updating to newer htsjdk snapshot

[lbergelson]

updating to a new htsjdk snapshot and updating our VariantContextWriters to compile

test may fail on this....

[droazen]

👍 looks good to me

[lbergelson]

@jean-philippe-martin We're seeing the same errors here as in yours pr. So it's something in the htsjdk update...

[cmnbroad]

@lbergelson @jean-philippe-martin I looked into this it a bit - it can be readily reproduced if the failing test is run through gradle, but not from IntelliJ, which is because when it runs through gradle, its using asyncIO for BAM writing, but when run from IntelliJ its using sync io.

I'm not sure why this issue is showing up now, but in the cases that fail, the async writer thread is being interrupted by this code which is called from the main thread when the PrintReads closeTool method closes the writer (I added a trace statement to verify this). When I run in IntelliJ and force async io to be used, the interrupt call still happens, but there is no test failure.

[cmnbroad]

One short term option would be to turn off async writing until we figure out the right fix.

[lbergelson]

The regression was introduced in samtools/htsjdk@1ade6c9

[codecov-io]

Codecov Report

Merging #3588 into master will increase coverage by 0.242%.
The diff coverage is 100%.

@@              Coverage Diff               @@
##              master    #3588       +/-   ##
==============================================
+ Coverage     79.738%   79.98%   +0.242%     
- Complexity     18153    18896      +743     
==============================================
  Files           1221     1222        +1     
  Lines          66623    68248     +1625     
  Branches       10411    10916      +505     
==============================================
+ Hits           53124    54585     +1461     
- Misses          9290     9367       +77     
- Partials        4209     4296       +87

Impacted Files	Coverage Δ	Complexity Δ
...e/hellbender/utils/variant/writers/GVCFWriter.java	93.258% <100%> (+0.155%)	33 <1> (+1)	⬆️
...ute/hellbender/tools/spark/pathseq/PSTreeNode.java	87.736% <0%> (-0.576%)	38% <0%> (+20%)
...ools/spark/pathseq/PSFilterArgumentCollection.java	79.487% <0%> (-0.513%)	2% <0%> (ø)
...nder/tools/spark/pathseq/PSPathogenTaxonScore.java	78.125% <0%> (-0.057%)	22% <0%> (+3%)
...ecaller/AssemblyBasedCallerArgumentCollection.java	100% <0%> (ø)	2% <0%> (+1%)	⬆️
...tools/spark/pathseq/PSScoreArgumentCollection.java	100% <0%> (ø)	1% <0%> (ø)	⬇️
...r/tools/spark/pathseq/HostAlignmentReadFilter.java	100% <0%> (ø)	17% <0%> (+11%)	⬆️
...umber/allelic/alleliccount/AllelicCountWriter.java	100% <0%> (ø)	2% <0%> (ø)	⬇️
...ender/utils/smithwaterman/SWPairwiseAlignment.java	86.1% <0%> (ø)	65% <0%> (?)
.../copynumber/allelic/alleliccount/AllelicCount.java	65.854% <0%> (+0.336%)	19% <0%> (+8%)	⬆️
... and 15 more

[lbergelson]

the htsjdk regression was fixed in samtools/htsjdk#992

[lbergelson]

@jean-philippe-martin I've updated the htsjdk snapshot, you should be unblocked after a rebase.

[jean-philippe-martin]

Thank you very much @lbergelson !