Reversing the unsynchronisation seems to work slightly differently
for ID3 v2.3 tags and v2.4 tags: v2.3 tags don't have syncsafe frame
sizes in the frame header, so the unsynchronisation is applied to
the whole frame data including all the frame headers. v2.4 frames
have sync-safe sizes, however, so the unsynchronisation only needs
to be applied to the actual frame data, and it seems that's what's
being done as well. So we need to undo the unsynchronisation on a
per-frame basis for v2.4 tags for things to work properly.
Fixes extraction of coverart/images from APIC frames in ID3 v2.4
tags (#588148).
Add unit test for this as well.
We didn't handle unsynchronization at all up to now, which might have
caused frames to not be extracted - esp. frames after an APIC picture
frame. Fixes#577468.
Original commit message from CVS:
* gst/id3demux/id3tags.c:
* gst/id3demux/id3tags.h:
* gst/id3demux/id3v2frames.c: (id3v2_tag_to_taglist):
We don't want the same string multiple times in a tag list for the
same tag ever, for any tag, not just for GST_TAG_GENRE, so make sure
this doesn't happen and remove special-case code for GST_TAG_GENRE.
Original commit message from CVS:
* gst/id3demux/gstid3demux.c:
* gst/id3demux/gstid3demux.h:
* gst/id3demux/id3tags.c:
* gst/id3demux/id3tags.h:
* gst/id3demux/id3v2frames.c:
Port ID3 tag demuxer over to the new GstTagDemux in -base
(now would be a good time to test re-importing your music
collection).
Original commit message from CVS:
* gst/id3demux/id3tags.c: (id3demux_id3v2_frames_to_tag_list):
* gst/id3demux/id3tags.h:
* gst/id3demux/id3v2frames.c: (id3demux_id3v2_parse_frame),
(parse_obsolete_tdat_frame):
Do not convert obsolete TDA/TDAT frames to TDRC frames, otherwise
the four-digit number will be interpreted as a year, whereas it is
month and day in DDMM format. Instead, parse TDAT frames and fix up
the date in the GST_TAG_DATE tag later if we also extracted a year.
Fixes#407349.
Original commit message from CVS:
* gst/id3demux/id3tags.c:
(id3demux_add_id3v2_frame_blob_to_taglist):
Extract frames for ID3v2 versions prior to ID3v2.3.0 properly as
well, and add the version to the blob's buffer caps, since that
information will be needed for deserialisation later on (#348644).
Original commit message from CVS:
* gst/id3demux/gstid3demux.c: (plugin_init):
* gst/id3demux/id3tags.c:
(id3demux_add_id3v2_frame_blob_to_taglist):
* gst/id3demux/id3tags.h:
On second thought, it might be wiser and more efficient
not to do tag registration from a streaming thread.
Original commit message from CVS:
* gst/id3demux/id3tags.c:
(id3demux_add_id3v2_frame_blob_to_taglist),
(id3demux_id3v2_frames_to_tag_list):
Put ID3v2 frames we can't parse as binary blobs into private
tags, so that they are not lost when retagging, at least once
id3v2mux has been taught to re-inject those frames again.
See bug #334375.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_parse_stream),
(gst_avi_demux_process_next_entry):
Fix some leaks.
* gst/id3demux/id3tags.c: (id3demux_id3v2_frames_to_tag_list):
Don't use \n in debug lines.
Original commit message from CVS:
* gst/apetag/gsttagdemux.c: (gst_tag_demux_chain):
* gst/id3demux/gstid3demux.c: (gst_id3demux_chain):
Don't attempt typefinding on too-short buffers that have been
completely trimmed away.
* gst/id3demux/id3tags.c: (id3demux_read_id3v2_tag):
Improve the debug output
Original commit message from CVS:
* gst/id3demux/id3tags.c: (id3demux_id3v2_frames_to_tag_list):
* gst/id3demux/id3v2frames.c: (id3v2_genre_fields_to_taglist):
Handle 0 data size in otherwise valid frames.
Handle numeric strings in 2.4.0 even when not in parentheses
Original commit message from CVS:
* gst/id3demux/id3tags.c: (id3demux_id3v2_frames_to_tag_list):
ID3 2.3.0 used synch-safe integers for the tag size, but not for the
frame size. (Fixes#331368)
Original commit message from CVS:
* gst/id3demux/gstid3demux.c: (gst_id3demux_chain),
(gst_id3demux_read_id3v1), (gst_id3demux_sink_activate),
(gst_id3demux_send_tag_event):
* gst/id3demux/id3tags.c: (id3demux_read_id3v1_tag):
Someone should kick my butt. Remove ID3v1 tags from the end of the
file.
Improve error messages. Send the TAG message as soon as we complete
typefinding, instead of waiting until we send the first buffer.
Downstream tag event is still sent before the first buffer.
Original commit message from CVS:
* gst/id3demux/id3tags.c: (id3demux_read_id3v2_tag):
* gst/id3demux/id3tags.h:
* gst/id3demux/id3v2frames.c: (id3demux_id3v2_parse_frame),
(parse_comment_frame), (parse_text_identification_frame),
(id3v2_tag_to_taglist), (id3v2_are_digits),
(id3v2_genre_string_to_taglist), (id3v2_genre_fields_to_taglist),
(parse_split_strings), (free_tag_strings):
Rewrite parsing of text tags to handle multiple NULL terminated
strings. Parse numeric genre strings and ID3v2 type
"(3)(6)Alternative" style genre strings.
Parse dates that are only YYYY or YYYY-mm format.
Original commit message from CVS:
* gst/id3demux/id3tags.c: (id3demux_read_id3v2_tag):
If a broken tag has 0 bytes payload, at least still skip
the 10 byte header