Closed mahendrapaipuri closed 1 year ago
Thanks @mahendrapaipuri!
Wondering if we could use the python_packages
field of the releaser to specify all the Python packages in the monorepo? Like the jupyverse
project does here: https://github.com/jupyter-server/jupyverse/blob/b3c3589ef71d9b8088fb86cea3c92746ea05d49b/pyproject.toml#L144-L164
We could check if it is working by looking at the uploaded artifacts by the check release workflow on CI.
This is what exactly I am looking at. As we do not have a pyproject.toml
on the root of the repo, we will have to add a .jupyter-releaser.toml
for config. JupyterLab is using sort of a dummy package name for the root package.json
and using build-before-npm
hook to build the packages. We should be able to do the same right?
We need to add script to package.json
to bump versions of all individual packages though.
We could also have a pyproject.toml
at the root level of the repo to hold the configuration I guess.
@jtpio I tried to integrate the check-release
workflow but with some strong assumptions.
package.json
. We should make sure that the version in the top level directory should be in sync will all packages.If we can pass --python-packages
arg at runtime, I think we can control the individual package release. But I guess that is not possible at the moment? Let me know, what do you think!
I have also reverted the versions of all packages so that we can bump them to 1.0.0
with jupyter-releaser. Replaced hypens with underscores in the package folders to get valid wheels.
Any updates on this @jtpio? I would like to use the new version of jupyterlab-topbar for JupyterLab 4.0 🙂
I think we should be good to go.
Maybe we can grab the latest version of the workflows from the releaser repo to get the proper configuration for using the PyPI trusted publishers: https://github.com/jupyter-server/jupyter_releaser/blob/main/example-workflows/publish-release.yml
Also it looks like we can version all the packages together for now, although in the long run it would be nice to allow versioning packages independently.
Also it looks like we can version all the packages together for now, although in the long run it would be nice to allow versioning packages independently.
@jtpio Agree! Lets do a release for Jupyterlab 4.0 now and we will see how we can version packages independently!
@jtpio Just checking in to see if there is anything missing in this PR to get it merged and do a release?
@mahendrapaipuri yes it should be good to go!
Merging now, will try to get a release out today. Thanks again!
Merging now, will try to get a release out today. Thanks again!
Just tried the prep release workflow specifying 1.0.0
as the version but it bumped to 0.6.1
instead:
Looks like there might be an issue when reading the new version after the bump.
Added jupyter-releaser workflows and a
CHANGELOG.md
file generated bygithub-activity
.