Updated DB + minor bug fix

If you explicitly provided a --scheme
AND
You got a null result (no matches)
THEN
The TSV ourput would have 7 empty allele columns instead of the actual number of alleles in the scheme.

Updated databases

Just updated the databases.

SO here is a poem:

Wun-Wun was a race horse
Tutu was one too
Wun-Wun won one
And Tutu won one too

The dog ate my homework

Should just be a long overdue update to the scheme/allele database.

And 2.22.0 is well suited to the year 2022.

It's what Jason wanted

• The --json now outputs a JSON array in case of multiple file inputs.
• Handle version parsing of BLAST+ 2.10
• Update all PubMLST bundled databases
• Document that scheme folder and scheme.txt file must be the same name
• Fix typos

Correctness is a novelty

• --novel out.fa now works correctly

• Now needs Perl 5.26 and BLAST+ 2.9.0

• New dependency any2fasta - no more temp files needed

• New script mlst-get_seqs -s SCHEME -t ST > genes.ffn

• Replace some bespoke functions with core Perl modules

• Latest databases, some new schemes

• Documentation updates

• More TRAVIS tests

Bad acs gone

• updated pubmlst databases
• a bunch of new schemes eg. mycobacteria
• removed faulty paralog acs genes in paeruginosa scheme

Guess what - 14 is unlucky too!

Let's also pretend 2.14 never existed. I think we are back to normal now.

• removed any -culling_limit as it is just wrong for this use case
• set --minscore=0 when --scheme specified to ensure any allele matches are shown
• don't print out rejected alleles unless --debug on
• added some new MOTDs
• don't print number of exclusions when there are none.

I guess 13 is unlucky after all

I left a major bug in 2.13 and have fixed it here I hope.
Let's pretend it never existed.

Limit the culling

• remove dependency on file
• add back -culling_limit 1 to avoid erroneous allele subhits (also increases speed)
• add latest pubmlst
• fix bug in download scripts mlst-download_pubmlst script.

Revenge of the null allele

• update pubmlst to 18 Aug 2018
• new vcholerae2 scheme
• multimatch alleles are now numerically sorted
• handle new null alleles eg pstS(0) in efaecium
• ensure fungal schemes are deleted