- Move field alias handling, shared-table qualification, and path pattern
normalization into FieldQuery.from_model with a query-specific
normalize_pattern hook.
- Update query parsing and plugin call sites to construct queries through
query_cls.from_model and remove redundant model-level field query helpers.
This keeps model entrypoints focused on orchestration and colocates
transformation behavior with the query types that own it.
It seems like previously filtering by flexible attributes did not work
- I'd receive '{"data": []}' trying to GET `/aura/tracks?filter[play_count]=11`
Now this works, not only for tracks, but for `/aura/artists` and
`/aura/albums` too.
Additionally, this improves `/aura/tracks` response time significantly.
I tried loading the default of 500 tracks from my library:
On `master`, it took ~20s
After this commit, it takes under 1s.
This will be help with testing each of the documents which do not
any more depend on the 'global' `current_app` and `request`. These two
can now be provided at the time the objects are instantiated.
When constructing paths to image files to serve, we previously spliced
strings from URL requests directly into the path to be opened. This is
theoretically worrisome because it could allow clients to read other
files that they are not supposed to read.
I'm not actually sure this is a real security problem because Flask's
URL parsing should probably rule out IDs that have `/` in them anyway.
But out of an abundance of caution, this now prevents paths from showing
up in IDs at all---and also prevents `.` and `..` from being valid
names.
displayable_path may remove 'bad' characters, yielding a wrong path.
Also use track.path rather than track.destination() as that is where
the file is actually located rather than where it should be located
according to the beets path system.
