1 /* -*- Mode: C; tab-width: 2; indent-tabs-mode: t; c-basic-offset: 2 -*- */
2 /* Copyright 2005 Jan Schmidt <thaytan@mad.scientist.com>
3 * Copyright 2002,2003 Scott Wheeler <wheeler@kde.org> (portions from taglib)
5 * This library is free software; you can redistribute it and/or
6 * modify it under the terms of the GNU Library General Public
7 * License as published by the Free Software Foundation; either
8 * version 2 of the License, or (at your option) any later version.
10 * This library is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * Library General Public License for more details.
15 * You should have received a copy of the GNU Library General Public
16 * License along with this library; if not, write to the
17 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
18 * Boston, MA 02111-1307, USA.
26 #include <gst/tag/tag.h>
30 #define HANDLE_INVALID_SYNCSAFE
32 static gboolean id3v2_frames_to_tag_list (ID3TagsWorking * work, guint size);
34 #ifndef GST_DISABLE_GST_DEBUG
36 #define GST_CAT_DEFAULT id3v2_ensure_debug_category()
39 id3v2_ensure_debug_category (void)
41 static gsize cat_gonce = 0;
43 if (g_once_init_enter (&cat_gonce)) {
46 cat = (gsize) _gst_debug_category_new ("id3v2", 0, "ID3v2 tag parsing");
48 g_once_init_leave (&cat_gonce, cat);
51 return (GstDebugCategory *) cat_gonce;
54 #endif /* GST_DISABLE_GST_DEBUG */
57 id3v2_read_synch_uint (const guint8 * data, guint size)
66 for (i = 0; i <= size; i++) {
67 invalid |= data[i] & 0x80;
68 result |= (data[i] & 0x7f) << ((size - i) * 7);
71 #ifdef HANDLE_INVALID_SYNCSAFE
73 GST_WARNING ("Invalid synch-safe integer in ID3v2 frame "
74 "- using the actual value instead");
76 for (i = 0; i <= size; i++) {
77 result |= data[i] << ((size - i) * 8);
85 * gst_tag_get_id3v2_tag_size:
86 * @buffer: buffer holding ID3v2 tag (or at least the start of one)
88 * Determines size of an ID3v2 tag on buffer containing at least ID3v2 header,
89 * i.e. at least #GST_TAG_ID3V2_HEADER_SIZE (10) bytes;
91 * Returns: Size of tag, or 0 if header is invalid or too small.
96 gst_tag_get_id3v2_tag_size (GstBuffer * buffer)
101 g_return_val_if_fail (buffer != NULL, 0);
103 if (GST_BUFFER_SIZE (buffer) < ID3V2_HDR_SIZE)
106 data = GST_BUFFER_DATA (buffer);
108 /* Check for 'ID3' string at start of buffer */
109 if (data[0] != 'I' || data[1] != 'D' || data[2] != '3') {
110 GST_DEBUG ("No ID3v2 tag in data");
117 /* Read the size from the header */
118 size = id3v2_read_synch_uint (data + 6, 4);
120 return ID3V2_HDR_SIZE;
122 size += ID3V2_HDR_SIZE;
124 /* Expand the read size to include a footer if there is one */
125 if ((flags & ID3V2_HDR_FLAG_FOOTER))
128 GST_DEBUG ("ID3v2 tag, size: %u bytes", size);
133 id3v2_ununsync_data (const guint8 * unsync_data, guint32 * size)
139 uu = out = g_malloc (*size);
141 for (end = unsync_data + *size; unsync_data < end - 1; ++unsync_data, ++uu) {
143 if (G_UNLIKELY (*unsync_data == 0xff && *(unsync_data + 1) == 0x00))
147 /* take care of last byte (if last two bytes weren't 0xff 0x00) */
148 if (unsync_data < end) {
154 GST_DEBUG ("size after un-unsyncing: %u (before: %u)", out_size, *size);
161 * gst_tag_list_from_id3v2_tag:
162 * @buffer: buffer to convert
164 * Creates a new tag list that contains the information parsed out of a
167 * Returns: A new #GstTagList with all tags that could be extracted from the
168 * given vorbiscomment buffer or NULL on error.
173 gst_tag_list_from_id3v2_tag (GstBuffer * buffer)
175 guint8 *data, *uu_data = NULL;
181 read_size = gst_tag_get_id3v2_tag_size (buffer);
183 /* Ignore tag if it has no frames attached, but skip the header then */
184 if (read_size < ID3V2_HDR_SIZE)
187 data = GST_BUFFER_DATA (buffer);
189 /* Read the version */
190 version = GST_READ_UINT16_BE (data + 3);
195 /* Validate the version. At the moment, we only support up to 2.4.0 */
196 if (ID3V2_VER_MAJOR (version) > 4 || ID3V2_VER_MINOR (version) > 0) {
197 GST_WARNING ("ID3v2 tag is from revision 2.%d.%d, "
198 "but decoder only supports 2.%d.%d. Ignoring as per spec.",
199 version >> 8, version & 0xff, ID3V2_VERSION >> 8, ID3V2_VERSION & 0xff);
203 GST_DEBUG ("ID3v2 header flags: %s %s %s %s",
204 (flags & ID3V2_HDR_FLAG_UNSYNC) ? "UNSYNC" : "",
205 (flags & ID3V2_HDR_FLAG_EXTHDR) ? "EXTENDED_HEADER" : "",
206 (flags & ID3V2_HDR_FLAG_EXPERIMENTAL) ? "EXPERIMENTAL" : "",
207 (flags & ID3V2_HDR_FLAG_FOOTER) ? "FOOTER" : "");
209 /* This shouldn't really happen! Caller should have checked first */
210 if (GST_BUFFER_SIZE (buffer) < read_size) {
212 ("Found ID3v2 tag with revision 2.%d.%d - need %u more bytes to read",
213 version >> 8, version & 0xff,
214 (guint) (read_size - GST_BUFFER_SIZE (buffer)));
218 GST_DEBUG ("Reading ID3v2 tag with revision 2.%d.%d of size %u", version >> 8,
219 version & 0xff, read_size);
221 GST_MEMDUMP ("ID3v2 tag", GST_BUFFER_DATA (buffer), read_size);
223 memset (&work, 0, sizeof (ID3TagsWorking));
224 work.buffer = buffer;
225 work.hdr.version = version;
226 work.hdr.size = read_size;
227 work.hdr.flags = flags;
228 work.hdr.frame_data = GST_BUFFER_DATA (buffer) + ID3V2_HDR_SIZE;
229 if (flags & ID3V2_HDR_FLAG_FOOTER)
230 work.hdr.frame_data_size = read_size - ID3V2_HDR_SIZE - 10;
232 work.hdr.frame_data_size = read_size - ID3V2_HDR_SIZE;
234 /* in v2.3 the frame sizes are not syncsafe, so the entire tag had to be
235 * unsynced. In v2.4 the frame sizes are syncsafe so it's just the frame
236 * data that needs un-unsyncing, but not the frame headers. */
237 if ((flags & ID3V2_HDR_FLAG_UNSYNC) != 0 && ID3V2_VER_MAJOR (version) <= 3) {
238 GST_DEBUG ("Un-unsyncing entire tag");
239 uu_data = id3v2_ununsync_data (work.hdr.frame_data,
240 &work.hdr.frame_data_size);
241 work.hdr.frame_data = uu_data;
242 GST_MEMDUMP ("ID3v2 tag (un-unsyced)", uu_data, work.hdr.frame_data_size);
245 id3v2_frames_to_tag_list (&work, work.hdr.frame_data_size);
253 id3v2_frame_hdr_size (guint id3v2ver)
255 /* ID3v2 < 2.3.0 only had 6 byte header */
256 switch (ID3V2_VER_MAJOR (id3v2ver)) {
268 static const gchar obsolete_frame_ids[][5] = {
269 {"CRM"}, {"EQU"}, {"LNK"}, {"RVA"}, {"TIM"}, {"TSI"}, /* From 2.2 */
270 {"EQUA"}, {"RVAD"}, {"TIME"}, {"TRDA"}, {"TSIZ"} /* From 2.3 */
273 static const struct ID3v2FrameIDConvert
277 } frame_id_conversions[] = {
303 "TDA", "TDAT"}, { /* obsolete, but we need to parse it anyway */
344 convert_fid_to_v240 (gchar * frame_id)
348 for (i = 0; i < G_N_ELEMENTS (obsolete_frame_ids); ++i) {
349 if (strncmp (frame_id, obsolete_frame_ids[i], 5) == 0)
353 for (i = 0; i < G_N_ELEMENTS (frame_id_conversions); ++i) {
354 if (strncmp (frame_id, frame_id_conversions[i].orig, 5) == 0) {
355 strcpy (frame_id, frame_id_conversions[i].new);
363 /* add unknown or unhandled ID3v2 frames to the taglist as binary blobs */
365 id3v2_add_id3v2_frame_blob_to_taglist (ID3TagsWorking * work, guint size)
371 guint frame_size, header_size;
374 switch (ID3V2_VER_MAJOR (work->hdr.version)) {
381 header_size = 4 + 4 + 2;
384 g_return_if_reached ();
387 frame_data = work->hdr.frame_data - header_size;
388 frame_size = size + header_size;
390 blob = gst_buffer_new_and_alloc (frame_size);
391 memcpy (GST_BUFFER_DATA (blob), frame_data, frame_size);
393 /* Sanitize frame id */
394 for (i = 0; i < 4; i++) {
395 if (!g_ascii_isalnum (frame_data[i]))
399 media_type = g_strdup_printf ("application/x-gst-id3v2-%c%c%c%c-frame",
400 g_ascii_tolower (frame_data[0]), g_ascii_tolower (frame_data[1]),
401 g_ascii_tolower (frame_data[2]), g_ascii_tolower (frame_data[3]));
402 caps = gst_caps_new_simple (media_type, "version", G_TYPE_INT,
403 (gint) ID3V2_VER_MAJOR (work->hdr.version), NULL);
404 gst_buffer_set_caps (blob, caps);
405 gst_caps_unref (caps);
408 /* gst_util_dump_mem (GST_BUFFER_DATA (blob), GST_BUFFER_SIZE (blob)); */
410 gst_tag_list_add (work->tags, GST_TAG_MERGE_APPEND,
411 GST_TAG_ID3V2_FRAME, blob, NULL);
412 gst_buffer_unref (blob);
416 id3v2_frames_to_tag_list (ID3TagsWorking * work, guint size)
418 guint frame_hdr_size;
421 /* Extended header if present */
422 if (work->hdr.flags & ID3V2_HDR_FLAG_EXTHDR) {
423 work->hdr.ext_hdr_size = id3v2_read_synch_uint (work->hdr.frame_data, 4);
424 if (work->hdr.ext_hdr_size < 6 ||
425 (work->hdr.ext_hdr_size) > work->hdr.frame_data_size) {
426 GST_DEBUG ("Invalid extended header. Broken tag");
429 work->hdr.ext_flag_bytes = work->hdr.frame_data[4];
430 if (5 + work->hdr.ext_flag_bytes > work->hdr.frame_data_size) {
432 ("Tag claims extended header, but doesn't have enough bytes. Broken tag");
436 work->hdr.ext_flag_data = work->hdr.frame_data + 5;
437 work->hdr.frame_data += work->hdr.ext_hdr_size;
438 work->hdr.frame_data_size -= work->hdr.ext_hdr_size;
441 start = GST_BUFFER_DATA (work->buffer);
442 frame_hdr_size = id3v2_frame_hdr_size (work->hdr.version);
443 if (work->hdr.frame_data_size <= frame_hdr_size) {
444 GST_DEBUG ("Tag has no data frames. Broken tag");
445 return FALSE; /* Must have at least one frame */
448 work->tags = gst_tag_list_new ();
450 while (work->hdr.frame_data_size > frame_hdr_size) {
451 guint frame_size = 0;
452 gchar frame_id[5] = "";
453 guint16 frame_flags = 0x0;
454 gboolean obsolete_id = FALSE;
455 gboolean read_synch_size = TRUE;
458 /* Read the header */
459 switch (ID3V2_VER_MAJOR (work->hdr.version)) {
463 frame_id[0] = work->hdr.frame_data[0];
464 frame_id[1] = work->hdr.frame_data[1];
465 frame_id[2] = work->hdr.frame_data[2];
468 obsolete_id = convert_fid_to_v240 (frame_id);
470 /* 3 byte non-synchsafe size */
471 frame_size = work->hdr.frame_data[3] << 16 |
472 work->hdr.frame_data[4] << 8 | work->hdr.frame_data[5];
476 read_synch_size = FALSE; /* 2.3 frame size is not synch-safe */
479 frame_id[0] = work->hdr.frame_data[0];
480 frame_id[1] = work->hdr.frame_data[1];
481 frame_id[2] = work->hdr.frame_data[2];
482 frame_id[3] = work->hdr.frame_data[3];
485 frame_size = id3v2_read_synch_uint (work->hdr.frame_data + 4, 4);
487 frame_size = GST_READ_UINT32_BE (work->hdr.frame_data + 4);
489 frame_flags = GST_READ_UINT16_BE (work->hdr.frame_data + 8);
491 if (ID3V2_VER_MAJOR (work->hdr.version) == 3) {
492 frame_flags &= ID3V2_3_FRAME_FLAGS_MASK;
493 obsolete_id = convert_fid_to_v240 (frame_id);
495 GST_DEBUG ("Ignoring v2.3 frame %s", frame_id);
500 work->hdr.frame_data += frame_hdr_size;
501 work->hdr.frame_data_size -= frame_hdr_size;
503 if (frame_size > work->hdr.frame_data_size || strcmp (frame_id, "") == 0)
504 break; /* No more frames to read */
506 /* Sanitize frame id */
507 switch (ID3V2_VER_MAJOR (work->hdr.version)) {
511 for (i = 0; i < 3; i++) {
512 if (!g_ascii_isalnum (frame_id[i]))
517 for (i = 0; i < 4; i++) {
518 if (!g_ascii_isalnum (frame_id[i]))
524 ("Frame @ %ld (0x%02lx) id %s size %u, next=%ld (0x%02lx) obsolete=%d",
525 (glong) (work->hdr.frame_data - start),
526 (glong) (work->hdr.frame_data - start), frame_id, frame_size,
527 (glong) (work->hdr.frame_data + frame_hdr_size + frame_size - start),
528 (glong) (work->hdr.frame_data + frame_hdr_size + frame_size - start),
530 #define flag_string(flag,str) \
531 ((frame_flags & (flag)) ? (str) : "")
532 GST_LOG ("Frame header flags: 0x%04x %s %s %s %s %s %s %s", frame_flags,
533 flag_string (ID3V2_FRAME_STATUS_FRAME_ALTER_PRESERVE, "ALTER_PRESERVE"),
534 flag_string (ID3V2_FRAME_STATUS_READONLY, "READONLY"),
535 flag_string (ID3V2_FRAME_FORMAT_GROUPING_ID, "GROUPING_ID"),
536 flag_string (ID3V2_FRAME_FORMAT_COMPRESSION, "COMPRESSION"),
537 flag_string (ID3V2_FRAME_FORMAT_ENCRYPTION, "ENCRYPTION"),
538 flag_string (ID3V2_FRAME_FORMAT_UNSYNCHRONISATION, "UNSYNC"),
539 flag_string (ID3V2_FRAME_FORMAT_DATA_LENGTH_INDICATOR, "LENGTH_IND"));
544 /* Now, read, decompress etc the contents of the frame
545 * into a TagList entry */
546 work->cur_frame_size = frame_size;
547 work->frame_id = frame_id;
548 work->frame_flags = frame_flags;
550 if (id3v2_parse_frame (work)) {
551 GST_LOG ("Extracted frame with id %s", frame_id);
553 GST_LOG ("Failed to extract frame with id %s", frame_id);
554 id3v2_add_id3v2_frame_blob_to_taglist (work, frame_size);
557 work->hdr.frame_data += frame_size;
558 work->hdr.frame_data_size -= frame_size;
561 if (gst_structure_n_fields (GST_STRUCTURE (work->tags)) == 0) {
562 GST_DEBUG ("Could not extract any frames from tag. Broken or empty tag");
563 gst_tag_list_free (work->tags);
568 /* Set day/month now if they were in a separate (obsolete) TDAT frame */
569 if (work->pending_day != 0 && work->pending_month != 0) {
572 if (gst_tag_list_get_date (work->tags, GST_TAG_DATE, &date)) {
573 g_date_set_day (date, work->pending_day);
574 g_date_set_month (date, work->pending_month);
575 gst_tag_list_add (work->tags, GST_TAG_MERGE_REPLACE, GST_TAG_DATE,