BEP-0008 - Update Namespace and Identifier Handling

[cthoyt]

After discussion with @sgebel, I've written this BEP to suggest an improvement to the <namespace>:<name> used in BEL terms. See the full proposal at https://github.com/belbio/bep/blob/add-bep-0008-obo-identifiers/docs/drafts/bep-0008.md

I propose that we (optionally) adopt the OBO syntax for references (<namespace>:<identifier>!<name>) anywhere that the old references were valid.

For example, p(HGNC:MAPT), could be optionally written using the OBO syntax of p(HGNC:6893!MAPT), which includes both the identifier (6893) and the label (MAPT) of the human Tau protein. See the BEP for more examples and rationale.

[wshayes]

I like this - great idea Charlie and @sgebel the name is an optional element so it is backwards compatible. In presentation mode, we can hide the id and just show the name - with a rollover to show the full official name. This solves some of the terminology evolution issues I have with name identifiers vs database id's.

I'm not a fan of the GO:GO:1234 - let's try to come up with a better way of handling identifiers.org ids and namespaces.

[cthoyt]

I think it's reasonable to bake in logic that if an identifier starts with the prefix and a semicolon, then one can be omitted. It would take only trivial logic (str.startswith()) when looking up information from MIRIAM on each one. But if we agree on that, let's make it explicit in this BEP

[wshayes]

That sounds good to me

[cthoyt]

Before this BEP is reviewed, I will submit another BEP for yet another way to define namespaces using MIRIAM, which will support this

[wshayes]

@cthoyt Please review the character/regex definitions. I'd like to make sure we can build a functional parser without too much difficulty before releasing this to production (e.g. the Reference Implementation). I'm planning on redoing my parser, but it is behind 3 higher priority tasks so it will be a month or so before I can do the reference implementation).