2 * Copyright (C) 2008 by INdT
4 * This program is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU Lesser General Public License
6 * as published by the Free Software Foundation; either version 2
7 * of the License, or (at your option) any later version.
9 * This program is distributed in the hope that it will be useful,
10 * but WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 * GNU General Public License for more details.
14 * You should have received a copy of the GNU Lesser General Public License
15 * along with this program; if not, write to the Free Software
16 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
18 * @author Andre Moreira Magalhaes <andre.magalhaes@openbossa.org>
24 * asf/wma file parser.
31 #define _XOPEN_SOURCE 600
32 #include <lightmediascanner_plugin.h>
33 #include <lightmediascanner_db.h>
34 #include <sys/types.h>
43 STREAM_TYPE_UNKNOWN = 0,
49 ATTR_TYPE_UNICODE = 0,
59 struct lms_string_size title;
60 struct lms_string_size artist;
61 struct lms_string_size album;
62 struct lms_string_size genre;
63 unsigned char trackno;
67 struct lms_plugin plugin;
68 lms_db_audio_t *audio_db;
69 lms_db_video_t *video_db;
70 lms_charset_conv_t *cs_conv;
73 static const char _name[] = "asf";
74 static const struct lms_string_size _exts[] = {
75 LMS_STATIC_STRING_SIZE(".wma"),
76 LMS_STATIC_STRING_SIZE(".wmv"),
77 LMS_STATIC_STRING_SIZE(".asf")
82 * Microsoft defines these 16-byte (128-bit) GUIDs as:
83 * first 8 bytes are in little-endian order
84 * next 8 bytes are in big-endian order
86 * Eg.: AaBbCcDd-EeFf-GgHh-IiJj-KkLlMmNnOoPp:
88 * to convert to byte string do as follow:
90 * $Dd $Cc $Bb $Aa $Ff $Ee $Hh $Gg $Ii $Jj $Kk $Ll $Mm $Nn $Oo $Pp
92 * See http://www.microsoft.com/windows/windowsmedia/forpros/format/asfspec.aspx
94 static const char header_guid[16] = "\x30\x26\xB2\x75\x8E\x66\xCF\x11\xA6\xD9\x00\xAA\x00\x62\xCE\x6C";
95 static const char file_properties_guid[16] = "\xA1\xDC\xAB\x8C\x47\xA9\xCF\x11\x8E\xE4\x00\xC0\x0C\x20\x53\x65";
96 static const char stream_properties_guid[16] = "\x91\x07\xDC\xB7\xB7\xA9\xCF\x11\x8E\xE6\x00\xC0\x0C\x20\x53\x65";
97 static const char stream_type_audio_guid[16] = "\x40\x9E\x69\xF8\x4D\x5B\xCF\x11\xA8\xFD\x00\x80\x5F\x5C\x44\x2B";
98 static const char stream_type_video_guid[16] = "\xC0\xEF\x19\xBC\x4D\x5B\xCF\x11\xA8\xFD\x00\x80\x5F\x5C\x44\x2B";
99 static const char content_description_guid[16] = "\x33\x26\xB2\x75\x8E\x66\xCF\x11\xA6\xD9\x00\xAA\x00\x62\xCE\x6C";
100 static const char extended_content_description_guid[16] = "\x40\xA4\xD0\xD2\x07\xE3\xD2\x11\x97\xF0\x00\xA0\xC9\x5E\xA8\x50";
101 static const char header_extension_guid[16] = "\xb5\x03\xbf_.\xa9\xcf\x11\x8e\xe3\x00\xc0\x0c Se";
102 static const char metadata_guid[16] = "\xEA\xCB\xF8\xC5\xAF[wH\204g\xAA\214D\xFAL\xCA";
103 static const char metadata_library_guid[16] = "\224\034#D\230\224\321I\241A\x1d\x13NEpT";
104 static const char content_encryption_object_guid[16] = "\xFB\xB3\x11\x22\x23\xBD\xD2\x11\xB4\xB7\x00\xA0\xC9\x55\xFC\x6E";
105 static const char extended_content_encryption_object_guid[16] = "\x14\xE6\x8A\x29\x22\x26\x17\x4C\xB9\x35\xDA\xE0\x7E\xE9\x28\x9C";
107 static const char attr_name_wm_album_title[26] = "\x57\x00\x4d\x00\x2f\x00\x41\x00\x6c\x00\x62\x00\x75\x00\x6d\x00\x54\x00\x69\x00\x74\x00\x6c\x00\x65\x00";
108 static const char attr_name_wm_genre[16] = "\x57\x00\x4d\x00\x2f\x00\x47\x00\x65\x00\x6e\x00\x72\x00\x65\x00";
109 static const char attr_name_wm_track_number[28] = "\x57\x00\x4d\x00\x2f\x00\x54\x00\x72\x00\x61\x00\x63\x00\x6b\x00\x4e\x00\x75\x00\x6d\x00\x62\x00\x65\x00\x72\x00";
112 _to_number(const char *data, unsigned int type_size, unsigned int data_size)
115 unsigned int last, i;
117 last = data_size > type_size ? type_size : data_size;
119 for (i = 0; i < last; i++)
120 sum |= (unsigned char) (data[i]) << (i * 8);
129 if (read(fd, &v, 2) != 2)
131 return (short) _to_number(v, sizeof(unsigned short), 2);
138 if (read(fd, &v, 4) != 4)
140 return (unsigned int) _to_number(v, sizeof(unsigned int), 4);
147 if (read(fd, &v, 8) != 8)
149 return _to_number(v, sizeof(unsigned long long), 8);
153 _read_string(int fd, size_t count, char **str, size_t *len)
156 size_t data_size, size;
159 lseek(fd, count, SEEK_CUR);
163 data = malloc(sizeof(char) * count);
164 data_size = read(fd, data, count);
165 if (data_size == -1) {
172 if (data[size - 1] != '\0' || data[size - 2] != '\0')
184 _parse_content_description(lms_charset_conv_t *cs_conv, int fd, struct asf_info *info)
186 int title_length = _read_word(fd);
187 int artist_length = _read_word(fd);
188 int copyright_length = _read_word(fd);
189 int comment_length = _read_word(fd);
190 int rating_length = _read_word(fd);
192 _read_string(fd, title_length, &info->title.str, &info->title.len);
193 lms_charset_conv_force(cs_conv, &info->title.str, &info->title.len);
194 _read_string(fd, artist_length, &info->artist.str, &info->artist.len);
195 lms_charset_conv_force(cs_conv, &info->artist.str, &info->artist.len);
196 /* ignore copyright, comment and rating */
197 lseek(fd, copyright_length + comment_length + rating_length, SEEK_CUR);
201 _parse_attribute_name(int fd,
204 size_t *attr_name_len,
208 int attr_name_length;
210 /* extended content descriptor */
212 attr_name_length = _read_word(fd);
213 _read_string(fd, attr_name_length, attr_name, attr_name_len);
214 *attr_type = _read_word(fd);
215 *attr_size = _read_word(fd);
217 /* metadata & metadata library */
219 lseek(fd, 2 + 2, SEEK_CUR); /* language and stream */
220 attr_name_length = _read_word(fd);
221 *attr_type = _read_word(fd);
222 *attr_size = _read_dword(fd);
223 _read_string(fd, attr_name_length, attr_name, attr_name_len);
228 _parse_attribute_string_data(lms_charset_conv_t *cs_conv,
232 size_t *attr_data_len)
234 _read_string(fd, attr_size, attr_data, attr_data_len);
235 lms_charset_conv_force(cs_conv, attr_data, attr_data_len);
239 _skip_attribute_data(int fd, int kind, int attr_type, int attr_size)
243 lseek(fd, 2, SEEK_CUR);
248 lseek(fd, 4, SEEK_CUR);
250 lseek(fd, 2, SEEK_CUR);
253 case ATTR_TYPE_DWORD:
254 lseek(fd, 4, SEEK_CUR);
257 case ATTR_TYPE_QWORD:
258 lseek(fd, 8, SEEK_CUR);
261 case ATTR_TYPE_UNICODE:
262 case ATTR_TYPE_BYTES:
264 lseek(fd, attr_size, SEEK_CUR);
273 _skip_attribute(int fd, int kind)
275 int attr_type, attr_size;
276 _parse_attribute_name(fd, kind, NULL, NULL, &attr_type, &attr_size);
277 _skip_attribute_data(fd, kind, attr_type, attr_size);
281 _parse_extended_content_description_object(lms_charset_conv_t *cs_conv, int fd, struct asf_info *info)
283 int count = _read_word(fd);
285 size_t attr_name_len;
286 int attr_type, attr_size;
289 _parse_attribute_name(fd, 0,
290 &attr_name, &attr_name_len,
291 &attr_type, &attr_size);
292 if (attr_type == ATTR_TYPE_UNICODE) {
293 if (memcmp(attr_name, attr_name_wm_album_title, attr_name_len) == 0)
294 _parse_attribute_string_data(cs_conv,
298 else if (memcmp(attr_name, attr_name_wm_genre, attr_name_len) == 0)
299 _parse_attribute_string_data(cs_conv,
303 else if (memcmp(attr_name, attr_name_wm_track_number, attr_name_len) == 0) {
306 _parse_attribute_string_data(cs_conv,
311 info->trackno = atoi(trackno);
316 _skip_attribute_data(fd, 0, attr_type, attr_size);
319 _skip_attribute_data(fd, 0, attr_type, attr_size);
326 _skip_metadata(int fd)
328 int count = _read_word(fd);
330 _skip_attribute(fd, 1);
334 _skip_metadata_library(int fd)
336 int count = _read_word(fd);
338 _skip_attribute(fd, 2);
342 _skip_header_extension(int fd)
345 long long size, data_size, data_pos;
347 lseek(fd, 18, SEEK_CUR);
348 data_size = _read_dword(fd);
350 while (data_pos < data_size) {
352 size = _read_qword(fd);
355 if (memcmp(guid, metadata_guid, 16) == 0)
357 else if (memcmp(guid, metadata_library_guid, 16) == 0)
358 _skip_metadata_library(fd);
360 lseek(fd, size - 24, SEEK_CUR);
366 _strstrip(char **str, unsigned int *p_len)
369 lms_strstrip(*str, p_len);
371 if (*p_len == 0 && *str) {
378 _match(struct plugin *p, const char *path, int len, int base)
382 i = lms_which_extension(path, len, _exts, LMS_ARRAY_SIZE(_exts));
386 return (void*)(i + 1);
390 _parse(struct plugin *plugin, struct lms_context *ctxt, const struct lms_file_info *finfo, void *match)
392 struct asf_info info = {{0}, {0}, {0}, {0}, 0};
393 struct lms_audio_info audio_info = {0};
394 struct lms_video_info video_info = {0};
395 int r, fd, num_objects, i;
398 int stream_type = STREAM_TYPE_UNKNOWN;
400 fd = open(finfo->path, O_RDONLY);
406 if (read(fd, &guid, 16) != 16) {
411 if (memcmp(guid, header_guid, 16) != 0) {
412 fprintf(stderr, "ERROR: invalid header (%s).\n", finfo->path);
417 size = _read_qword(fd);
418 num_objects = _read_dword(fd);
420 lseek(fd, 2, SEEK_CUR);
422 for (i = 0; i < num_objects; ++i) {
424 size = _read_qword(fd);
426 if (memcmp(guid, file_properties_guid, 16) == 0)
427 lseek(fd, size - 24, SEEK_CUR);
428 else if (memcmp(guid, stream_properties_guid, 16) == 0) {
430 if (memcmp(guid, stream_type_audio_guid, 16) == 0)
431 stream_type = STREAM_TYPE_AUDIO;
432 else if (memcmp(guid, stream_type_video_guid, 16) == 0)
433 stream_type = STREAM_TYPE_VIDEO;
434 lseek(fd, size - 40, SEEK_CUR);
436 else if (memcmp(guid, content_description_guid, 16) == 0)
437 _parse_content_description(plugin->cs_conv, fd, &info);
438 else if (memcmp(guid, extended_content_description_guid, 16) == 0)
439 _parse_extended_content_description_object(plugin->cs_conv, fd, &info);
440 else if (memcmp(guid, header_extension_guid, 16) == 0)
441 _skip_header_extension(fd);
442 else if (memcmp(guid, content_encryption_object_guid, 16) == 0 ||
443 memcmp(guid, extended_content_encryption_object_guid, 16) == 0) {
444 /* ignore DRM'd files */
445 fprintf(stderr, "ERROR: ignoring DRM'd file %s\n", finfo->path);
450 lseek(fd, size - 24, SEEK_CUR);
453 /* try to define stream type by extension */
454 if (stream_type == STREAM_TYPE_UNKNOWN) {
455 int ext_idx = ((int)match) - 1;
456 if (strcmp(_exts[ext_idx].str, ".wma") == 0)
457 stream_type = STREAM_TYPE_AUDIO;
458 /* consider wmv and asf as video */
460 stream_type = STREAM_TYPE_VIDEO;
463 _strstrip(&info.title.str, &info.title.len);
464 _strstrip(&info.artist.str, &info.artist.len);
465 _strstrip(&info.album.str, &info.album.len);
466 _strstrip(&info.genre.str, &info.genre.len);
468 if (!info.title.str) {
470 ext_idx = ((int)match) - 1;
471 info.title.len = finfo->path_len - finfo->base - _exts[ext_idx].len;
472 info.title.str = malloc((info.title.len + 1) * sizeof(char));
473 memcpy(info.title.str, finfo->path + finfo->base, info.title.len);
474 info.title.str[info.title.len] = '\0';
475 lms_charset_conv(ctxt->cs_conv, &info.title.str, &info.title.len);
479 fprintf(stderr, "file %s info\n", finfo->path);
480 fprintf(stderr, "\ttitle='%s'\n", info.title.str);
481 fprintf(stderr, "\tartist='%s'\n", info.artist.str);
482 fprintf(stderr, "\talbum='%s'\n", info.album.str);
483 fprintf(stderr, "\tgenre='%s'\n", info.genre.str);
484 fprintf(stderr, "\ttrackno=%d\n", info.trackno);
487 if (stream_type == STREAM_TYPE_AUDIO) {
488 audio_info.id = finfo->id;
489 audio_info.title = info.title;
490 audio_info.artist = info.artist;
491 audio_info.album = info.album;
492 audio_info.genre = info.genre;
493 audio_info.trackno = info.trackno;
494 r = lms_db_audio_add(plugin->audio_db, &audio_info);
497 video_info.id = finfo->id;
498 video_info.title = info.title;
499 video_info.artist = info.artist;
500 r = lms_db_video_add(plugin->video_db, &video_info);
505 free(info.title.str);
507 free(info.artist.str);
509 free(info.album.str);
511 free(info.genre.str);
513 posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED);
520 _setup(struct plugin *plugin, struct lms_context *ctxt)
522 plugin->audio_db = lms_db_audio_new(ctxt->db);
523 if (!plugin->audio_db)
525 plugin->video_db = lms_db_video_new(ctxt->db);
526 if (!plugin->video_db)
528 plugin->cs_conv = lms_charset_conv_new();
529 if (!plugin->cs_conv)
531 lms_charset_conv_add(plugin->cs_conv, "UTF-16LE");
537 _start(struct plugin *plugin, struct lms_context *ctxt)
540 r = lms_db_audio_start(plugin->audio_db);
541 r |= lms_db_video_start(plugin->video_db);
546 _finish(struct plugin *plugin, struct lms_context *ctxt)
548 if (plugin->audio_db)
549 lms_db_audio_free(plugin->audio_db);
550 if (plugin->video_db)
551 lms_db_video_free(plugin->video_db);
553 lms_charset_conv_free(plugin->cs_conv);
559 _close(struct plugin *plugin)
565 API struct lms_plugin *
566 lms_plugin_open(void)
568 struct plugin *plugin;
570 plugin = (struct plugin *)malloc(sizeof(*plugin));
571 plugin->plugin.name = _name;
572 plugin->plugin.match = (lms_plugin_match_fn_t)_match;
573 plugin->plugin.parse = (lms_plugin_parse_fn_t)_parse;
574 plugin->plugin.close = (lms_plugin_close_fn_t)_close;
575 plugin->plugin.setup = (lms_plugin_setup_fn_t)_setup;
576 plugin->plugin.start = (lms_plugin_start_fn_t)_start;
577 plugin->plugin.finish = (lms_plugin_finish_fn_t)_finish;
579 return (struct lms_plugin *)plugin;