Opened 4 years ago
Last modified 4 years ago
#32540 closed Cleanup/optimization
Only do "top level" detection in DiscoverRunner.build_suite() when needed — at Version 1
| Reported by: | Chris Jerdonek | Owned by: | nobody |
|---|---|---|---|
| Component: | Testing framework | Version: | dev |
| Severity: | Normal | Keywords: | DiscoverRunner, build_suite, discovery |
| Cc: | Triage Stage: | Ready for checkin | |
| Has patch: | yes | Needs documentation: | no |
| Needs tests: | no | Patch needs improvement: | no |
| Easy pickings: | no | UI/UX: | no |
Description (last modified by )
Currently, DiscoverRunner.build_suite() is a bit confusing to read because it does "top level" detection for test discovery even when discovery won't be taking place.
My suggestion is to move the top-level detection logic (and large code comment) into a find_top_level(top_level) function that is placed immediately before where is_discoverable() is defined. Then, only call find_top_level() when needed, namely right before where self.test_loader.discover() is called.
Similarly, the kwargs = discover_kwargs.copy() line can also go into that if block right before that line, because that's the only code path where kwargs is actually used. Without this change, it's hard for one to notice that the kwargs are in fact only used for that one line.
Together, this will cut down on the size of build_suite() quite a bit and make it easier to understand.