You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
one could use that as a backup if the parsing directly fails... (through then there is probably also nothing on pypi).
You use the orginal metadata parser from pkg and importlib_metadata, right?
One could also request the metadata for each python dependency that way to fill from that the metadata of the dependencies required for codemeta, or parse the metadata from the dependencies (which would only work if installed or?).
The text was updated successfully, but these errors were encountered:
one could use that as a backup if the parsing directly fails... (through then there is probably also nothing on pypi).
That sounds like a good idea yes. It should be quite doable. We also have
Github and Gitlab API support after all, adding PyPi to the mix as a potential
source makes sense. Though indeed I'd rather have it as a fallback and not rely on it primarily, as
I think it's best if we can get as much from the git clone itself.
I'm definitely open to contributions on this.
You use the orginal metadata parser from pkg and importlib_metadata, right?
Indeed, that saves having to parse pyproject.toml or setup.py or setup.cfg directly and seems to most elegant way.
One could also request the metadata for each python dependency that way to
fill from that the metadata of the dependencies required for codemeta, or
parse the metadata from the dependencies (which would only work if installed
or?).
The actual dependencies seem not be listed in the pypi JSON. But from the source repo, it would work yeah.
Just an idea:
Pypi has an API for the basic metadata, for example see: https://pypi.org/pypi/codemetapy/json
one could use that as a backup if the parsing directly fails... (through then there is probably also nothing on pypi).
You use the orginal metadata parser from pkg and importlib_metadata, right?
One could also request the metadata for each python dependency that way to fill from that the metadata of the dependencies required for codemeta, or parse the metadata from the dependencies (which would only work if installed or?).
The text was updated successfully, but these errors were encountered: