keithj / cl-sam Goto Github PK

It fails to compile in the latest quicklisp, because it relies on CFFI declaring a version, but apparently non-release copies of CFFI, including the one in quicklisp, do not declare a :version.

Use the template/fragment notation instead of read/forward/reverse in function and argument names.

The reference sequence name should match the regex in the spec, or raise an error.

The new CIGAR string operations '=' and 'X' should be supported.

1.3 introduces a non-backwards compatible change in permitted sequence characters; from [NTGCA=]+ to [A-Za-z=.]+. The current validation is too strict.

Permitted read names have changed too. @ character is no longer allowed. Limited to between 1 and 255 chars.

The sanity checks for the bitwise flags added in 1.3 should be used to avoid raising errors.

Great repo! Is it still being maintained? I forked it to https://github.com/biolisp

Hi,

The find-bam-index/1 fails for me in LispWorks. The reason is that the pathnames returned by merge-pathnames on the .bai files and the pathnames returned returned by find-bam-index don't exactly match. However, when comparing their namestrings, everything is fine. I changed the three ensure forms to this:

(ensure (equal (namestring (merge-pathnames ...))
(namestring (find-bam-index (merge-pathnames ...)))))

Then this test passes for me in LispWorks. All the other tests already pass anyway.

Best,
Pascal

Add restarts to allow recovery from bad headers or clashes during header merging.

There should be functions to build PG chains from PP tags, to raise errors on orphan PP tags and to merge PP tag sets.