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>
39 # include "../extras/dirent.h"
45 * What's between a directory and a filename?
48 #define DIRECTORY_SEPARATOR '\\'
50 #define DIRECTORY_SEPARATOR '/'
54 * We are not going to autodetect more subtitle files than this.
56 #define MAX_SUBTITLE_FILES 128
60 * The possible extentions for subtitle files we support
62 static const char * sub_exts[] = { "utf", "utf8", "utf-8", "sub", "srt", "smi", "txt", "ssa", NULL};
63 /* extensions from unsupported types */
64 /* rt, aqt, jss, js, ass */
66 static void strcpy_trim( char *d, char *s )
68 /* skip leading whitespace */
69 while( *s && !isalnum(*s) )
76 while( *s && isalnum(*s) )
82 /* trim excess whitespace */
83 while( *s && !isalnum(*s) )
93 static void strcpy_strip_ext( char *d, char *s )
95 char *tmp = strrchr(s, '.');
102 strncpy(d, s, tmp - s);
112 static void strcpy_get_ext( char *d, char *s )
114 char *tmp = strrchr(s, '.');
119 } else strcpy( d, tmp + 1 );
122 static int whiteonly( char *s )
126 if( isalnum( *s ) ) return 0;
132 typedef struct _subfn
138 static int compare_sub_priority( const void *a, const void *b )
140 if (((subfn*)a)->priority > ((subfn*)b)->priority)
145 if (((subfn*)a)->priority < ((subfn*)b)->priority)
150 return strcoll(((subfn*)a)->psz_fname, ((subfn*)b)->psz_fname);
154 * Convert a list of paths separated by ',' to a char**
156 static char **paths_to_list( char *psz_dir, char *psz_path )
158 unsigned int i, k, i_nb_subdirs;
159 char **subdirs; /* list of subdirectories to look in */
162 for( k = 0; k < strlen( psz_path ); k++ )
164 if( psz_path[k] == ',' )
170 if( i_nb_subdirs > 0 )
172 char *psz_parser, *psz_temp;
174 subdirs = (char**)malloc( sizeof(char*) * ( i_nb_subdirs + 1 ) );
175 memset( subdirs, 0, sizeof(char*) * ( i_nb_subdirs + 1 ) );
177 psz_parser = psz_path;
178 while( psz_parser && *psz_parser )
181 psz_subdir = psz_parser;
182 psz_parser = strchr( psz_subdir, ',' );
187 while( *psz_parser == ' ' )
192 if( strlen( psz_subdir ) > 0 )
194 psz_temp = (char *)malloc( strlen(psz_dir)
195 + strlen(psz_subdir) + 2 );
198 sprintf( psz_temp, "%s%s%c",
199 psz_subdir[0] == '.' ? psz_dir : "",
201 psz_subdir[strlen(psz_subdir) - 1] ==
202 DIRECTORY_SEPARATOR ? '\0' : DIRECTORY_SEPARATOR );
203 subdirs[i] = psz_temp;
218 * Detect subtitle files.
220 * When called this function will split up the psz_fname string into a
221 * directory, filename and extension. It then opens the directory
222 * in which the file resides and tries to find possible matches of
226 * \param p_this the calling \ref input_thread_t
227 * \param psz_path a list of subdirectories (separated by a ',') to look in.
228 * \param psz_fname the complete filename to base the search on.
229 * \return a NULL terminated array of filenames with detected possible subtitles.
230 * The array contains max MAX_SUBTITLE_FILES items and you need to free it after use.
232 char **subtitles_Detect( input_thread_t *p_this, char *psz_path,
235 /* variables to be used for derivatives of psz_fname */
236 char *f_dir, *f_fname, *f_fname_noext, *f_fname_trim, *tmp;
237 /* variables to be used for derivatives FILE *f */
238 char *tmp_fname_noext, *tmp_fname_trim, *tmp_fname_ext, *tmpresult;
241 int len, i, j, i_sub_count;
242 subfn *result; /* unsorted results */
243 char **result2; /* sorted results */
244 char **tmp_subdirs, **subdirs; /* list of subdirectories to look in */
251 len = strlen( psz_fname ) > 256 ? strlen( psz_fname ) : 256;
253 f_dir = (char*)malloc(len);
254 f_fname = (char*)malloc(len);
255 f_fname_noext = (char*)malloc(len);
256 f_fname_trim = (char*)malloc(len);
258 tmp_fname_noext = (char*)malloc(len);
259 tmp_fname_trim = (char*)malloc(len);
260 tmp_fname_ext = (char*)malloc(len);
262 tmpresult = (char*)malloc(len);
264 result = (subfn*)malloc( sizeof(subfn) * MAX_SUBTITLE_FILES );
265 memset( result, 0, sizeof(subfn) * MAX_SUBTITLE_FILES );
267 /* extract filename & dirname from psz_fname */
268 tmp = strrchr( psz_fname, DIRECTORY_SEPARATOR );
272 strncpy( f_fname, tmp + 1, len - 1 );
273 f_fname[len - 1] = 0;
274 pos = tmp - psz_fname + 1;
275 strncpy( f_dir, psz_fname, __MIN(pos,len-1) );
276 f_dir[__MIN(pos,len-1)] = 0;
280 strncpy( f_fname, psz_fname, len - 1 );
281 f_fname[len - 1] = 0;
285 strcpy_strip_ext( f_fname_noext, f_fname );
286 strcpy_trim( f_fname_trim, f_fname_noext );
287 var_Get( p_this, "sub-autodetect-fuzzy", &fuzzy );
289 tmp_subdirs = paths_to_list( f_dir, psz_path );
290 subdirs = tmp_subdirs;
292 for( j = -1; j == -1 || ( j >= 0 && subdirs != NULL && *subdirs != NULL );
295 d = opendir( j < 0 ? f_dir : *subdirs );
299 msg_Dbg( p_this, "looking for a subtitle file in %s",
300 j < 0 ? f_dir : *subdirs );
301 while( ( de = readdir( d ) ) )
303 /* retrieve various parts of the filename */
304 strcpy_strip_ext( tmp_fname_noext, de->d_name );
305 strcpy_get_ext( tmp_fname_ext, de->d_name );
306 strcpy_trim( tmp_fname_trim, tmp_fname_noext );
308 /* does it end with a subtitle extension? */
310 for( i = 0; sub_exts[i]; i++ )
312 if( strcmp( sub_exts[i], tmp_fname_ext ) == 0 )
315 msg_Dbg( p_this, "found a possible subtitle: %s",
321 /* we have a (likely) subtitle file */
325 if( !i_prio && !strcmp( tmp_fname_trim, f_fname_trim ) )
327 /* matches the movie name exactly */
331 ( tmp = strstr( tmp_fname_trim, f_fname_trim ) ) )
333 /* contains the movie name */
334 tmp += strlen( f_fname_trim );
335 if( whiteonly( tmp ) )
337 /* chars in front of the movie name */
342 /* chars after (and possibly in front of)
349 /* doesn't contain the movie name */
350 if( j == 0 ) i_prio = 1;
353 if( i_prio >= fuzzy.i_int )
355 sprintf( tmpresult, "%s%s", j == -1 ? f_dir : *subdirs,
357 msg_Dbg( p_this, "autodetected subtitle: %s with "
358 "priority %d", de->d_name, i_prio );
359 if( ( f = fopen( tmpresult, "rt" ) ) )
362 result[i_sub_count].priority = i_prio;
363 result[i_sub_count].psz_fname = strdup(tmpresult);
368 if( i_sub_count >= MAX_SUBTITLE_FILES ) break;
375 if( tmp_subdirs ) free( tmp_subdirs );
379 free( f_fname_noext );
380 free( f_fname_trim );
382 free( tmp_fname_noext );
383 free( tmp_fname_trim );
384 free( tmp_fname_ext );
388 qsort( result, i_sub_count, sizeof( subfn ), compare_sub_priority );
390 result2 = (char**)malloc( sizeof(char*) * ( i_sub_count + 1 ) );
391 memset( result2, 0, sizeof(char*) * ( i_sub_count + 1 ) );
393 for( i = 0; i < i_sub_count; i++ )
395 result2[i] = result[i].psz_fname;
397 result2[i_sub_count] = NULL;