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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * @author Andre Moreira Magalhaes <andre.magalhaes@openbossa.org>
24 * asf/wma file parser.
27 * http://www.microsoft.com/en-us/download/details.aspx?id=14995
30 #include <lightmediascanner_plugin.h>
31 #include <lightmediascanner_db.h>
32 #include <shared/util.h>
36 #include <sys/types.h>
44 #define NSEC100_PER_SEC 10000000ULL
45 #define MSEC_PER_SEC 1000ULL
48 STREAM_TYPE_UNKNOWN = 0,
54 ATTR_TYPE_UNICODE = 0,
64 struct lms_string_size title;
65 struct lms_string_size artist;
66 struct lms_string_size album;
67 struct lms_string_size genre;
68 unsigned char trackno;
72 struct lms_plugin plugin;
73 lms_db_audio_t *audio_db;
74 lms_db_video_t *video_db;
75 lms_charset_conv_t *cs_conv;
78 static const char _name[] = "asf";
79 static const struct lms_string_size _container = LMS_STATIC_STRING_SIZE("asf");
80 static const struct lms_string_size _exts[] = {
81 LMS_STATIC_STRING_SIZE(".wma"),
82 LMS_STATIC_STRING_SIZE(".wmv"),
83 LMS_STATIC_STRING_SIZE(".asf")
85 static const char *_cats[] = {
90 static const char *_authors[] = {
91 "Andre Moreira Magalhaes",
97 enum StreamTypes type;
103 uint32_t sampling_rate;
109 /* TODO: Add the gazillion of possible codecs -- possibly a task to gperf */
110 static const struct {
112 struct lms_string_size name;
114 { 0x0160, LMS_STATIC_STRING_SIZE("wmav1") },
115 { 0x0161, LMS_STATIC_STRING_SIZE("wmav2") },
116 { 0x0162, LMS_STATIC_STRING_SIZE("wmavpro") },
117 { 0x0163, LMS_STATIC_STRING_SIZE("wmavlossless") },
118 { 0x1600, LMS_STATIC_STRING_SIZE("aac") },
119 { 0x706d, LMS_STATIC_STRING_SIZE("aac") },
120 { 0x4143, LMS_STATIC_STRING_SIZE("aac") },
121 { 0xA106, LMS_STATIC_STRING_SIZE("aac") },
122 { 0xF1AC, LMS_STATIC_STRING_SIZE("flac") },
123 { 0x0055, LMS_STATIC_STRING_SIZE("mp3") },
129 * Microsoft defines these 16-byte (128-bit) GUIDs as:
130 * first 8 bytes are in little-endian order
131 * next 8 bytes are in big-endian order
133 * Eg.: AaBbCcDd-EeFf-GgHh-IiJj-KkLlMmNnOoPp:
135 * to convert to byte string do as follow:
137 * $Dd $Cc $Bb $Aa $Ff $Ee $Hh $Gg $Ii $Jj $Kk $Ll $Mm $Nn $Oo $Pp
139 * See http://www.microsoft.com/windows/windowsmedia/forpros/format/asfspec.aspx
141 static const char header_guid[16] = "\x30\x26\xB2\x75\x8E\x66\xCF\x11\xA6\xD9\x00\xAA\x00\x62\xCE\x6C";
142 static const char file_properties_guid[16] = "\xA1\xDC\xAB\x8C\x47\xA9\xCF\x11\x8E\xE4\x00\xC0\x0C\x20\x53\x65";
143 static const char stream_properties_guid[16] = "\x91\x07\xDC\xB7\xB7\xA9\xCF\x11\x8E\xE6\x00\xC0\x0C\x20\x53\x65";
144 static const char stream_type_audio_guid[16] = "\x40\x9E\x69\xF8\x4D\x5B\xCF\x11\xA8\xFD\x00\x80\x5F\x5C\x44\x2B";
145 static const char stream_type_video_guid[16] = "\xC0\xEF\x19\xBC\x4D\x5B\xCF\x11\xA8\xFD\x00\x80\x5F\x5C\x44\x2B";
146 static const char content_description_guid[16] = "\x33\x26\xB2\x75\x8E\x66\xCF\x11\xA6\xD9\x00\xAA\x00\x62\xCE\x6C";
147 static const char extended_content_description_guid[16] = "\x40\xA4\xD0\xD2\x07\xE3\xD2\x11\x97\xF0\x00\xA0\xC9\x5E\xA8\x50";
148 static const char header_extension_guid[16] = "\xb5\x03\xbf_.\xa9\xcf\x11\x8e\xe3\x00\xc0\x0c Se";
149 static const char metadata_guid[16] = "\xEA\xCB\xF8\xC5\xAF[wH\204g\xAA\214D\xFAL\xCA";
150 static const char metadata_library_guid[16] = "\224\034#D\230\224\321I\241A\x1d\x13NEpT";
151 static const char content_encryption_object_guid[16] = "\xFB\xB3\x11\x22\x23\xBD\xD2\x11\xB4\xB7\x00\xA0\xC9\x55\xFC\x6E";
152 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";
154 static const char attr_name_wm_album_artist[28] = "\x57\x00\x4d\x00\x2f\x00\x41\x00\x6c\x00\x62\x00\x75\x00\x6d\x00\x41\x00\x72\x00\x74\x00\x69\x00\x73\x00\x74\x00";
155 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";
156 static const char attr_name_wm_genre[16] = "\x57\x00\x4d\x00\x2f\x00\x47\x00\x65\x00\x6e\x00\x72\x00\x65\x00";
157 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";
160 _to_number(const char *data, unsigned int type_size, unsigned int data_size)
163 unsigned int last, i;
165 last = data_size > type_size ? type_size : data_size;
167 for (i = 0; i < last; i++)
168 sum |= (unsigned char) (data[i]) << (i * 8);
177 if (read(fd, &v, 2) != 2)
179 return (short) _to_number(v, sizeof(unsigned short), 2);
186 if (read(fd, &v, 4) != 4)
188 return (unsigned int) _to_number(v, sizeof(unsigned int), 4);
195 if (read(fd, &v, 8) != 8)
197 return _to_number(v, sizeof(unsigned long long), 8);
201 _read_string(int fd, size_t count, char **str, unsigned int *len)
204 ssize_t data_size, size;
206 data = malloc(sizeof(char) * count);
207 data_size = read(fd, data, count);
208 if (data_size == -1) {
215 if (data[size - 1] != '\0' || data[size - 2] != '\0')
227 _parse_file_properties(lms_charset_conv_t *cs_conv, int fd,
228 struct lms_audio_info *info)
233 uint64_t creation_date;
234 uint64_t data_packets_count;
235 uint64_t play_duration;
236 uint64_t send_duration;
239 uint32_t min_data_packet_size;
240 uint32_t max_data_packet_size;
241 uint32_t max_bitrate;
242 } __attribute__((packed)) props;
245 r = read(fd, &props, sizeof(props));
246 if (r != sizeof(props))
250 if (le32toh(props.flags) & 0x1)
253 /* ASF spec 01.20.06 sec. 3.2: we need to subtract the preroll value from
254 * the duration in order to obtain the real duration */
255 info->length = (unsigned int)((le64toh(props.play_duration) / NSEC100_PER_SEC)
256 - le64toh(props.preroll) / MSEC_PER_SEC);
261 static struct lms_string_size
262 _codec_id_to_str(uint16_t id)
266 for (i = 0; _codecs[i].name.str != NULL; i++)
267 if (_codecs[i].id == id)
268 return _codecs[i].name;
270 return _codecs[i].name;
274 _parse_stream_properties(int fd, struct stream **pstream)
277 char stream_type[16];
278 char error_correction_type[16];
279 uint64_t time_offset;
280 uint32_t type_specific_len;
281 uint32_t error_correction_data_len;
283 uint32_t reserved; /* don't use, unaligned */
284 } __attribute__((packed)) props;
291 s = calloc(1, sizeof(struct stream));
295 r = read(fd, &props, sizeof(props));
296 if (r != sizeof(props))
299 if (memcmp(props.stream_type, stream_type_audio_guid, 16) == 0)
300 s->type = STREAM_TYPE_AUDIO;
301 else if (memcmp(props.stream_type, stream_type_video_guid, 16) == 0)
302 s->type = STREAM_TYPE_VIDEO;
308 s->id = le16toh(props.flags) & 0x7F;
309 /* Not a valid stream */
313 if (s->type == STREAM_TYPE_AUDIO) {
314 if (le32toh(props.type_specific_len) < 18)
317 s->audio.codec_id = _read_word(fd);
318 s->audio.channels = _read_word(fd);
319 s->audio.sampling_rate = _read_dword(fd);
320 s->audio.byterate = _read_dword(fd);
334 _parse_content_description(lms_charset_conv_t *cs_conv, int fd, struct asf_info *info)
336 int title_length = _read_word(fd);
337 int artist_length = _read_word(fd);
338 int copyright_length = _read_word(fd);
339 int comment_length = _read_word(fd);
340 int rating_length = _read_word(fd);
342 _read_string(fd, title_length, &info->title.str, &info->title.len);
343 lms_charset_conv_force(cs_conv, &info->title.str, &info->title.len);
344 _read_string(fd, artist_length, &info->artist.str, &info->artist.len);
345 lms_charset_conv_force(cs_conv, &info->artist.str, &info->artist.len);
346 /* ignore copyright, comment and rating */
347 lseek(fd, copyright_length + comment_length + rating_length, SEEK_CUR);
351 _parse_attribute_name(int fd,
353 unsigned int *attr_name_len,
357 int attr_name_length;
359 attr_name_length = _read_word(fd);
360 _read_string(fd, attr_name_length, attr_name, attr_name_len);
361 *attr_type = _read_word(fd);
362 *attr_size = _read_word(fd);
366 _parse_attribute_string_data(lms_charset_conv_t *cs_conv,
370 unsigned int *attr_data_len)
372 _read_string(fd, attr_size, attr_data, attr_data_len);
373 lms_charset_conv_force(cs_conv, attr_data, attr_data_len);
377 _skip_attribute_data(int fd, int kind, int attr_type, int attr_size)
381 lseek(fd, 2, SEEK_CUR);
386 lseek(fd, 4, SEEK_CUR);
388 lseek(fd, 2, SEEK_CUR);
391 case ATTR_TYPE_DWORD:
392 lseek(fd, 4, SEEK_CUR);
395 case ATTR_TYPE_QWORD:
396 lseek(fd, 8, SEEK_CUR);
399 case ATTR_TYPE_UNICODE:
400 case ATTR_TYPE_BYTES:
402 lseek(fd, attr_size, SEEK_CUR);
411 _parse_extended_content_description_object(lms_charset_conv_t *cs_conv, int fd, struct asf_info *info)
413 int count = _read_word(fd);
415 unsigned int attr_name_len;
416 int attr_type, attr_size;
419 _parse_attribute_name(fd,
420 &attr_name, &attr_name_len,
421 &attr_type, &attr_size);
422 if (attr_type == ATTR_TYPE_UNICODE) {
423 if (memcmp(attr_name, attr_name_wm_album_title, attr_name_len) == 0)
424 _parse_attribute_string_data(cs_conv,
428 else if (memcmp(attr_name, attr_name_wm_genre, attr_name_len) == 0)
429 _parse_attribute_string_data(cs_conv,
433 else if (memcmp(attr_name, attr_name_wm_album_artist, attr_name_len) == 0)
434 _parse_attribute_string_data(cs_conv,
438 else if (memcmp(attr_name, attr_name_wm_track_number, attr_name_len) == 0) {
440 unsigned int trackno_len;
441 _parse_attribute_string_data(cs_conv,
446 info->trackno = atoi(trackno);
451 _skip_attribute_data(fd, 0, attr_type, attr_size);
454 _skip_attribute_data(fd, 0, attr_type, attr_size);
460 _match(struct plugin *p, const char *path, int len, int base)
464 i = lms_which_extension(path, len, _exts, LMS_ARRAY_SIZE(_exts));
468 return (void*)(i + 1);
472 _parse(struct plugin *plugin, struct lms_context *ctxt, const struct lms_file_info *finfo, void *match)
474 struct asf_info info = { };
475 struct lms_audio_info audio_info = { };
476 struct lms_video_info video_info = { };
477 int r, fd, num_objects, i;
480 int stream_type = STREAM_TYPE_UNKNOWN;
481 struct stream *streams = NULL;
483 fd = open(finfo->path, O_RDONLY);
489 if (read(fd, &guid, 16) != 16) {
494 if (memcmp(guid, header_guid, 16) != 0) {
495 fprintf(stderr, "ERROR: invalid header (%s).\n", finfo->path);
500 size = _read_qword(fd);
501 num_objects = _read_dword(fd);
503 lseek(fd, 2, SEEK_CUR);
505 for (i = 0; i < num_objects; ++i) {
507 size = _read_qword(fd);
509 if (memcmp(guid, file_properties_guid, 16) == 0) {
510 r = _parse_file_properties(plugin->cs_conv, fd, &audio_info);
513 lseek(fd, size - (24 + r), SEEK_CUR);
514 } else if (memcmp(guid, stream_properties_guid, 16) == 0) {
516 r = _parse_stream_properties(fd, &s);
520 lseek(fd, size - (24 + r), SEEK_CUR);
524 if (stream_type != STREAM_TYPE_VIDEO)
525 stream_type = s->type;
529 } else if (memcmp(guid, content_description_guid, 16) == 0)
530 _parse_content_description(plugin->cs_conv, fd, &info);
531 else if (memcmp(guid, extended_content_description_guid, 16) == 0)
532 _parse_extended_content_description_object(plugin->cs_conv, fd, &info);
533 else if (memcmp(guid, content_encryption_object_guid, 16) == 0 ||
534 memcmp(guid, extended_content_encryption_object_guid, 16) == 0) {
535 /* ignore DRM'd files */
536 fprintf(stderr, "ERROR: ignoring DRM'd file %s\n", finfo->path);
540 lseek(fd, size - 24, SEEK_CUR);
543 /* try to define stream type by extension */
544 if (stream_type == STREAM_TYPE_UNKNOWN) {
545 long ext_idx = ((long)match) - 1;
546 if (strcmp(_exts[ext_idx].str, ".wma") == 0)
547 stream_type = STREAM_TYPE_AUDIO;
548 /* consider wmv and asf as video */
550 stream_type = STREAM_TYPE_VIDEO;
553 lms_string_size_strip_and_free(&info.title);
554 lms_string_size_strip_and_free(&info.artist);
555 lms_string_size_strip_and_free(&info.album);
556 lms_string_size_strip_and_free(&info.genre);
559 info.title = str_extract_name_from_path(finfo->path, finfo->path_len,
561 &_exts[((long) match) - 1],
565 fprintf(stderr, "file %s info\n", finfo->path);
566 fprintf(stderr, "\ttitle='%s'\n", info.title.str);
567 fprintf(stderr, "\tartist='%s'\n", info.artist.str);
568 fprintf(stderr, "\talbum='%s'\n", info.album.str);
569 fprintf(stderr, "\tgenre='%s'\n", info.genre.str);
570 fprintf(stderr, "\ttrackno=%d\n", info.trackno);
573 audio_info.container = _container;
575 if (stream_type == STREAM_TYPE_AUDIO) {
576 audio_info.id = finfo->id;
577 audio_info.title = info.title;
578 audio_info.artist = info.artist;
579 audio_info.album = info.album;
580 audio_info.genre = info.genre;
581 audio_info.trackno = info.trackno;
583 audio_info.channels = streams->audio.channels;
584 audio_info.bitrate = streams->audio.byterate * 8;
585 audio_info.sampling_rate = streams->audio.sampling_rate;
586 audio_info.codec = _codec_id_to_str(streams->audio.codec_id);
588 r = lms_db_audio_add(plugin->audio_db, &audio_info);
590 video_info.id = finfo->id;
591 video_info.title = info.title;
592 video_info.artist = info.artist;
593 r = lms_db_video_add(plugin->video_db, &video_info);
598 struct stream *s = streams;
603 free(info.title.str);
604 free(info.artist.str);
605 free(info.album.str);
606 free(info.genre.str);
608 posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED);
615 _setup(struct plugin *plugin, struct lms_context *ctxt)
617 plugin->audio_db = lms_db_audio_new(ctxt->db);
618 if (!plugin->audio_db)
620 plugin->video_db = lms_db_video_new(ctxt->db);
621 if (!plugin->video_db)
623 plugin->cs_conv = lms_charset_conv_new();
624 if (!plugin->cs_conv)
626 lms_charset_conv_add(plugin->cs_conv, "UTF-16LE");
632 _start(struct plugin *plugin, struct lms_context *ctxt)
635 r = lms_db_audio_start(plugin->audio_db);
636 r |= lms_db_video_start(plugin->video_db);
641 _finish(struct plugin *plugin, struct lms_context *ctxt)
643 if (plugin->audio_db)
644 lms_db_audio_free(plugin->audio_db);
645 if (plugin->video_db)
646 lms_db_video_free(plugin->video_db);
648 lms_charset_conv_free(plugin->cs_conv);
654 _close(struct plugin *plugin)
660 API struct lms_plugin *
661 lms_plugin_open(void)
663 struct plugin *plugin;
665 plugin = (struct plugin *)malloc(sizeof(*plugin));
666 plugin->plugin.name = _name;
667 plugin->plugin.match = (lms_plugin_match_fn_t)_match;
668 plugin->plugin.parse = (lms_plugin_parse_fn_t)_parse;
669 plugin->plugin.close = (lms_plugin_close_fn_t)_close;
670 plugin->plugin.setup = (lms_plugin_setup_fn_t)_setup;
671 plugin->plugin.start = (lms_plugin_start_fn_t)_start;
672 plugin->plugin.finish = (lms_plugin_finish_fn_t)_finish;
674 return (struct lms_plugin *)plugin;
677 API const struct lms_plugin_info *
678 lms_plugin_info(void)
680 static struct lms_plugin_info info = {
683 "Microsoft WMA, WMV and ASF",
686 "http://lms.garage.maemo.org"