- Create a new release branch:
git checkout -n release_1.0.x
(replace .x with the actual version). - Install
tbump
withpip install tbump
. - Bump the version with
tbump --only-patch 1.0.x
- Save, sign and commit your changes:
git commit -s -m "Release 1.0.x"
. - Open a PR with your release branch:
git push -u origin release_1.0.x
. - Once your PR has been merged (!), pull the new main branch
git checkout main && git pull upstream main
. - Add a new release tag:
git tag -a 1.0.x -m "Release 1.0.x"
. - Push the new tag to GitHub:
git push upstream --tags
. - Create a new conda environment:
conda create -n release_grid -c conda-forge python=3.8 python-build
. - Activate the environment:
conda activate release_grid
.
- Make sure the
dist
folder under theipydatagrid
root folder is empty - delete previous release tarballs if they're present. - Build the Python release by running
python -m build
in the root directory. - Check the
dist
folder for the output from the build process. You should see the following files:- ipydatagrid-1.0.x-py3-none-any.whl
- ipydatagrid-1.0.x.tar.gz
- index.js
- index.js.map
- index.js.LICENSE.txt
- Delete all *.js, *.map and *.txt files:
rm *.js *.map *.txt
. Only the wheel and tarball files should remain. - Inspect the contents of the tarball to see everything is in check:
tar -tvf ipydatagrid-1.0.x.tar.gz
. - Install
twine
by runningpip install twine
. - Once
twine
is installed, upload the Python release to pypi:twine upload dist/*
. You will need to have upload credentials foripydatagrid
on pypi for this step to work. If the step was successful, you should see an output similar to the one below:
Uploading ipydatagrid-1.0.x-py3-none-any.whl
100%|████████████████████████████████████████████████████████████████████████████████████████████| 3.38M/3.38M [00:04<00:00, 720kB/s]
Uploading ipydatagrid-1.0.x.tar.gz
100%|███████████████████████████████████████████████████████████████████████████████████████████| 22.4M/22.4M [00:07<00:00, 3.22MB/s]
View at:
https://pypi.org/project/ipydatagrid/1.0.x/
- In the root folder of the
ipydatagrid
, runnpm publish --dry-run
. This will perform a dry run release so you can check everything looks right before an actual release. It should generate an output similar to the one below:
webpack 5.34.0 compiled with 3 warnings in 93596 ms
npm notice
npm notice 📦 [email protected]
npm notice === Tarball Contents ===
npm notice 4.5kB style/feathergrid.css
npm notice 56B style/jupyter-widget.css
npm notice 4.5kB lib/core/barrenderer.js
npm notice 11.1kB lib/cellrenderer.js
npm notice 165.1kB lib/core/datagrid.js
npm notice 19.5kB lib/datagrid.js
npm notice 1.3kB lib/extension.js
npm notice 30.4kB lib/feathergrid.js
npm notice 45.1kB lib/core/filterMenu.js
npm notice 13.7kB lib/core/graphicscontext.js
npm notice 5.2kB lib/core/gridContextMenu.js
npm notice 13.6kB lib/core/headerRenderer.js
npm notice 1.1MB dist/index.js
npm notice 930B lib/index.js
npm notice 1.4kB lib/keyhandler.js
npm notice 3.1kB lib/plugin.js
npm notice 114B lib/core/transform.js
npm notice 10.8kB lib/core/transformExecutors.js
npm notice 7.1kB lib/core/transformStateManager.js
npm notice 6.2kB lib/utils.js
npm notice 4.6kB lib/core/valueRenderer.js
npm notice 3.8kB lib/vegaexpr.js
npm notice 584B lib/version.js
npm notice 7.8kB lib/core/view.js
npm notice 16.7kB lib/core/viewbasedjsonmodel.js
npm notice 4.0kB package.json
npm notice 3.1kB lib/core/barrenderer.js.map
npm notice 7.7kB lib/cellrenderer.js.map
npm notice 128.2kB lib/core/datagrid.js.map
npm notice 15.8kB lib/datagrid.js.map
npm notice 368B lib/extension.js.map
npm notice 22.8kB lib/feathergrid.js.map
npm notice 31.4kB lib/core/filterMenu.js.map
npm notice 13.1kB lib/core/graphicscontext.js.map
npm notice 2.8kB lib/core/gridContextMenu.js.map
npm notice 10.6kB lib/core/headerRenderer.js.map
npm notice 4.3MB dist/index.js.map
npm notice 189B lib/index.js.map
npm notice 689B lib/keyhandler.js.map
npm notice 1.6kB lib/plugin.js.map
npm notice 117B lib/core/transform.js.map
npm notice 8.4kB lib/core/transformExecutors.js.map
npm notice 4.9kB lib/core/transformStateManager.js.map
npm notice 4.1kB lib/utils.js.map
npm notice 3.7kB lib/core/valueRenderer.js.map
npm notice 2.7kB lib/vegaexpr.js.map
npm notice 287B lib/version.js.map
npm notice 4.1kB lib/core/view.js.map
npm notice 9.9kB lib/core/viewbasedjsonmodel.js.map
npm notice 4.6kB README.md
npm notice 416B style/icons/arrow-down-short-dark.svg
npm notice 416B style/icons/arrow-down-short.svg
npm notice 400B style/icons/arrow-up-short-dark.svg
npm notice 400B style/icons/arrow-up-short.svg
npm notice 349B style/icons/filter-dark.svg
npm notice 349B style/icons/filter.svg
npm notice 3.8kB lib/core/barrenderer.d.ts
npm notice 4.5kB lib/cellrenderer.d.ts
npm notice 38.0kB lib/core/datagrid.d.ts
npm notice 2.6kB lib/datagrid.d.ts
npm notice 25B lib/extension.d.ts
npm notice 3.9kB lib/feathergrid.d.ts
npm notice 10.0kB lib/core/filterMenu.d.ts
npm notice 5.5kB lib/core/graphicscontext.d.ts
npm notice 3.3kB lib/core/gridContextMenu.d.ts
npm notice 2.5kB lib/core/headerRenderer.d.ts
npm notice 103B lib/index.d.ts
npm notice 584B lib/keyhandler.d.ts
npm notice 231B lib/plugin.d.ts
npm notice 1.6kB lib/core/transform.d.ts
npm notice 3.2kB lib/core/transformExecutors.d.ts
npm notice 2.7kB lib/core/transformStateManager.d.ts
npm notice 1.4kB lib/utils.d.ts
npm notice 1.4kB lib/core/valueRenderer.d.ts
npm notice 1.4kB lib/vegaexpr.d.ts
npm notice 262B lib/version.d.ts
npm notice 4.7kB lib/core/view.d.ts
npm notice 11.8kB lib/core/viewbasedjsonmodel.d.ts
npm notice 1.5kB LICENSE.txt
npm notice === Tarball Details ===
npm notice name: ipydatagrid
npm notice version: 1.0.5
npm notice package size: 1.3 MB
npm notice unpacked size: 6.2 MB
npm notice shasum: f8743bf45c0667b2f99631b73f5df465534549fc
npm notice integrity: sha512-4E1xna+4Sy33k[...]eQx2mpnwvYKOw==
npm notice total files: 79
npm notice
+ [email protected]
One specific item to watch for is the inclusion of dist/index.js
in the output generated by the console, as that's the file being used by CDNs, which voila
relies on as a fallback.
- If everything looks right, you can now publish to NPM. Login to NPM:
npm login
and enter your credentials. - Run
npm publish
in the root directory ofipydatagrid
. You will need to have upload credentials foripydatagrid
on NPM for this step to work. If the publishing step was successful, you will see an output on your console which is similar to the one from step (15).
- Deactivate the release environment:
conda deactivate
. - Delete the release environment:
conda env remove -n release_grid
.
Congratulations! You successfully published ipydatagrid
!