsers, user_data);
 102
 103         g_match_info_free (match_info);
 104         g_regex_unref (uri_regex);
 105 }
 106
 107 void
 108 empathy_string_match_smiley (const gchar *text,
 109                              gssize len,
 110                              EmpathyStringReplace replace_func,
 111                              EmpathyStringParser *sub_parsers,
 112                              gpointer user_data)
 113 {
 114         guint last = 0;
 115         EmpathySmileyManager *smiley_manager;
 116         GSList *hits, *l;
 117
 118         smiley_manager = empathy_smiley_manager_dup_singleton ();
 119         hits = empathy_smiley_manager_parse_len (smiley_manager, text, len);
 120
 121         for (l = hits; l; l = l->next) {
 122                 EmpathySmileyHit *hit = l->data;
 123
 124                 if (hit->start > last) {
 125                         /* Append the text between last smiley (or the
 126                          * start of the message) and this smiley */
 127                         empathy_string_parser_substr (text + last,
 128                                                       hit->start - last,
 129                                                       sub_parsers, user_data);
 130                 }
 131
 132                 replace_func (text + hit->start, hit->end - hit->start,
 133                               hit, user_data);
 134
 135                 last = hit->end;
 136
 137                 empathy_smiley_hit_free (hit);
 138         }
 139         g_slist_free (hits);
 140         g_object_unref (smiley_manager);
 141
 142         empathy_string_parser_substr (text + last, len - last,
 143                                       sub_parsers, user_data);
 144 }
 145
 146 void
 147 empathy_string_match_all (const gchar *text,
 148                           gssize len,
 149                           EmpathyStringReplace replace_func,
 150                           EmpathyStringParser *sub_parsers,
 151                           gpointer user_data)
 152 {
 153         replace_func (text, len, NULL, user_data);
 154 }
 155
 156 void
 157 empathy_string_replace_link (const gchar *text,
 158                              gssize len,
 159                              gpointer match_data,
 160                              gpointer user_data)
 161 {
 162         GString *string = user_data;
 163         gchar *real_url;
 164         gchar *escaped;
 165
 166         real_url = empathy_make_absolute_url_len (text, len);
 167
 168         /* The thing we are making a link of may contain
 169          * characters which need escaping */
 170         escaped = g_markup_escape_text (text, len);
 171
 172         /* Append the link inside <a href=""></a> tag */
 173         g_string_append_printf (string, "<a href=\"%s\">%s</a>",
 174                                 real_url, escaped);
 175
 176         g_free (real_url);
 177         g_free (escaped);
 178 }
 179
 180 void
 181 empathy_string_replace_escaped (const gchar *text,
 182                                 gssize len,
 183                                 gpointer match_data,
 184                                 gpointer user_data)
 185 {
 186         GString *string = user_data;
 187         gchar *escaped;
 188         guint i;
 189         gssize escaped_len, old_len;
 190
 191         escaped = g_markup_escape_text (text, len);
 192         escaped_len = strlen (escaped);
 193
 194         /* Allocate more space to string (we really need a g_string_extend...) */
 195         old_len = string->len;
 196         g_string_set_size (string, old_len + escaped_len);
 197         g_string_truncate (string, old_len);
 198
 199         /* Remove '\r' */
 200         for (i = 0; i < escaped_len; i++) {
 201                 if (escaped[i] != '\r')
 202                         g_string_append_c (string, escaped[i]);
 203         }
 204
 205         g_free (escaped);
 206 }
 207
 208 gchar *
 209 empathy_add_link_markup (const gchar *text)
 210 {
 211         EmpathyStringParser parsers[] = {
 212                 {empathy_string_match_link, empathy_string_replace_link},
 213                 {empathy_string_match_all, empathy_string_replace_escaped},
 214                 {NULL, NULL}
 215         };
 216         GString *string;
 217
 218         g_return_val_if_fail (text != NULL, NULL);
 219
 220         /* GtkLabel with links could make infinite loop because of
 221          * GNOME bug #612066. It is fixed in GTK >= 2.18.8 and GTK >= 2.19.7.
 222          * FIXME: Remove this check once we have an hard dep on GTK 2.20 */
 223         if (gtk_check_version (2, 18, 8) != NULL ||
 224             (gtk_minor_version == 19 && gtk_micro_version < 7)) {
 225                 return g_markup_escape_text (text, -1);
 226         }
 227
 228         string = g_string_sized_new (strlen (text));
 229         empathy_string_parser_substr (text, -1, parsers, string);
 230
 231         return g_string_free (string, FALSE);
 232 }
 233