Accession order fixes

[cbielow]

Enhances scripts/find_next_available_number.py to report the gap size and also now detects an unsorted CV (which would cause reporting of false positive gaps).

e.g.

Found gap of size 62: ('MS', 1000938) to ('MS', 1000999)
...
Found gap of size 1: ('MS', 1003236) to ('MS', 1003236)
Found gap of size 2: ('MS', 1003407) to ('MS', 1003408)               <-- false positive...
Found gap of size 2996587: ('MS', 1003413) to ('MS', 3999999)
Found gap of size 1: ('MS', 4000011) to ('MS', 4000011)
Found gap of size 25: ('MS', 4000025) to ('MS', 4000049)
CV is not sorted! MS:1003407 found after MS:4000181
Found gap of size 1000965: ('PEFF', 1036) to ('PEFF', 1002000)

Ordering of accessions is restored as well in this PR (eliminating CV is not sorted! MS:1003407 found after MS:4000181).

This might be worth adding to GH actions? i.e. fail if not sorted?

[cbielow]

Do we need to increase the version number? It's just a re-ordering of terms. Content is identical.

[edeutsch]

yes, I think the GitHub actions/checking requires every change to have a bump in version number.

[cbielow]

good to go

[mobiusklein]

Good catch, I had not considered the CV being unsorted by accession when I wrote the niche finder.

It is highly improbable that this breaks things for people directly consuming the OBO file.