Add support for parsing chapters from MKV files

[tymmesyde]

Hi,
this include changes to support parsing MKV chapters using the recent changes to introduce metadata Chapter interface
This is already implemented for MP4: Nero & QuickTime
Used this doc as reference: https://www.matroska.org/technical/elements.html

Added sample_with_chapters.mkv for mkvSample_withChapters test
Generated with:

;FFMETADATA1
[CHAPTER]
TIMEBASE=1/1000
START=0
END=1999
title=Chapter 1

[CHAPTER]
TIMEBASE=1/1000
START=2000
END=4999
title=Chapter 2

ffmpeg \                                                                                       
  -f lavfi -i testsrc=duration=5:size=320x240:rate=30 \
  -f lavfi -i sine=frequency=440:duration=5 \
  -i chapters.txt \
  -map 0:v -map 1:a -map_chapters 2 \
  -c:v libx264 -c:a aac \
  sample_with_chapters.mkv

[icbaker]

I can't work out how to delete the stale comment made against an old copy of the code, sorry for the duplication.

[icbaker]

I didn't find this documented in https://datatracker.ietf.org/doc/rfc9559/ - where is this defined?

[tymmesyde]

I don't see it there, i followed this spec: https://www.matroska.org/technical/elements.html
So i guess it's not a RFC standard? Should i remove related logic?

[tymmesyde]

It is specified in the sources tho: https://github.com/ietf-wg-cellar/matroska-specification/blob/master/ebml_matroska.xml#L1521, probably was not deployed to the webiste

[icbaker]

This is documented in https://datatracker.ietf.org/doc/rfc9559/ with range: not 0 (1-18446744073709551615) which I think means:

1. We should use long here (since 18446744073709551615 = 2^64 - 1 - so actually even Java signed long kinda isn't large enough if you consider signedness - but better than int)
2. We should use 0 as the 'unset' value, not Format.NO_VALUE which is -1 in Java signed int math, which is18446744073709551615 as a 64-bit unsigned int (i.e. a technically valid value).

[srgsf]

it is defaulted to C.TIME_UNSET but you're comparing it with Format.NO_VALUE

[srgsf]

again, C.TIME_UNSET should be used.