[Feature]: Use standard in/out

[MillironX]

Have each of the haplink commands read from standard input and write to standard output unless appropriate file flags are called.

Additional Removed Arguments

haplink variants

• --bam
  • Can be replaced by stdin in SAM format
• --output
  Can be replaced by stdout in VCF format

haplink haplotypes

• --variants
  • Can be replaced by stdin in VCF format
• --output
  • Can be replaced by stdout in YAML format

haplink sequences

• --haplotypes
  • Can be replaced by stdin in YAML format
• --output
  • Can be replaced by stdout in FASTA format

Example Usage

Command

minimap2 -ax map-ont --MD example/reference.fasta reads.fastq \
  | tee output.sam \
  | haplink variants --reference example/reference.fasta \
  | tee output.vcf \
  | haplink haplotypes  --bam output.sam \
  | haplink sequences --reference example/reference.fasta \
  > output.fasta

Context

Most bioinformatic tools work with standard in/out. It would be convenient if HapLink did this, too, as it could then be added to cli pipelines for more efficient work.

Possible Implementation

Each of the stdin/stdout parameters will need to be made optional, with manual checks for stdin. Also, read/write operations will need to open stdin/out instead of files.

The big problem here is that https://github.com/BioJulia/XAM.jl does not support a unified API for SAM and BAM records (see BioJulia/XAM.jl#25). Although https://github.com/samtools/htslib and related tools do not seem to care, https://github.com/genome/bam-readcount does require an indexed BAM file, so there will need to be a way to create, sort, and index a BAM file is SAM input is given. There will still need to be a way to sort and index BAM input if given via stdin.

[MillironX]

Partially implemented with #35. Stdout is used unless a parameter is passed, but stdin is more tricky.