Support reading CRAM files with embedded references

[athos]

Note: This PR will be marked as ready for review after #306 has been merged.

This PR allows the CRAM reader to handle files with embedded references. With this update, we no longer need to specify any reference file to read CRAM files with embedded references.

Specification

According to the CRAM specification, if the encoder embeds a portion of the reference sequence into a slice, it allocates a dedicated block to store the raw (ASCII character) base sequence data for the reference sequence covering that slice. The content ID of this allocated block is pointed to by the "embedded reference bases block content ID" field in the slice header. If there is no embedded reference for the slice, this field's value will be -1. (See §8.5 for more details)

Implementation

In this change, the CRAM reader first checks the "embedded reference bases block content ID" field in the slice header to identify the block allocated for the embedded reference. If this value is positive, it creates a dedicated seq resolver based on the content of the block associated with the specified content ID. This new seq resolver is responsible for resolving reference sequences covered by the embedded reference and helps with the decoding of CRAM records in the slice.

Notes

This PR also makes some slight changes to the signature of the seq resolver protocol method to make the new seq resolver for embedded reference a little bit more efficient.

Note that this PR does not include additional test code, although you can easily verify the change using a simple snippet with a test file 0600_mapped.cram from hts-specs (which is documented as being encoded with embedded references) as follows:

(require '[cljam.io.cram :as cram])

(with-open [cram-rdr (cram/reader "0600_mapped.cram")]
  (= ["ATTTTTCGGGTTTTTTGAAAATGTAGCTACAGAAAGTTTGTTTAAATATCTTGTTTTCTTGCACTTTGTGCAGAATT"
      "CCCTTTCAGAAAAATTATTTTTAAGAATTTTTCATRYTAGGAATATTGTTATTTCAGAAAATAGCTAAATGTGATTTCTGTAATTTTGCCTGCTAAAGGG"]
     (map :seq (cram/read-alignments cram-rdr))))

[codecov]

Codecov Report

Attention: Patch coverage is 47.36842% with 10 lines in your changes are missing coverage. Please review.

Project coverage is 88.40%. Comparing base (11d39b0) to head (fae8525).

Files	Patch %	Lines
src/cljam/io/cram/reader.clj	33.33%	10 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #307      +/-   ##
==========================================
- Coverage   88.50%   88.40%   -0.11%     
==========================================
  Files          95       95              
  Lines        8214     8228      +14     
  Branches      506      506              
==========================================
+ Hits         7270     7274       +4     
- Misses        438      448      +10     
  Partials      506      506              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[alumi]

Thank you for working on this feature! I added some comments. Please take a look 🙏

[alumi]

Thank you for updating! LGTM 👍

[athos]

Rebased onto the latest master branch.

[athos]

Rebased onto the latest master branch.

[matsutomo81]

Thank you for this update. LGTM 👍

I haven't merged it because I thought you might want to handle the commit integration.

[athos]

Thank you both for reviewing! I've squashed the commits and merged the change.