Set tab delimiter in manpage for tabix GFF3 sort

[cmdcolin]

This can help if there are spaces in the GFF3 file e.g. in column 2 or 3. Was found in the wild by my co-worker on a gff file here https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/002/285/GCF_000002285.3_CanFam3.1/GCF_000002285.3_CanFam3.1_genomic.gff.gz

Another alternative but similar command is

awk '$1 ~ /^#/ {print $0;next} {print $0 | "sort -t\"\t\" -k1,1 -k4,4n"}' file.gff > file.sorted.gff

But, the PR here keeps it as is

[daviesrob]

This is a good idea, but $'\t' to get the tab is a bash-ism. "`printf '\t'`" is more portable, and even works in csh.

I quite like the awk version as it only makes one pass through the file, although sadly it seems to be slightly slower than making two passes with grep. Presumably that's because awk is doing more work splitting the input up.

[cmdcolin]

as long as we are looking at compatibility, the current one actually has issue in zsh (needed the ^ to be inside of quotes I think) so updated again :)

also incorporated the printf \t! good catch

[cmdcolin]

also interesting that the awk is a bit slower. also, if could be helpful to make typing out the command easier. for that, can potentially offer people a bash command like:

function gffsort() {
  grep "^#" $1;
  grep -v "^#" $1 | sort -t"`printf '\t'`" -k1,1 -k4,4n;
}

then

gffsort input.gff | bgzip> out.gff.gz
tabix -p gff out.gff.gz

[daviesrob]

The update is mostly OK, apart from the \t got interpreted as a troff macro, and disappeared from the formatted output. I'll push a tiny fix.

The slowness with awk is most likely due to it splitting the entire line into fields. grep can get away with just looking at the first character, which is much faster.