1 /*****************************************************************************
3 *****************************************************************************
4 * Copyright (C) 2003 VideoLAN
5 * $Id: subtitles.c,v 1.4 2003/10/11 22:40:05 hartman Exp $
7 * Authors: Derk-Jan Hartman <hartman at videolan.org>
8 * This is adapted code from the GPL'ed MPlayer (http://mplayerhq.hu)
10 * This program is free software; you can redistribute it and/or modify
11 * it under the terms of the GNU General Public License as published by
12 * the Free Software Foundation; either version 2 of the License, or
13 * (at your option) any later version.
15 * This program is distributed in the hope that it will be useful,
16 * but WITHOUT ANY WARRANTY; without even the implied warranty of
17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 * GNU General Public License for more details.
20 * You should have received a copy of the GNU General Public License
21 * along with this program; if not, write to the Free Software
22 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111, USA.
23 *****************************************************************************/
27 * This file contains functions to dectect subtitle files.
32 #include <vlc/input.h>
39 * What's between a directory and a filename?
42 #define DIRECTORY_SEPARATOR '\\'
44 #define DIRECTORY_SEPARATOR '/'
48 * We are not going to autodetect more subtitle files than this.
50 #define MAX_SUBTITLE_FILES 128
54 * The possible extentions for subtitle files we support
56 static const char * sub_exts[] = { "utf", "utf8", "utf-8", "sub", "srt", "smi", "txt", "ssa", NULL};
57 /* extensions from unsupported types */
58 /* rt, aqt, jss, js, ass */
60 static void strcpy_trim( char *d, char *s )
62 /* skip leading whitespace */
63 while( *s && !isalnum(*s) )
70 while( *s && isalnum(*s) )
76 /* trim excess whitespace */
77 while( *s && !isalnum(*s) )
87 static void strcpy_strip_ext( char *d, char *s )
89 char *tmp = strrchr(s, '.');
96 strncpy(d, s, tmp - s);
106 static void strcpy_get_ext( char *d, char *s )
108 char *tmp = strrchr(s, '.');
113 } else strcpy( d, tmp + 1 );
116 static int whiteonly( char *s )
120 if( isalnum( *s ) ) return 0;
126 typedef struct _subfn
132 static int compare_sub_priority( const void *a, const void *b )
134 if (((subfn*)a)->priority > ((subfn*)b)->priority) {
136 } else if (((subfn*)a)->priority < ((subfn*)b)->priority) {
139 return strcoll(((subfn*)a)->psz_fname, ((subfn*)b)->psz_fname);
144 * Detect subtitle files.
146 * When called this function will split up the psz_fname string into a
147 * directory, filename and extension. It then opens the directory
148 * in which the file resides and tries to find possible matches of
152 * \param p_this the calling \ref input_thread_t
153 * \param psz_path a subdirectory to look into. This is not used atm.
154 * \param psz_fname the complete filename to base the search on.
155 * \return a NULL terminated array of filenames with detected possible subtitles.
156 * The array contains max MAX_SUBTITLE_FILES items and you need to free it after use.
158 char** subtitles_Detect( input_thread_t *p_this, char *psz_path, char *psz_fname )
160 /* variables to be used for derivatives of psz_fname */
161 char *f_dir, *f_fname, *f_fname_noext, *f_fname_trim, *tmp;
162 /* variables to be used for derivatives FILE *f */
163 char *tmp_fname_noext, *tmp_fname_trim, *tmp_fname_ext, *tmpresult;
166 int len, i, j, i_sub_count;
167 subfn *result; /* unsorted results */
168 char **result2; /* sorted results */
175 len = ( strlen( psz_fname ) > 256 ? strlen( psz_fname ) : 256 ) +
176 ( strlen( psz_path ) > 256 ? strlen( psz_path ) : 256 ) + 2;
178 f_dir = (char*)malloc(len);
179 f_fname = (char*)malloc(len);
180 f_fname_noext = (char*)malloc(len);
181 f_fname_trim = (char*)malloc(len);
183 tmp_fname_noext = (char*)malloc(len);
184 tmp_fname_trim = (char*)malloc(len);
185 tmp_fname_ext = (char*)malloc(len);
187 tmpresult = (char*)malloc(len);
189 result = (subfn*)malloc( sizeof(subfn) * MAX_SUBTITLE_FILES );
190 memset( result, 0, sizeof(subfn) * MAX_SUBTITLE_FILES );
192 /* extract filename & dirname from psz_fname */
193 tmp = strrchr( psz_fname, DIRECTORY_SEPARATOR );
197 strcpy( f_fname, tmp + 1 );
198 pos = tmp - psz_fname;
199 strncpy( f_dir, psz_fname, pos + 1 );
204 strcpy( f_fname, psz_fname );
208 strcpy_strip_ext( f_fname_noext, f_fname );
209 strcpy_trim( f_fname_trim, f_fname_noext );
210 var_Get( p_this, "sub-autodetect-fuzzy", &fuzzy );
213 for( j = 0; j <= 1; j++)
215 d = opendir( j == 0 ? f_dir : psz_path );
219 while( ( de = readdir( d ) ) )
221 /* retrieve various parts of the filename */
222 strcpy_strip_ext( tmp_fname_noext, de->d_name );
223 strcpy_get_ext( tmp_fname_ext, de->d_name );
224 strcpy_trim( tmp_fname_trim, tmp_fname_noext );
226 /* does it end with a subtitle extension? */
228 for( i = 0; sub_exts[i]; i++ )
230 if( strcmp(sub_exts[i], tmp_fname_ext ) == 0 )
233 msg_Dbg( p_this, "found a possible subtitle: %s", de->d_name );
238 /* we have a (likely) subtitle file */
242 if( !i_prio && strcmp( tmp_fname_trim, f_fname_trim ) == 0 )
244 /* matches the movie name exactly */
247 if( !i_prio && ( tmp = strstr( tmp_fname_trim, f_fname_trim ) ) )
249 /* contains the movie name */
250 tmp += strlen( f_fname_trim );
251 if( whiteonly( tmp ) )
253 /* chars in front of the movie name */
258 /* chars after (and possibly in front of) the movie name */
264 /* doesn't contain the movie name */
265 if( j == 0 ) i_prio = 1;
268 if( i_prio >= fuzzy.i_int )
270 sprintf( tmpresult, "%s%s", j == 0 ? f_dir : psz_path, de->d_name );
271 msg_Dbg( p_this, "autodetected subtitle: %s with priority %d", de->d_name, i_prio );
272 if( ( f = fopen( tmpresult, "rt" ) ) )
275 result[i_sub_count].priority = i_prio;
276 result[i_sub_count].psz_fname = strdup( tmpresult );
282 if( i_sub_count >= MAX_SUBTITLE_FILES ) break;
290 free( f_fname_noext );
291 free( f_fname_trim );
293 free( tmp_fname_noext );
294 free( tmp_fname_trim );
295 free( tmp_fname_ext );
299 qsort( result, i_sub_count, sizeof( subfn ), compare_sub_priority );
301 result2 = (char**)malloc( sizeof(char*) * ( i_sub_count + 1 ) );
302 memset( result2, 0, sizeof(char*) * ( i_sub_count + 1 ) );
304 for( i = 0; i < i_sub_count; i++ )
306 result2[i] = result[i].psz_fname;
308 result2[i_sub_count] = NULL;