Nils Homer

@msto I'd like to review this one before it gets merged

Does this belong in pysam, or in htslib and exposed in pysam? Have we asked them if they would accept a contribution? And if not, why?

I am ok with this. @tfenne ?

Sgtm

One solution would be to override the `values` method when writing. Another solution would be to have a parent class with _only_ the attributes written to file, and a sub-class...

For Illumina, we could have up to four separate FASTQs: R1, R2, I1, and I2. In some cases, we have multiple files for R1. So really, rather than paired, I...

I like it, but would only make one modification that we can certainly sequence fragment reads and a single (or dual) barcodes. E.g. R1/I1, or R1/I1/I2. So I would make...

I think it's a footgun to have it default to a SAM file type if writing to standard output in library code. Let's go without it defaulting to SAM please.

I think the follow-up question deserves a separate issue. See comment in the code https://github.com/pysam-developers/pysam/issues/655 and a similar issue our other rust repo: https://github.com/fulcrumgenomics/fgoxide/pull/9.

It might be useful to see what others (reference impls) have done to merge headers and seqdicts, as well as tools themselves (e.g. `samtools merge`). You may * htsjdk: [here](https://github.com/samtools/htsjdk/blob/master/src/main/java/htsjdk/samtools/SamFileHeaderMerger.java)...