ENH: add support for --addfragments flag

[misialq]

Adds supports for --addfragments flag for addition of fragmentary sequences to existing alignments.

Since it's just a small variant of mafft_add, rather than creating a new action I introduced a new parameter in the existing one that will determine whether the --add flag (current behaviour) or --addfragments one should be used.

[mikerobeson]

Rename fragments to short_fragments and default to False?

and rephrase to something like:
'Optimize for the addition for short sequence fragments. For example, primer or amplicon sequences'

[thermokarst]

I like taking the approach of ensuring the plugin parameter name matches the underlying tool flag name - so in this case, addfragments. I like to think that this helps minimize "translation" confusion when trying to figure out how a wrapper-plugin (like this one) does what it does. What do you think, @mikerobeson?

[mikerobeson]

Gotcha. That makes sense to me. fragments it is. :-)

[thermokarst]

I'm proposing a third option: addfragments (instead of fragments, or short_fragments), that way the param names match!

[mikerobeson]

I like addfragments. I suppose this would be defaulted False? That is it would do add by default?

[misialq]

Alright, changed the param name to addfragments. Not passing the param flag would default to a plain add.

[mikerobeson]

I was just wondering if we should be a tiny bit more explicit to avoid confusion:

'addfragments': 'Optimize for the addition of short sequence '
                'fragments (for example, primer or amplicon '
                'sequences). If not set, default sequence addition '
                'is used.'

[misialq]

True, good point - updated.

[mikerobeson]

Everything looks fine to me. Just wanted to voice my approval. :-)

[thermokarst]

This is fantastic, thanks so much for this @misialq!

cc @gregcaporaso

[gregcaporaso]

Thanks so much for the contribution @misialq and for the review @mikerobeson! I have a use case for this now, so very excited to try it out!