Have filenameIdentify use datasetContext to determine which filename rules to use.

[rwblair]

Currently is applying derived filename rules when it shouldn't be. This makes it difficult to notify a user that has a derivatives dataset mistakenly listed in dataset_description as a derivative. Something that may have never happened before. I'm fine overlooking that for now.

Should we have deriv datasets not apply any schema.rules.files.raw rules? I think thats why we went through all the trouble of making $refs work for filenames in the schema.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 86.59%. Comparing base (dc61d0f) to head (c5b348e).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1939      +/-   ##
==========================================
+ Coverage   85.58%   86.59%   +1.01%     
==========================================
  Files          91      130      +39     
  Lines        3768     5970    +2202     
  Branches     1207     1484     +277     
==========================================
+ Hits         3225     5170    +1945     
- Misses        457      709     +252     
- Partials       86       91       +5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[effigies]

Should we have deriv datasets not apply any schema.rules.files.raw rules?

Raw files are valid in derivative datasets, so there shouldn't be a need to prune them. They may be redundant with the more lax derivatives rules, however, assuming we're thorough enough in keeping derivative counterparts to all raw rules.

[effigies]

This LGTM. Don't know if your question and my answer impact whether you want to do anything else. Go ahead and merge, if not.

[rwblair]

Given how marginal the performance gains from pruning the deriv branch were I'll leave it how it is right now.