[WIP] Ann io

[joshuak94]

This is a very first draft/WIP for the annotation IO. This will cover annotation file types (BED, bedGraph, wiggle, etc.).

At the moment I've only implemented a very basic BED format (three columns, chrom, chromStart, chromEnd) and the BED header.

TODO

• Extend BED3 format to full BED format w/ optional fields.
• Implement the writer.
• Add support for at the very least, bigbed (binary & indexed BED), wig, and bigwig files.
• Allow reading BAM files as annotation files??

[h-2]

Thanks for working on this first of all! The structure seems fine to me!

Regarding the name: I still think that "ann_io" is sounds like the name of a secret agent 🕵🏻‍♀️ but I also don't have a great replacement, so we can stick with this for now. We can also discuss with @smehringer to see what she thinks.

Note that I have changed a small thing about reader_base in #47, so the inheritance works a little different now.

[h-2]

I know these use camelCase in the specification, but it looks very strange to have that mixed with the other formatting styles in this library.
Can we change this to having snake_case or do you think that will confuse users?

[joshuak94]

I could change it! I just did it this way because I wasn't sure if I should be consistent with the specs or with our code haha.

[h-2]

Yeah, I think we should stick with the code style for now. I would love to have a table like this in the documentation at some point:

bio::field::	bio::fasta	bio::fastq	bio::vcf	bio::bcf	bio::sam	bio::bed
::id == ::qname	description line	description line	ID	ID	QNAME	name
::seq	sequence data	sequence data	–	–	SEQ	–
::chrom == ::rname	–	--	CHROM	CHROM	RNAME	chrom
::qual == ::mapq	–	quality data	QUAL	QUAL	MAPQ	–
::pos == ::chrom_start	–	–	POS	POS	POS	chromStart

...

So, instead of having individual documentation for all the fields, one big table with the format-specific terminology would be more helpful I think.

[smehringer]

For the naming I opened a discussion thread: #51