Update Protein Ontology regex specification

[nataled]

Prefix

pr

Explanation

The given regex in Bioregistry is incomplete. In general there are two types of local identifiers in PR:

• purely digital
• UniProtKB identifier (including isoforms)

The full regex is thus: ^(\d+|[OPQ][0-9][A-Z0-9]{3}0-9?|[A-NR-Z]0-9{1,2}(-\d+)?)$

Contributor ORCID

0000-0001-5809-9523

[cthoyt]

Awesome, thanks @nataled.

[nataled]

The regex given above is from the OBO PURL system, but I can see that it is not as precise as the one I use internally:

^(?:\d{9}|[OPQ][0-9][A-Z0-9]{3}0-9?|[A-NR-Z]0-9{1,2}(?:-\d+)?)$

This version specifies the allowed number of digits (nine) and also specifies that the parentheses are not intended for string capture. Perhaps this version will work better?

[cthoyt]

Yes, this one appears to work, but what is going on with the colons?

[nataled]

The ?: at the beginning of each parenthetical expression says "don't save the contents of the parentheses match". Those parenthetical expressions are there only because they represent optional parts. Taught to me and recommended by James Overton.

[nataled]

I noticed that the first regex appears to have links in it. When I did a copy/paste they showed up.