1 /*****************************************************************************
3 *****************************************************************************
4 * Copyright (C) 2003-2006 the VideoLAN team
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., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301, USA.
23 *****************************************************************************/
27 * This file contains functions to dectect subtitle files.
32 #include <vlc/input.h>
49 #include "input_internal.h"
52 * We are not going to autodetect more subtitle files than this.
54 #define MAX_SUBTITLE_FILES 128
58 * The possible extensions 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 strlcpy(d, s, tmp - s + 1 );
108 static void strcpy_get_ext( char *d, char *s )
110 char *tmp = strrchr(s, '.');
115 } else strcpy( d, tmp + 1 );
118 static int whiteonly( char *s )
122 if( isalnum( *s ) ) return 0;
128 typedef struct _subfn
135 static int compare_sub_priority( const void *a, const void *b )
137 if (((subfn*)a)->priority > ((subfn*)b)->priority)
142 if (((subfn*)a)->priority < ((subfn*)b)->priority)
148 return strcoll(((subfn*)a)->psz_fname, ((subfn*)b)->psz_fname);
150 return strcmp(((subfn*)a)->psz_fname, ((subfn*)b)->psz_fname);
155 * Check if a file ends with a subtitle extension
157 int subtitles_Filter( const char *psz_dir_content )
159 const char *tmp = strrchr( psz_dir_content, '.');
167 for( i = 0; sub_exts[i]; i++ )
168 if( strcmp( sub_exts[i], tmp ) == 0 )
176 * Convert a list of paths separated by ',' to a char**
178 static char **paths_to_list( char *psz_dir, char *psz_path )
180 unsigned int i, k, i_nb_subdirs;
181 char **subdirs; /* list of subdirectories to look in */
183 if( !psz_dir ) return NULL;
184 if( !psz_path ) return NULL;
187 for( k = 0; k < strlen( psz_path ); k++ )
189 if( psz_path[k] == ',' )
195 if( i_nb_subdirs > 0 )
197 char *psz_parser = NULL, *psz_temp = NULL;
199 subdirs = (char**)malloc( sizeof(char*) * ( i_nb_subdirs + 1 ) );
200 memset( subdirs, 0, sizeof(char*) * ( i_nb_subdirs + 1 ) );
202 psz_parser = psz_path;
203 while( psz_parser && *psz_parser )
206 psz_subdir = psz_parser;
207 psz_parser = strchr( psz_subdir, ',' );
212 while( *psz_parser == ' ' )
217 if( strlen( psz_subdir ) > 0 )
219 psz_temp = (char *)malloc( strlen(psz_dir)
220 + strlen(psz_subdir) + 2 );
223 sprintf( psz_temp, "%s%s%c",
224 psz_subdir[0] == '.' ? psz_dir : "",
226 psz_subdir[strlen(psz_subdir) - 1] ==
227 DIR_SEP_CHAR ? '\0' : DIR_SEP_CHAR );
228 subdirs[i] = psz_temp;
244 * Detect subtitle files.
246 * When called this function will split up the psz_name string into a
247 * directory, filename and extension. It then opens the directory
248 * in which the file resides and tries to find possible matches of
252 * \param p_this the calling \ref input_thread_t
253 * \param psz_path a list of subdirectories (separated by a ',') to look in.
254 * \param psz_name the complete filename to base the search on.
255 * \return a NULL terminated array of filenames with detected possible subtitles.
256 * The array contains max MAX_SUBTITLE_FILES items and you need to free it after use.
258 char **subtitles_Detect( input_thread_t *p_this, char *psz_path,
262 int j, i_result2, i_sub_count = 0, i_fname_len = 0;
263 char *f_dir = NULL, *f_fname = NULL, *f_fname_noext = NULL, *f_fname_trim = NULL;
266 char **tmp_subdirs, **subdirs; /* list of subdirectories to look in */
268 subfn *result = NULL; /* unsorted results */
269 char **result2; /* sorted results */
271 char *psz_fname_original = strdup( psz_name );
272 char *psz_fname = psz_fname_original;
274 if( psz_fname == NULL ) return NULL;
276 if( !strncmp( psz_fname, "file://", 7 ) )
281 /* extract filename & dirname from psz_fname */
282 tmp = strrchr( psz_fname, DIR_SEP_CHAR );
287 f_fname = malloc( strlen(tmp) );
289 strcpy( f_fname, tmp+1 ); // we skip the separator, so it will still fit in the allocated space
290 dirlen = strlen(psz_fname) - strlen(tmp) + 2; // add the separator
291 f_dir = malloc( dirlen + 1 );
293 strlcpy( f_dir, psz_fname, dirlen );
297 /* Get the current working directory */
300 f_dir = getcwd( NULL, 0 );
304 if( psz_fname_original ) free( psz_fname_original );
307 dirlen = strlen( f_dir );
308 f_dir = (char *)realloc(f_dir, dirlen +2 );
309 f_dir[dirlen] = DIR_SEP_CHAR;
310 f_dir[dirlen+1] = '\0';
311 f_fname = FromLocaleDup( psz_fname );
314 i_fname_len = strlen( f_fname );
315 f_fname_noext = malloc(i_fname_len + 1);
316 f_fname_trim = malloc(i_fname_len + 1 );
318 strcpy_strip_ext( f_fname_noext, f_fname );
319 strcpy_trim( f_fname_trim, f_fname_noext );
321 result = (subfn*)malloc( sizeof(subfn) * MAX_SUBTITLE_FILES );
323 memset( result, 0, sizeof(subfn) * MAX_SUBTITLE_FILES );
325 var_Get( p_this, "sub-autodetect-fuzzy", &fuzzy );
327 tmp_subdirs = paths_to_list( f_dir, psz_path );
328 subdirs = tmp_subdirs;
330 for( j = -1; (j == -1) || ( (j >= 0) && (subdirs != NULL) &&
331 (*subdirs != NULL) ); j++)
333 const char *psz_dir = j < 0 ? f_dir : *subdirs;
334 char **ppsz_dir_content;
337 if( psz_dir == NULL )
340 /* parse psz_src dir */
341 i_dir_content = utf8_scandir( psz_dir, &ppsz_dir_content,
342 subtitles_Filter, NULL );
344 if( i_dir_content != -1 )
348 msg_Dbg( p_this, "looking for a subtitle file in %s", psz_dir );
349 for( a = 0; a < i_dir_content; a++ )
351 char *psz_name = ppsz_dir_content[a];
352 char tmp_fname_noext[strlen( psz_name ) + 1];
353 char tmp_fname_trim[strlen( psz_name ) + 1];
354 char tmp_fname_ext[strlen( psz_name ) + 1];
358 if( psz_name == NULL )
361 /* retrieve various parts of the filename */
362 strcpy_strip_ext( tmp_fname_noext, psz_name );
363 strcpy_get_ext( tmp_fname_ext, psz_name );
364 strcpy_trim( tmp_fname_trim, tmp_fname_noext );
366 if( !i_prio && !strcmp( tmp_fname_trim, f_fname_trim ) )
368 /* matches the movie name exactly */
372 ( tmp = strstr( tmp_fname_trim, f_fname_trim ) ) )
374 /* contains the movie name */
375 tmp += strlen( f_fname_trim );
376 if( whiteonly( tmp ) )
378 /* chars in front of the movie name */
383 /* chars after (and possibly in front of)
390 /* doesn't contain the movie name */
391 if( j == 0 ) i_prio = 1;
393 if( i_prio >= fuzzy.i_int )
395 char psz_path[strlen( psz_dir ) + strlen( psz_name ) + 1];
398 sprintf( psz_path, "%s%s", psz_dir, psz_name );
400 "autodetected subtitle: %s with priority %d",
403 if( utf8_stat( psz_path, &st ) && S_ISREG( st.st_mode ) )
406 "autodetected subtitle: %s with priority %d",
408 result[i_sub_count].priority = i_prio;
409 result[i_sub_count].psz_fname = strdup( psz_path );
410 result[i_sub_count].psz_ext = strdup(tmp_fname_ext);
415 msg_Dbg( p_this, "fopen failed" );
418 if( i_sub_count >= MAX_SUBTITLE_FILES ) break;
420 for( a = 0; a < i_dir_content; a++ )
421 free( ppsz_dir_content[a] );
422 if( ppsz_dir_content ) free( ppsz_dir_content );
424 if( j >= 0 ) if( *subdirs ) free( *subdirs++ );
427 if( tmp_subdirs ) free( tmp_subdirs );
428 if( f_fname_trim ) free( f_fname_trim );
429 if( f_fname_noext ) free( f_fname_noext );
430 if( f_fname ) free( f_fname );
431 if( f_dir ) free( f_dir );
433 qsort( result, i_sub_count, sizeof( subfn ), compare_sub_priority );
435 result2 = (char**)malloc( sizeof(char*) * ( i_sub_count + 1 ) );
437 memset( result2, 0, sizeof(char*) * ( i_sub_count + 1 ) );
440 for( j = 0; j < i_sub_count; j++ )
442 if( result[j].psz_ext && !strcasecmp( result[j].psz_ext, "sub" ) )
445 for( i = 0; i < i_sub_count; i++ )
447 if( result[i].psz_fname && result[j].psz_fname &&
448 !strncasecmp( result[j].psz_fname, result[i].psz_fname,
449 sizeof( result[j].psz_fname) - 4 ) &&
450 !strcasecmp( result[i].psz_ext, "idx" ) )
453 if( i >= i_sub_count )
455 result2[i_result2] = result[j].psz_fname;
461 result2[i_result2] = result[j].psz_fname;
466 if( psz_fname_original ) free( psz_fname_original );
467 if( result ) free( result );