GoRule 57 update and test for issue #473

[dustine32]

@dougli1sqrd I'll probably need to update the mgi.yaml and wb.yaml files in go-site as well, right? Example:

filter_for:
  provided_by:
    - MGI

[kltm]

What's the difference between filter_for and filter_out? We'll also need to add this to the schema (where we also probably want to document).

[dustine32]

@kltm Oh yeah, I forgot to update the datasets.schema.yaml, thanks! I'll make a ticket and PR in go-site for this.

Existing filter_out defines annotation values that would trigger a line to be filtered out, whereas filter_for (maybe we should change this to filter_in) defines annotation values that a line must have to prevent being filtered out.

The example in issue #473 shows filter_for being used to filter the MGI GPAD, keeping only the lines where Provided_by=MGI and excluding all other lines (e.g. UniProt, SynGO).

[kltm]

@dustine32 Just as a note, I'm fine with properties that have long_and_human_readable_names.
That said, whatever works for you, I just want to make sure that people reading the metadata without all of the context have some idea what's going on (at some point in the future).

[dustine32]

Linking schema PR: geneontology/go-site#1554

[dustine32]

Thinking about changing filter_for to only_allow as @dougli1sqrd suggested. Still brainstorming though.