Ensure pypi-normalized names when fetching from npe2api #1335
Conversation
aganders3
added
the
bug-fix
data-workflows/plugin/processor.py
Outdated
| pypi_plugin_version = pypi_latest_plugins.get(name) | ||
| if pypi_plugin_version == version: | ||
| continue | ||
|
|
||
| if pypi_plugin_version is None and is_plugin_active(name, version): |
There was a problem hiding this comment.
As we are fetching data from npe2api and not pypi, could we simply remove the check for is_plugin_active?
There was a problem hiding this comment.
Yes, I think we can remove it. Do you mean in addition to the other changes here, or instead?
I still think it makes sense to do the PyPI normalization to ensure a single version of each plugin. In prod I think this also gives continuity with past versions (though I'm not sure this really matters).
There was a problem hiding this comment.
I removed this check in bcbdd73. The check is then unused so I removed the code and tests with it. I'm happy to do whatever is cleanest here 😅.
There was a problem hiding this comment.
If we removed the is_plugin_active check, it should also take care of removing the duplicate entry. So one of those checks should be enough.
There was a problem hiding this comment.
I agree the normalization of the names would be nice to have. But if we go down that route, we should also normalize the names in results returned by npe2api, as all of them were not normalized. ex: AlveolEye, acquifer-napari.
|
I'm not sure why the Push Remote Dev workflow is failing. It looks okay to me in Terraform. |
Description
This coerces npe2api names to be PyPI-normalized. Thanks to @manasaV3 for debugging the issue and @neuromusic for the link to the normalization docs.
We may want to also fix this in npe2api, but I'm not sure the consequences of changing that right now.
Closes #1334