CRAM validation / test files #512
Conversation
jkbonfield
force-pushed
the
CRAM_tests
branch
4 times, most recently
from
600f8ca
to
461f136
Compare
jkbonfield
changed the title
jkbonfield
force-pushed
the
CRAM_tests
branch
5 times, most recently
from
4e60e79
to
da888d6
Compare
The files are numbered such that a natural sort order will mean the earlier files may be decodable without yet being able to decode any of the later files. Therefore this may optionally be considered as a roadmap for development of new CRAM decoders. Indices are also tested.
|
Finally beaten the .md file into some reasonable state. I think I'm happy with this as a starting point for CRAM 3.0. CRAM 3.1 would just be a few new codecs included (which are tested outside of CRAM in their own repository already), so it'd be easy enough to add them when we're ready. What I don't have are failure tests, but they're not as vital as the "passed" ones. I've included pre-made CRAMs and the samtools decoded versions of their SAM files. We probably want something similar for BAM too, but I'm all worn out in specs testing land atm. You can see the formatted documentation here: https://github.com/jkbonfield/hts-specs/blob/CRAM_tests/test/cram/3.0/CRAM.md |
|
As discussed on the File Formats call 2021-04-01 between @jmarshall, @lbergelson, @tcezard and @tskir, this PR will be merged in its current form in four weeks (2021-04-29), unless there are strong objections. |
The files are numbered such that a natural sort order will mean the earlier files may be decodable without yet being able to decode any of the later files. Therefore this may be considered as a roadmap for development of new CRAM decoders.
Not yet complete, but obviously better than what was there before. :-)
I made the PR so people can provide feedback while this is progressing.