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 #include <lightmediascanner_plugin.h>
32 #include <lightmediascanner_db.h>
33 #include <sys/types.h>
42 STREAM_TYPE_UNKNOWN = 0,
48 ATTR_TYPE_UNICODE = 0,
58 struct lms_string_size title;
59 struct lms_string_size artist;
60 struct lms_string_size album;
61 struct lms_string_size genre;
62 unsigned char trackno;
66 struct lms_plugin plugin;
67 lms_db_audio_t *audio_db;
68 lms_db_video_t *video_db;
69 lms_charset_conv_t *cs_conv;
72 static const char _name[] = "asf";
73 static const struct lms_string_size _exts[] = {
74 LMS_STATIC_STRING_SIZE(".wma"),
75 LMS_STATIC_STRING_SIZE(".wmv"),
76 LMS_STATIC_STRING_SIZE(".asf")
81 * Microsoft defines these 16-byte (128-bit) GUIDs as:
82 * first 8 bytes are in little-endian order
83 * next 8 bytes are in big-endian order
85 * Eg.: AaBbCcDd-EeFf-GgHh-IiJj-KkLlMmNnOoPp:
87 * to convert to byte string do as follow:
89 * $Dd $Cc $Bb $Aa $Ff $Ee $Hh $Gg $Ii $Jj $Kk $Ll $Mm $Nn $Oo $Pp
91 * See http://www.microsoft.com/windows/windowsmedia/forpros/format/asfspec.aspx
93 static const char header_guid[16] = "\x30\x26\xB2\x75\x8E\x66\xCF\x11\xA6\xD9\x00\xAA\x00\x62\xCE\x6C";
94 static const char file_properties_guid[16] = "\xA1\xDC\xAB\x8C\x47\xA9\xCF\x11\x8E\xE4\x00\xC0\x0C\x20\x53\x65";
95 static const char stream_properties_guid[16] = "\x91\x07\xDC\xB7\xB7\xA9\xCF\x11\x8E\xE6\x00\xC0\x0C\x20\x53\x65";
96 static const char stream_type_audio_guid[16] = "\x40\x9E\x69\xF8\x4D\x5B\xCF\x11\xA8\xFD\x00\x80\x5F\x5C\x44\x2B";
97 static const char stream_type_video_guid[16] = "\xC0\xEF\x19\xBC\x4D\x5B\xCF\x11\xA8\xFD\x00\x80\x5F\x5C\x44\x2B";
98 static const char content_description_guid[16] = "\x33\x26\xB2\x75\x8E\x66\xCF\x11\xA6\xD9\x00\xAA\x00\x62\xCE\x6C";
99 static const char extended_content_description_guid[16] = "\x40\xA4\xD0\xD2\x07\xE3\xD2\x11\x97\xF0\x00\xA0\xC9\x5E\xA8\x50";
100 static const char header_extension_guid[16] = "\xb5\x03\xbf_.\xa9\xcf\x11\x8e\xe3\x00\xc0\x0c Se";
101 static const char metadata_guid[16] = "\xEA\xCB\xF8\xC5\xAF[wH\204g\xAA\214D\xFAL\xCA";
102 static const char metadata_library_guid[16] = "\224\034#D\230\224\321I\241A\x1d\x13NEpT";
103 static const char content_encryption_object_guid[16] = "\xFB\xB3\x11\x22\x23\xBD\xD2\x11\xB4\xB7\x00\xA0\xC9\x55\xFC\x6E";
104 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";
106 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";
107 static const char attr_name_wm_genre[16] = "\x57\x00\x4d\x00\x2f\x00\x47\x00\x65\x00\x6e\x00\x72\x00\x65\x00";
108 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";
111 _to_number(const char *data, unsigned int type_size, unsigned int data_size)
114 unsigned int last, i;
116 last = data_size > type_size ? type_size : data_size;
118 for (i = 0; i < last; i++)
119 sum |= (unsigned char) (data[i]) << (i * 8);
128 if (read(fd, &v, 2) != 2)
130 return (short) _to_number(v, sizeof(unsigned short), 2);
137 if (read(fd, &v, 4) != 4)
139 return (unsigned int) _to_number(v, sizeof(unsigned int), 4);
146 if (read(fd, &v, 8) != 8)
148 return _to_number(v, sizeof(unsigned long long), 8);
152 _read_string(int fd, size_t count, char **str, size_t *len)
155 size_t data_size, size;
158 lseek(fd, count, SEEK_CUR);
162 data = malloc(sizeof(char) * count);
163 data_size = read(fd, data, count);
164 if (data_size == -1) {
171 if (data[size - 1] != '\0' || data[size - 2] != '\0')
183 _parse_content_description(lms_charset_conv_t *cs_conv, int fd, struct asf_info *info)
185 int title_length = _read_word(fd);
186 int artist_length = _read_word(fd);
187 int copyright_length = _read_word(fd);
188 int comment_length = _read_word(fd);
189 int rating_length = _read_word(fd);
191 _read_string(fd, title_length, &info->title.str, &info->title.len);
192 lms_charset_conv_force(cs_conv, &info->title.str, &info->title.len);
193 _read_string(fd, artist_length, &info->artist.str, &info->artist.len);
194 lms_charset_conv_force(cs_conv, &info->artist.str, &info->artist.len);
195 /* ignore copyright, comment and rating */
196 lseek(fd, copyright_length + comment_length + rating_length, SEEK_CUR);
200 _parse_attribute_name(int fd,
203 size_t *attr_name_len,
207 int attr_name_length;
209 /* extended content descriptor */
211 attr_name_length = _read_word(fd);
212 _read_string(fd, attr_name_length, attr_name, attr_name_len);
213 *attr_type = _read_word(fd);
214 *attr_size = _read_word(fd);
216 /* metadata & metadata library */
218 lseek(fd, 2 + 2, SEEK_CUR); /* language and stream */
219 attr_name_length = _read_word(fd);
220 *attr_type = _read_word(fd);
221 *attr_size = _read_dword(fd);
222 _read_string(fd, attr_name_length, attr_name, attr_name_len);
227 _parse_attribute_string_data(lms_charset_conv_t *cs_conv,
231 size_t *attr_data_len)
233 _read_string(fd, attr_size, attr_data, attr_data_len);
234 lms_charset_conv_force(cs_conv, attr_data, attr_data_len);
238 _skip_attribute_data(int fd, int kind, int attr_type, int attr_size)
242 lseek(fd, 2, SEEK_CUR);
247 lseek(fd, 4, SEEK_CUR);
249 lseek(fd, 2, SEEK_CUR);
252 case ATTR_TYPE_DWORD:
253 lseek(fd, 4, SEEK_CUR);
256 case ATTR_TYPE_QWORD:
257 lseek(fd, 8, SEEK_CUR);
260 case ATTR_TYPE_UNICODE:
261 case ATTR_TYPE_BYTES:
263 lseek(fd, attr_size, SEEK_CUR);
272 _skip_attribute(int fd, int kind)
274 int attr_type, attr_size;
275 _parse_attribute_name(fd, kind, NULL, NULL, &attr_type, &attr_size);
276 _skip_attribute_data(fd, kind, attr_type, attr_size);
280 _parse_extended_content_description_object(lms_charset_conv_t *cs_conv, int fd, struct asf_info *info)
282 int count = _read_word(fd);
284 size_t attr_name_len;
285 int attr_type, attr_size;
288 _parse_attribute_name(fd, 0,
289 &attr_name, &attr_name_len,
290 &attr_type, &attr_size);
291 if (attr_type == ATTR_TYPE_UNICODE) {
292 if (memcmp(attr_name, attr_name_wm_album_title, attr_name_len) == 0)
293 _parse_attribute_string_data(cs_conv,
297 else if (memcmp(attr_name, attr_name_wm_genre, attr_name_len) == 0)
298 _parse_attribute_string_data(cs_conv,
302 else if (memcmp(attr_name, attr_name_wm_track_number, attr_name_len) == 0) {
305 _parse_attribute_string_data(cs_conv,
310 info->trackno = atoi(trackno);
315 _skip_attribute_data(fd, 0, attr_type, attr_size);
318 _skip_attribute_data(fd, 0, attr_type, attr_size);
325 _skip_metadata(int fd)
327 int count = _read_word(fd);
329 _skip_attribute(fd, 1);
333 _skip_metadata_library(int fd)
335 int count = _read_word(fd);
337 _skip_attribute(fd, 2);
341 _skip_header_extension(int fd)
344 long long size, data_size, data_pos;
346 lseek(fd, 18, SEEK_CUR);
347 data_size = _read_dword(fd);
349 while (data_pos < data_size) {
351 size = _read_qword(fd);
354 if (memcmp(guid, metadata_guid, 16) == 0)
356 else if (memcmp(guid, metadata_library_guid, 16) == 0)
357 _skip_metadata_library(fd);
359 lseek(fd, size - 24, SEEK_CUR);
365 _strstrip(char **str, unsigned int *p_len)
368 lms_strstrip(*str, p_len);
370 if (*p_len == 0 && *str) {
377 _match(struct plugin *p, const char *path, int len, int base)
381 i = lms_which_extension(path, len, _exts, LMS_ARRAY_SIZE(_exts));
385 return (void*)(i + 1);
389 _parse(struct plugin *plugin, struct lms_context *ctxt, const struct lms_file_info *finfo, void *match)
391 struct asf_info info = {{0}, {0}, {0}, {0}, 0};
392 struct lms_audio_info audio_info = {0};
393 struct lms_video_info video_info = {0};
394 int r, fd, num_objects, i;
397 int stream_type = STREAM_TYPE_UNKNOWN;
399 fd = open(finfo->path, O_RDONLY);
405 if (read(fd, &guid, 16) != 16) {
410 if (memcmp(guid, header_guid, 16) != 0) {
411 fprintf(stderr, "ERROR: invalid header (%s).\n", finfo->path);
416 size = _read_qword(fd);
417 num_objects = _read_dword(fd);
419 lseek(fd, 2, SEEK_CUR);
421 for (i = 0; i < num_objects; ++i) {
423 size = _read_qword(fd);
425 if (memcmp(guid, file_properties_guid, 16) == 0)
426 lseek(fd, size - 24, SEEK_CUR);
427 else if (memcmp(guid, stream_properties_guid, 16) == 0) {
429 if (memcmp(guid, stream_type_audio_guid, 16) == 0)
430 stream_type = STREAM_TYPE_AUDIO;
431 else if (memcmp(guid, stream_type_video_guid, 16) == 0)
432 stream_type = STREAM_TYPE_VIDEO;
433 lseek(fd, size - 40, SEEK_CUR);
435 else if (memcmp(guid, content_description_guid, 16) == 0)
436 _parse_content_description(plugin->cs_conv, fd, &info);
437 else if (memcmp(guid, extended_content_description_guid, 16) == 0)
438 _parse_extended_content_description_object(plugin->cs_conv, fd, &info);
439 else if (memcmp(guid, header_extension_guid, 16) == 0)
440 _skip_header_extension(fd);
441 else if (memcmp(guid, content_encryption_object_guid, 16) == 0 ||
442 memcmp(guid, extended_content_encryption_object_guid, 16) == 0) {
443 /* ignore DRM'd files */
444 fprintf(stderr, "ERROR: ignoring DRM'd file %s\n", finfo->path);
449 lseek(fd, size - 24, SEEK_CUR);
452 /* try to define stream type by extension */
453 if (stream_type == STREAM_TYPE_UNKNOWN) {
454 int ext_idx = ((int)match) - 1;
455 if (strcmp(_exts[ext_idx].str, ".wma") == 0)
456 stream_type = STREAM_TYPE_AUDIO;
457 /* consider wmv and asf as video */
459 stream_type = STREAM_TYPE_VIDEO;
462 _strstrip(&info.title.str, &info.title.len);
463 _strstrip(&info.artist.str, &info.artist.len);
464 _strstrip(&info.album.str, &info.album.len);
465 _strstrip(&info.genre.str, &info.genre.len);
467 if (!info.title.str) {
469 ext_idx = ((int)match) - 1;
470 info.title.len = finfo->path_len - finfo->base - _exts[ext_idx].len;
471 info.title.str = malloc((info.title.len + 1) * sizeof(char));
472 memcpy(info.title.str, finfo->path + finfo->base, info.title.len);
473 info.title.str[info.title.len] = '\0';
474 lms_charset_conv(ctxt->cs_conv, &info.title.str, &info.title.len);
478 fprintf(stderr, "file %s info\n", finfo->path);
479 fprintf(stderr, "\ttitle='%s'\n", info.title.str);
480 fprintf(stderr, "\tartist='%s'\n", info.artist.str);
481 fprintf(stderr, "\talbum='%s'\n", info.album.str);
482 fprintf(stderr, "\tgenre='%s'\n", info.genre.str);
483 fprintf(stderr, "\ttrackno=%d\n", info.trackno);
486 if (stream_type == STREAM_TYPE_AUDIO) {
487 audio_info.id = finfo->id;
488 audio_info.title = info.title;
489 audio_info.artist = info.artist;
490 audio_info.album = info.album;
491 audio_info.genre = info.genre;
492 audio_info.trackno = info.trackno;
493 r = lms_db_audio_add(plugin->audio_db, &audio_info);
496 video_info.id = finfo->id;
497 video_info.title = info.title;
498 video_info.artist = info.artist;
499 r = lms_db_video_add(plugin->video_db, &video_info);
504 free(info.title.str);
506 free(info.artist.str);
508 free(info.album.str);
510 free(info.genre.str);
512 posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED);
519 _setup(struct plugin *plugin, struct lms_context *ctxt)
521 plugin->audio_db = lms_db_audio_new(ctxt->db);
522 if (!plugin->audio_db)
524 plugin->video_db = lms_db_video_new(ctxt->db);
525 if (!plugin->video_db)
527 plugin->cs_conv = lms_charset_conv_new();
528 if (!plugin->cs_conv)
530 lms_charset_conv_add(plugin->cs_conv, "UTF-16LE");
536 _start(struct plugin *plugin, struct lms_context *ctxt)
539 r = lms_db_audio_start(plugin->audio_db);
540 r |= lms_db_video_start(plugin->video_db);
545 _finish(struct plugin *plugin, struct lms_context *ctxt)
547 if (plugin->audio_db)
548 lms_db_audio_free(plugin->audio_db);
549 if (plugin->video_db)
550 lms_db_video_free(plugin->video_db);
552 lms_charset_conv_free(plugin->cs_conv);
558 _close(struct plugin *plugin)
564 API struct lms_plugin *
565 lms_plugin_open(void)
567 struct plugin *plugin;
569 plugin = (struct plugin *)malloc(sizeof(*plugin));
570 plugin->plugin.name = _name;
571 plugin->plugin.match = (lms_plugin_match_fn_t)_match;
572 plugin->plugin.parse = (lms_plugin_parse_fn_t)_parse;
573 plugin->plugin.close = (lms_plugin_close_fn_t)_close;
574 plugin->plugin.setup = (lms_plugin_setup_fn_t)_setup;
575 plugin->plugin.start = (lms_plugin_start_fn_t)_start;
576 plugin->plugin.finish = (lms_plugin_finish_fn_t)_finish;
578 return (struct lms_plugin *)plugin;