1 /*****************************************************************************
3 *****************************************************************************
4 * Copyright (C) 2003-2004 VideoLAN
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>
37 # include "../extras/dirent.h"
43 * What's between a directory and a filename?
46 #define DIRECTORY_SEPARATOR '\\'
48 #define DIRECTORY_SEPARATOR '/'
52 * We are not going to autodetect more subtitle files than this.
54 #define MAX_SUBTITLE_FILES 128
58 * The possible extentions for subtitle files we support
60 static const char * sub_exts[] = { "utf", "utf8", "utf-8", "sub", "srt", "smi", "txt", "ssa", "idx", NULL};
61 /* extensions from unsupported types */
62 /* rt, aqt, jss, js, ass */
64 static void strcpy_trim( char *d, char *s )
66 /* skip leading whitespace */
67 while( *s && !isalnum(*s) )
74 while( *s && isalnum(*s) )
80 /* trim excess whitespace */
81 while( *s && !isalnum(*s) )
91 static void strcpy_strip_ext( char *d, char *s )
93 char *tmp = strrchr(s, '.');
100 strncpy(d, s, tmp - s);
110 static void strcpy_get_ext( char *d, char *s )
112 char *tmp = strrchr(s, '.');
117 } else strcpy( d, tmp + 1 );
120 static int whiteonly( char *s )
124 if( isalnum( *s ) ) return 0;
130 typedef struct _subfn
136 static int compare_sub_priority( const void *a, const void *b )
138 if (((subfn*)a)->priority > ((subfn*)b)->priority)
143 if (((subfn*)a)->priority < ((subfn*)b)->priority)
148 return strcoll(((subfn*)a)->psz_fname, ((subfn*)b)->psz_fname);
152 * Convert a list of paths separated by ',' to a char**
154 static char **paths_to_list( char *psz_dir, char *psz_path )
156 unsigned int i, k, i_nb_subdirs;
157 char **subdirs; /* list of subdirectories to look in */
160 for( k = 0; k < strlen( psz_path ); k++ )
162 if( psz_path[k] == ',' )
168 if( i_nb_subdirs > 0 )
170 char *psz_parser, *psz_temp;
172 subdirs = (char**)malloc( sizeof(char*) * ( i_nb_subdirs + 1 ) );
173 memset( subdirs, 0, sizeof(char*) * ( i_nb_subdirs + 1 ) );
175 psz_parser = psz_path;
176 while( psz_parser && *psz_parser )
179 psz_subdir = psz_parser;
180 psz_parser = strchr( psz_subdir, ',' );
185 while( *psz_parser == ' ' )
190 if( strlen( psz_subdir ) > 0 )
192 psz_temp = (char *)malloc( strlen(psz_dir)
193 + strlen(psz_subdir) + 2 );
196 sprintf( psz_temp, "%s%s%c",
197 psz_subdir[0] == '.' ? psz_dir : "",
199 psz_subdir[strlen(psz_subdir) - 1] ==
200 DIRECTORY_SEPARATOR ? '\0' : DIRECTORY_SEPARATOR );
201 subdirs[i] = psz_temp;
216 * Detect subtitle files.
218 * When called this function will split up the psz_fname string into a
219 * directory, filename and extension. It then opens the directory
220 * in which the file resides and tries to find possible matches of
224 * \param p_this the calling \ref input_thread_t
225 * \param psz_path a list of subdirectories (separated by a ',') to look in.
226 * \param psz_fname the complete filename to base the search on.
227 * \return a NULL terminated array of filenames with detected possible subtitles.
228 * The array contains max MAX_SUBTITLE_FILES items and you need to free it after use.
230 char **subtitles_Detect( input_thread_t *p_this, char *psz_path,
233 /* variables to be used for derivatives of psz_fname */
234 char *f_dir, *f_fname, *f_fname_noext, *f_fname_trim, *tmp;
235 /* variables to be used for derivatives FILE *f */
236 char *tmp_fname_noext, *tmp_fname_trim, *tmp_fname_ext, *tmpresult;
238 int len, i, j, i_sub_count;
239 subfn *result; /* unsorted results */
240 char **result2; /* sorted results */
241 char **tmp_subdirs, **subdirs; /* list of subdirectories to look in */
247 if( !strncmp( psz_fname, "file://", 7 ) )
253 len = strlen( psz_fname ) > 256 ? strlen( psz_fname ) : 256;
255 f_dir = (char*)malloc(len);
256 f_fname = (char*)malloc(len);
257 f_fname_noext = (char*)malloc(len);
258 f_fname_trim = (char*)malloc(len);
260 tmp_fname_noext = (char*)malloc(len);
261 tmp_fname_trim = (char*)malloc(len);
262 tmp_fname_ext = (char*)malloc(len);
264 tmpresult = (char*)malloc(len);
266 result = (subfn*)malloc( sizeof(subfn) * MAX_SUBTITLE_FILES );
267 memset( result, 0, sizeof(subfn) * MAX_SUBTITLE_FILES );
269 /* extract filename & dirname from psz_fname */
270 tmp = strrchr( psz_fname, DIRECTORY_SEPARATOR );
274 strncpy( f_fname, tmp + 1, len - 1 );
275 f_fname[len - 1] = 0;
276 pos = tmp - psz_fname + 1;
277 strncpy( f_dir, psz_fname, __MIN(pos,len-1) );
278 f_dir[__MIN(pos,len-1)] = 0;
282 strncpy( f_fname, psz_fname, len - 1 );
283 f_fname[len - 1] = 0;
287 strcpy_strip_ext( f_fname_noext, f_fname );
288 strcpy_trim( f_fname_trim, f_fname_noext );
289 var_Get( p_this, "sub-autodetect-fuzzy", &fuzzy );
291 tmp_subdirs = paths_to_list( f_dir, psz_path );
292 subdirs = tmp_subdirs;
294 for( j = -1; j == -1 || ( j >= 0 && subdirs != NULL && *subdirs != NULL );
297 d = opendir( j < 0 ? f_dir : *subdirs );
301 msg_Dbg( p_this, "looking for a subtitle file in %s",
302 j < 0 ? f_dir : *subdirs );
303 while( ( de = readdir( d ) ) )
305 /* retrieve various parts of the filename */
306 strcpy_strip_ext( tmp_fname_noext, de->d_name );
307 strcpy_get_ext( tmp_fname_ext, de->d_name );
308 strcpy_trim( tmp_fname_trim, tmp_fname_noext );
310 /* does it end with a subtitle extension? */
312 for( i = 0; sub_exts[i]; i++ )
314 if( strcmp( sub_exts[i], tmp_fname_ext ) == 0 )
317 msg_Dbg( p_this, "found a possible subtitle: %s",
323 /* we have a (likely) subtitle file */
327 if( !i_prio && !strcmp( tmp_fname_trim, f_fname_trim ) )
329 /* matches the movie name exactly */
333 ( tmp = strstr( tmp_fname_trim, f_fname_trim ) ) )
335 /* contains the movie name */
336 tmp += strlen( f_fname_trim );
337 if( whiteonly( tmp ) )
339 /* chars in front of the movie name */
344 /* chars after (and possibly in front of)
351 /* doesn't contain the movie name */
352 if( j == 0 ) i_prio = 1;
355 if( i_prio >= fuzzy.i_int )
357 sprintf( tmpresult, "%s%s", j == -1 ? f_dir : *subdirs,
359 msg_Dbg( p_this, "autodetected subtitle: %s with "
360 "priority %d", de->d_name, i_prio );
361 if( ( f = fopen( tmpresult, "rt" ) ) )
364 result[i_sub_count].priority = i_prio;
365 result[i_sub_count].psz_fname = strdup(tmpresult);
370 if( i_sub_count >= MAX_SUBTITLE_FILES ) break;
374 if( j >= 0 ) free( *subdirs++ );
377 if( tmp_subdirs ) free( tmp_subdirs );
381 free( f_fname_noext );
382 free( f_fname_trim );
384 free( tmp_fname_noext );
385 free( tmp_fname_trim );
386 free( tmp_fname_ext );
390 qsort( result, i_sub_count, sizeof( subfn ), compare_sub_priority );
392 result2 = (char**)malloc( sizeof(char*) * ( i_sub_count + 1 ) );
393 memset( result2, 0, sizeof(char*) * ( i_sub_count + 1 ) );
395 for( i = 0; i < i_sub_count; i++ )
397 result2[i] = result[i].psz_fname;
399 result2[i_sub_count] = NULL;