Coding 2022-07-02
I was poking at MOTR with new metrics stuff, and it had some recommendations, but making such changes isn't worth it until I've done a round of normal refactoring. I have a few other things I want to do today, so I'm going to try to get through this entry quickly.
My basic goal for now is to re-assess the pytest wrapper, and figure out what's going to need to change.
Comparing against other wrappers, I need, in approximate order:
- Program subclass
- Python versions ValueAdaptor
- Pip args Label
- Entry alias
- CommandBuilder alias
- OptionBuilder alias
- COMMAND global
- BASE global; things get tricky here, because as part of my paranoia about letting these programs touch the disk willy-nilly, I right now have it try to default to providing -p no:cacheprovider; on balance, it probably makes sense to ditch this and move that to the pytest config file
- So, assume that NO_CACHE and DEFAULT get deleted, and BASE gets the same niceties as the other wrappers
- The last thing in the wrapper, then, is the junit_xml function, which I now remember I never finished updating the code so that didn't need to be a function.
So, there are some fairly mechanical updates to the pytest wrapper, but before I deal with those, I need to:
- Disable caching in the pytest config
- Rework motr._api.helpers.reports.accumulator to get data from a parameter and also inject additional segments, forcing the file argument to be just a basename; this is technically less flexible, but I don't think it matters
Anyway, I'll try to get to those this weekend, and try to handle other things now.
Good night.