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 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 _container = LMS_STATIC_STRING_SIZE("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")
79 static const char *_cats[] = {
84 static const char *_authors[] = {
85 "Andre Moreira Magalhaes",
90 struct lms_stream base;
92 unsigned int sampling_rate;
96 /* TODO: Add the gazillion of possible codecs -- possibly a task to gperf */
99 struct lms_string_size name;
101 { 0x0160, LMS_STATIC_STRING_SIZE("wmav1") },
102 { 0x0161, LMS_STATIC_STRING_SIZE("wmav2") },
103 { 0x0162, LMS_STATIC_STRING_SIZE("wmavpro") },
104 { 0x0163, LMS_STATIC_STRING_SIZE("wmavlossless") },
105 { 0x1600, LMS_STATIC_STRING_SIZE("aac") },
106 { 0x706d, LMS_STATIC_STRING_SIZE("aac") },
107 { 0x4143, LMS_STATIC_STRING_SIZE("aac") },
108 { 0xA106, LMS_STATIC_STRING_SIZE("aac") },
109 { 0xF1AC, LMS_STATIC_STRING_SIZE("flac") },
110 { 0x0055, LMS_STATIC_STRING_SIZE("mp3") },
116 * Microsoft defines these 16-byte (128-bit) GUIDs as:
117 * first 8 bytes are in little-endian order
118 * next 8 bytes are in big-endian order
120 * Eg.: AaBbCcDd-EeFf-GgHh-IiJj-KkLlMmNnOoPp:
122 * to convert to byte string do as follow:
124 * $Dd $Cc $Bb $Aa $Ff $Ee $Hh $Gg $Ii $Jj $Kk $Ll $Mm $Nn $Oo $Pp
126 * See http://www.microsoft.com/windows/windowsmedia/forpros/format/asfspec.aspx
128 static const char header_guid[16] = "\x30\x26\xB2\x75\x8E\x66\xCF\x11\xA6\xD9\x00\xAA\x00\x62\xCE\x6C";
129 static const char file_properties_guid[16] = "\xA1\xDC\xAB\x8C\x47\xA9\xCF\x11\x8E\xE4\x00\xC0\x0C\x20\x53\x65";
130 static const char stream_properties_guid[16] = "\x91\x07\xDC\xB7\xB7\xA9\xCF\x11\x8E\xE6\x00\xC0\x0C\x20\x53\x65";
131 static const char stream_type_audio_guid[16] = "\x40\x9E\x69\xF8\x4D\x5B\xCF\x11\xA8\xFD\x00\x80\x5F\x5C\x44\x2B";
132 static const char stream_type_video_guid[16] = "\xC0\xEF\x19\xBC\x4D\x5B\xCF\x11\xA8\xFD\x00\x80\x5F\x5C\x44\x2B";
133 static const char content_description_guid[16] = "\x33\x26\xB2\x75\x8E\x66\xCF\x11\xA6\xD9\x00\xAA\x00\x62\xCE\x6C";
134 static const char extended_content_description_guid[16] = "\x40\xA4\xD0\xD2\x07\xE3\xD2\x11\x97\xF0\x00\xA0\xC9\x5E\xA8\x50";
135 static const char header_extension_guid[16] = "\xb5\x03\xbf_.\xa9\xcf\x11\x8e\xe3\x00\xc0\x0c Se";
136 static const char metadata_guid[16] = "\xEA\xCB\xF8\xC5\xAF[wH\204g\xAA\214D\xFAL\xCA";
137 static const char metadata_library_guid[16] = "\224\034#D\230\224\321I\241A\x1d\x13NEpT";
138 static const char content_encryption_object_guid[16] = "\xFB\xB3\x11\x22\x23\xBD\xD2\x11\xB4\xB7\x00\xA0\xC9\x55\xFC\x6E";
139 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";
141 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";
142 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";
143 static const char attr_name_wm_genre[16] = "\x57\x00\x4d\x00\x2f\x00\x47\x00\x65\x00\x6e\x00\x72\x00\x65\x00";
144 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";
147 _to_number(const char *data, unsigned int type_size, unsigned int data_size)
150 unsigned int last, i;
152 last = data_size > type_size ? type_size : data_size;
154 for (i = 0; i < last; i++)
155 sum |= (unsigned char) (data[i]) << (i * 8);
164 if (read(fd, &v, 2) != 2)
166 return (short) _to_number(v, sizeof(unsigned short), 2);
173 if (read(fd, &v, 4) != 4)
175 return (unsigned int) _to_number(v, sizeof(unsigned int), 4);
182 if (read(fd, &v, 8) != 8)
184 return _to_number(v, sizeof(unsigned long long), 8);
188 _read_string(int fd, size_t count, char **str, unsigned int *len)
191 ssize_t data_size, size;
193 data = malloc(sizeof(char) * count);
194 data_size = read(fd, data, count);
195 if (data_size == -1) {
202 if (data[size - 1] != '\0' || data[size - 2] != '\0')
214 _parse_file_properties(lms_charset_conv_t *cs_conv, int fd,
215 struct lms_audio_info *info)
220 uint64_t creation_date;
221 uint64_t data_packets_count;
222 uint64_t play_duration;
223 uint64_t send_duration;
226 uint32_t min_data_packet_size;
227 uint32_t max_data_packet_size;
228 uint32_t max_bitrate;
229 } __attribute__((packed)) props;
232 r = read(fd, &props, sizeof(props));
233 if (r != sizeof(props))
237 if (le32toh(props.flags) & 0x1)
240 /* ASF spec 01.20.06 sec. 3.2: we need to subtract the preroll value from
241 * the duration in order to obtain the real duration */
242 info->length = (unsigned int)((le64toh(props.play_duration) / NSEC100_PER_SEC)
243 - le64toh(props.preroll) / MSEC_PER_SEC);
248 static struct lms_string_size
249 _codec_id_to_str(uint16_t id)
253 for (i = 0; _codecs[i].name.str != NULL; i++)
254 if (_codecs[i].id == id)
255 return _codecs[i].name;
257 return _codecs[i].name;
261 _parse_stream_properties(int fd, struct stream **pstream)
264 char stream_type[16];
265 char error_correction_type[16];
266 uint64_t time_offset;
267 uint32_t type_specific_len;
268 uint32_t error_correction_data_len;
270 uint32_t reserved; /* don't use, unaligned */
271 } __attribute__((packed)) props;
278 s = calloc(1, sizeof(struct stream));
282 r = read(fd, &props, sizeof(props));
283 if (r != sizeof(props))
286 if (memcmp(props.stream_type, stream_type_audio_guid, 16) == 0)
287 s->base.type = LMS_STREAM_TYPE_AUDIO;
288 else if (memcmp(props.stream_type, stream_type_video_guid, 16) == 0)
289 s->base.type = LMS_STREAM_TYPE_VIDEO;
295 s->base.stream_id = le16toh(props.flags) & 0x7F;
296 /* Not a valid stream */
297 if (!s->base.stream_id)
300 if (s->base.type == LMS_STREAM_TYPE_AUDIO) {
301 if (le32toh(props.type_specific_len) < 18)
304 s->base.codec = _codec_id_to_str(_read_word(fd));
305 s->base.audio.channels = _read_word(fd);
306 s->priv.sampling_rate = _read_dword(fd);
307 s->base.audio.bitrate = _read_dword(fd) * 8;
321 _parse_content_description(lms_charset_conv_t *cs_conv, int fd, struct asf_info *info)
323 int title_length = _read_word(fd);
324 int artist_length = _read_word(fd);
325 int copyright_length = _read_word(fd);
326 int comment_length = _read_word(fd);
327 int rating_length = _read_word(fd);
329 _read_string(fd, title_length, &info->title.str, &info->title.len);
330 lms_charset_conv_force(cs_conv, &info->title.str, &info->title.len);
331 _read_string(fd, artist_length, &info->artist.str, &info->artist.len);
332 lms_charset_conv_force(cs_conv, &info->artist.str, &info->artist.len);
333 /* ignore copyright, comment and rating */
334 lseek(fd, copyright_length + comment_length + rating_length, SEEK_CUR);
338 _parse_attribute_name(int fd,
340 unsigned int *attr_name_len,
344 int attr_name_length;
346 attr_name_length = _read_word(fd);
347 _read_string(fd, attr_name_length, attr_name, attr_name_len);
348 *attr_type = _read_word(fd);
349 *attr_size = _read_word(fd);
353 _parse_attribute_string_data(lms_charset_conv_t *cs_conv,
357 unsigned int *attr_data_len)
359 _read_string(fd, attr_size, attr_data, attr_data_len);
360 lms_charset_conv_force(cs_conv, attr_data, attr_data_len);
364 _skip_attribute_data(int fd, int kind, int attr_type, int attr_size)
368 lseek(fd, 2, SEEK_CUR);
373 lseek(fd, 4, SEEK_CUR);
375 lseek(fd, 2, SEEK_CUR);
378 case ATTR_TYPE_DWORD:
379 lseek(fd, 4, SEEK_CUR);
382 case ATTR_TYPE_QWORD:
383 lseek(fd, 8, SEEK_CUR);
386 case ATTR_TYPE_UNICODE:
387 case ATTR_TYPE_BYTES:
389 lseek(fd, attr_size, SEEK_CUR);
398 _parse_extended_content_description_object(lms_charset_conv_t *cs_conv, int fd, struct asf_info *info)
400 int count = _read_word(fd);
402 unsigned int attr_name_len;
403 int attr_type, attr_size;
406 _parse_attribute_name(fd,
407 &attr_name, &attr_name_len,
408 &attr_type, &attr_size);
409 if (attr_type == ATTR_TYPE_UNICODE) {
410 if (memcmp(attr_name, attr_name_wm_album_title, attr_name_len) == 0)
411 _parse_attribute_string_data(cs_conv,
415 else if (memcmp(attr_name, attr_name_wm_genre, attr_name_len) == 0)
416 _parse_attribute_string_data(cs_conv,
420 else if (memcmp(attr_name, attr_name_wm_album_artist, attr_name_len) == 0)
421 _parse_attribute_string_data(cs_conv,
425 else if (memcmp(attr_name, attr_name_wm_track_number, attr_name_len) == 0) {
427 unsigned int trackno_len;
428 _parse_attribute_string_data(cs_conv,
433 info->trackno = atoi(trackno);
438 _skip_attribute_data(fd, 0, attr_type, attr_size);
441 _skip_attribute_data(fd, 0, attr_type, attr_size);
447 _match(struct plugin *p, const char *path, int len, int base)
451 i = lms_which_extension(path, len, _exts, LMS_ARRAY_SIZE(_exts));
455 return (void*)(i + 1);
459 _parse(struct plugin *plugin, struct lms_context *ctxt, const struct lms_file_info *finfo, void *match)
461 struct asf_info info = { };
462 struct lms_audio_info audio_info = { };
463 struct lms_video_info video_info = { };
464 int r, fd, num_objects, i;
467 enum lms_stream_type stream_type = LMS_STREAM_TYPE_UNKNOWN;
468 struct stream *streams = NULL;
470 fd = open(finfo->path, O_RDONLY);
476 if (read(fd, &guid, 16) != 16) {
481 if (memcmp(guid, header_guid, 16) != 0) {
482 fprintf(stderr, "ERROR: invalid header (%s).\n", finfo->path);
487 size = _read_qword(fd);
488 num_objects = _read_dword(fd);
490 lseek(fd, 2, SEEK_CUR);
492 for (i = 0; i < num_objects; ++i) {
494 size = _read_qword(fd);
496 if (memcmp(guid, file_properties_guid, 16) == 0) {
497 r = _parse_file_properties(plugin->cs_conv, fd, &audio_info);
500 lseek(fd, size - (24 + r), SEEK_CUR);
501 } else if (memcmp(guid, stream_properties_guid, 16) == 0) {
503 r = _parse_stream_properties(fd, &s);
507 lseek(fd, size - (24 + r), SEEK_CUR);
511 if (stream_type != LMS_STREAM_TYPE_VIDEO)
512 stream_type = s->base.type;
514 s->base.next = (struct lms_stream *) streams;
516 } else if (memcmp(guid, content_description_guid, 16) == 0)
517 _parse_content_description(plugin->cs_conv, fd, &info);
518 else if (memcmp(guid, extended_content_description_guid, 16) == 0)
519 _parse_extended_content_description_object(plugin->cs_conv, fd, &info);
520 else if (memcmp(guid, content_encryption_object_guid, 16) == 0 ||
521 memcmp(guid, extended_content_encryption_object_guid, 16) == 0) {
522 /* ignore DRM'd files */
523 fprintf(stderr, "ERROR: ignoring DRM'd file %s\n", finfo->path);
527 lseek(fd, size - 24, SEEK_CUR);
530 /* try to define stream type by extension */
531 if (stream_type == LMS_STREAM_TYPE_UNKNOWN) {
532 long ext_idx = ((long)match) - 1;
533 if (strcmp(_exts[ext_idx].str, ".wma") == 0)
534 stream_type = LMS_STREAM_TYPE_AUDIO;
535 /* consider wmv and asf as video */
537 stream_type = LMS_STREAM_TYPE_VIDEO;
540 lms_string_size_strip_and_free(&info.title);
541 lms_string_size_strip_and_free(&info.artist);
542 lms_string_size_strip_and_free(&info.album);
543 lms_string_size_strip_and_free(&info.genre);
546 info.title = str_extract_name_from_path(finfo->path, finfo->path_len,
548 &_exts[((long) match) - 1],
552 fprintf(stderr, "file %s info\n", finfo->path);
553 fprintf(stderr, "\ttitle='%s'\n", info.title.str);
554 fprintf(stderr, "\tartist='%s'\n", info.artist.str);
555 fprintf(stderr, "\talbum='%s'\n", info.album.str);
556 fprintf(stderr, "\tgenre='%s'\n", info.genre.str);
557 fprintf(stderr, "\ttrackno=%d\n", info.trackno);
560 audio_info.container = _container;
562 if (stream_type == LMS_STREAM_TYPE_AUDIO) {
563 audio_info.id = finfo->id;
564 audio_info.title = info.title;
565 audio_info.artist = info.artist;
566 audio_info.album = info.album;
567 audio_info.genre = info.genre;
568 audio_info.trackno = info.trackno;
570 audio_info.channels = streams->base.audio.channels;
571 audio_info.bitrate = streams->base.audio.bitrate;
572 audio_info.sampling_rate = streams->priv.sampling_rate;
573 audio_info.codec = streams->base.codec;
575 r = lms_db_audio_add(plugin->audio_db, &audio_info);
577 video_info.id = finfo->id;
578 video_info.title = info.title;
579 video_info.artist = info.artist;
580 r = lms_db_video_add(plugin->video_db, &video_info);
585 struct stream *s = streams;
586 streams = (struct stream *) s->base.next;
590 free(info.title.str);
591 free(info.artist.str);
592 free(info.album.str);
593 free(info.genre.str);
595 posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED);
602 _setup(struct plugin *plugin, struct lms_context *ctxt)
604 plugin->audio_db = lms_db_audio_new(ctxt->db);
605 if (!plugin->audio_db)
607 plugin->video_db = lms_db_video_new(ctxt->db);
608 if (!plugin->video_db)
610 plugin->cs_conv = lms_charset_conv_new();
611 if (!plugin->cs_conv)
613 lms_charset_conv_add(plugin->cs_conv, "UTF-16LE");
619 _start(struct plugin *plugin, struct lms_context *ctxt)
622 r = lms_db_audio_start(plugin->audio_db);
623 r |= lms_db_video_start(plugin->video_db);
628 _finish(struct plugin *plugin, struct lms_context *ctxt)
630 if (plugin->audio_db)
631 lms_db_audio_free(plugin->audio_db);
632 if (plugin->video_db)
633 lms_db_video_free(plugin->video_db);
635 lms_charset_conv_free(plugin->cs_conv);
641 _close(struct plugin *plugin)
647 API struct lms_plugin *
648 lms_plugin_open(void)
650 struct plugin *plugin;
652 plugin = (struct plugin *)malloc(sizeof(*plugin));
653 plugin->plugin.name = _name;
654 plugin->plugin.match = (lms_plugin_match_fn_t)_match;
655 plugin->plugin.parse = (lms_plugin_parse_fn_t)_parse;
656 plugin->plugin.close = (lms_plugin_close_fn_t)_close;
657 plugin->plugin.setup = (lms_plugin_setup_fn_t)_setup;
658 plugin->plugin.start = (lms_plugin_start_fn_t)_start;
659 plugin->plugin.finish = (lms_plugin_finish_fn_t)_finish;
661 return (struct lms_plugin *)plugin;
664 API const struct lms_plugin_info *
665 lms_plugin_info(void)
667 static struct lms_plugin_info info = {
670 "Microsoft WMA, WMV and ASF",
673 "http://lms.garage.maemo.org"