signature name multifasta input

[phiweger]

When using multifasta, the --singleton option of sourmash compute is really useful. However, when one later queries each signature against an SBT, e.g. with sourmash categorize, the output format is:

filename, some_hit_id, similarity
filename, another_hit_id, similarity
...

Is there an option to carry the fasta header, so that it appears instead of the same filename againandagainandagain? I tried --name-from-first, but I think it was not meant for this purpose right?

Thx!

[phiweger]

To make this more clear perhaps:

Given a multifasta

>FOO
ACTGATAC
>...
...

and after sourmash compute --singleton, sourmash categorize reports to stderr something like

loaded query: FOO ... (k=16, DNA)
for FOO, found: 0.9 BAR

However, the --csv flag will write the following information to disk:

multifasta.json,BAR,0.9

So what I do is parse the stderr, which is a bit inconvenient. This would be nice:

FOO,BAR,0.9

[ctb]

@phiweger could you take a look at #421 and see if it fixes this? Apologies to take so long!