[ADAM-1768] Add InFormatter for unpaired FASTQ.

[fnothaft]

Resolves #1768. Still testing; not ready for merge. Thoughts on how to handle the configuration properties that are shared with InterleavedFASTQInFormatter? I'd rather not duplicate them.

[heuermh]

Single is kinda redundant, how about FASTQInFormatter?

[fnothaft]

SGTM!

[heuermh]

Thoughts on how to handle the configuration properties that are shared with InterleavedFASTQInFormatter? I'd rather not duplicate them.

The way you have them now, keyed by fields on FragmentRDD, seems fine to me. Or are you asking a different question?

[AmplabJenkins]

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/ADAM-prb/2436/
Test PASSed.

[fnothaft]

The way you have them now, keyed by fields on FragmentRDD, seems fine to me. Or are you asking a different question?

No, that's exactly what I'm asking. I guess what I don't like about having them in FragmentRDD is that now you're using them with AlignmentRecordRDD in addition to FragmentRDD. Just seems a bit counterintuitive/awkward for a user.

[AmplabJenkins]

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/ADAM-prb/2437/
Test PASSed.

[heuermh]

No, that's exactly what I'm asking. I guess what I don't like about having them in FragmentRDD is that now you're using them with AlignmentRecordRDD in addition to FragmentRDD. Just seems a bit counterintuitive/awkward for a user.

I see. How about adding those key fields to both AlignmentRecordRDD and FragmentRDD?

Alternatively, we could put them onGenotypeRDD and then refer by specific subclass (AlignmentRecordRDD.WRITE_SUFFIXES) but that is somewhat poor code style.

[heuermh]

Created fnothaft#21 to address comment above

[fnothaft]

Alternatively, we could put them on GenotypeRDD and then refer by specific subclass (AlignmentRecordRDD.WRITE_SUFFIXES) but that is somewhat poor code style.

I assume that by GenotypeRDD, you meant RecordGroupGenomicRDD?

[fnothaft]

I'm fine with fnothaft#21 for now, so let's go with that.

[fnothaft]

@heuermh Do you want me to squash down your commit into mine, or leave it separate?

[AmplabJenkins]

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/ADAM-prb/2454/
Test PASSed.

[heuermh]

Squash is ok

[fnothaft]

Squash is ok

Sounds good! I am going to test this later today, and will squash this down once it is OK to merge.

[fnothaft]

I've tested this and it is good to go. Squashing now.

[AmplabJenkins]

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/ADAM-prb/2456/
Test PASSed.

[heuermh]

Thank you, @fnothaft