1 /*****************************************************************************
3 *****************************************************************************
4 * Copyright (C) 2003-2004 VideoLAN
5 * $Id: subtitles.c,v 1.9 2004/01/26 19:20:10 gbazin 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 # 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 * Detect subtitle files.
156 * When called this function will split up the psz_fname string into a
157 * directory, filename and extension. It then opens the directory
158 * in which the file resides and tries to find possible matches of
162 * \param p_this the calling \ref input_thread_t
163 * \param psz_path a subdirectory to look into. This is not used atm.
164 * \param psz_fname the complete filename to base the search on.
165 * \return a NULL terminated array of filenames with detected possible subtitles.
166 * The array contains max MAX_SUBTITLE_FILES items and you need to free it after use.
168 char **subtitles_Detect( input_thread_t *p_this, char *psz_path,
171 /* variables to be used for derivatives of psz_fname */
172 char *f_dir, *f_fname, *f_fname_noext, *f_fname_trim, *tmp;
173 /* variables to be used for derivatives FILE *f */
174 char *tmp_fname_noext, *tmp_fname_trim, *tmp_fname_ext, *tmpresult;
177 int len, i, j, i_sub_count;
178 subfn *result; /* unsorted results */
179 char **result2; /* sorted results */
186 len = ( strlen( psz_fname ) > 256 ? strlen( psz_fname ) : 256 ) +
187 ( strlen( psz_path ) > 256 ? strlen( psz_path ) : 256 ) + 2;
189 f_dir = (char*)malloc(len);
190 f_fname = (char*)malloc(len);
191 f_fname_noext = (char*)malloc(len);
192 f_fname_trim = (char*)malloc(len);
194 tmp_fname_noext = (char*)malloc(len);
195 tmp_fname_trim = (char*)malloc(len);
196 tmp_fname_ext = (char*)malloc(len);
198 tmpresult = (char*)malloc(len);
200 result = (subfn*)malloc( sizeof(subfn) * MAX_SUBTITLE_FILES );
201 memset( result, 0, sizeof(subfn) * MAX_SUBTITLE_FILES );
203 /* extract filename & dirname from psz_fname */
204 tmp = strrchr( psz_fname, DIRECTORY_SEPARATOR );
208 strncpy( f_fname, tmp + 1, len - 1 );
209 f_fname[len - 1] = 0;
210 pos = tmp - psz_fname + 1;
211 strncpy( f_dir, psz_fname, __MIN(pos,len-1) );
212 f_dir[__MIN(pos,len-1)] = 0;
216 strncpy( f_fname, psz_fname, len - 1 );
217 f_fname[len - 1] = 0;
221 strcpy_strip_ext( f_fname_noext, f_fname );
222 strcpy_trim( f_fname_trim, f_fname_noext );
223 var_Get( p_this, "sub-autodetect-fuzzy", &fuzzy );
226 for( j = 0; j <= 1; j++)
228 d = opendir( j == 0 ? f_dir : psz_path );
232 while( ( de = readdir( d ) ) )
234 /* retrieve various parts of the filename */
235 strcpy_strip_ext( tmp_fname_noext, de->d_name );
236 strcpy_get_ext( tmp_fname_ext, de->d_name );
237 strcpy_trim( tmp_fname_trim, tmp_fname_noext );
239 /* does it end with a subtitle extension? */
241 for( i = 0; sub_exts[i]; i++ )
243 if( strcmp(sub_exts[i], tmp_fname_ext ) == 0 )
246 msg_Dbg( p_this, "found a possible subtitle: %s", de->d_name );
251 /* we have a (likely) subtitle file */
255 if( !i_prio && strcmp( tmp_fname_trim, f_fname_trim ) == 0 )
257 /* matches the movie name exactly */
260 if( !i_prio && ( tmp = strstr( tmp_fname_trim, f_fname_trim ) ) )
262 /* contains the movie name */
263 tmp += strlen( f_fname_trim );
264 if( whiteonly( tmp ) )
266 /* chars in front of the movie name */
271 /* chars after (and possibly in front of)
278 /* doesn't contain the movie name */
279 if( j == 0 ) i_prio = 1;
282 if( i_prio >= fuzzy.i_int )
284 sprintf( tmpresult, "%s%s", j == 0 ? f_dir : psz_path, de->d_name );
285 msg_Dbg( p_this, "autodetected subtitle: %s with priority %d", de->d_name, i_prio );
286 if( ( f = fopen( tmpresult, "rt" ) ) )
289 result[i_sub_count].priority = i_prio;
290 result[i_sub_count].psz_fname = strdup( tmpresult );
296 if( i_sub_count >= MAX_SUBTITLE_FILES ) break;
304 free( f_fname_noext );
305 free( f_fname_trim );
307 free( tmp_fname_noext );
308 free( tmp_fname_trim );
309 free( tmp_fname_ext );
313 qsort( result, i_sub_count, sizeof( subfn ), compare_sub_priority );
315 result2 = (char**)malloc( sizeof(char*) * ( i_sub_count + 1 ) );
316 memset( result2, 0, sizeof(char*) * ( i_sub_count + 1 ) );
318 for( i = 0; i < i_sub_count; i++ )
320 result2[i] = result[i].psz_fname;
322 result2[i_sub_count] = NULL;