@@ -20,23 +20,28 @@ To perform a release in GitHub, the following steps are needed:
2020## PyPI Python Package Release
2121
2222At the moment Python package (whl files) is created automatically,
23- upon each successful Travis CI on master branch. At the end of
24- each Travis CI build on master branch, all whl files
25- (2.7, 3.4, 3.5, 3.6, 3.7 on Linux and 2.7 on macOS) are pushed to
26- Dropbox and are available in:
27-
28- https://www.dropbox.com/sh/dg0npidir5v1xki/AACor-91kbJh1ScqAdYpxdEca?dl=0
23+ upon each successful GitHub Actions workflow on master branch. At the end of
24+ each GitHub Actions Workflow build on master branch, all whl files
25+ (3.5, 3.6, 3.7 on Linux, macOS, and Windows) are pushed to part of the
26+ artifact that is available to download. To download the artifact,
27+ go into the tab of [ Actions] ( https://github.com/tensorflow/io/actions ) ,
28+ click and get into the build of the commit ready for release, and download
29+ the artifact named ` tensorflow-io-release ` only. Once the artifact is
30+ uncompressed locally, all whl files on Linux, macOS, and Windows will be
31+ available.
2932
3033To perform a release in PyPI, first make sure the binary whl files
31- are the correct one from corresponding Travis CI build number .
32- This could be verified by checking the Travis CI history where at
34+ are the correct one from corresponding GitHub Actions build commit .
35+ This could be verified by checking the GitHub Actions log history where at
3336the end of the log, the sha256 of all whl files are calculated and displayed.
34- The sha256 of each file displayed on Travis CI log should match the sha256
35- of the files downloaded from Dropbox .
37+ The sha256 of each file displayed on GitHub Actions log should match the sha256
38+ of the files extracted from artifact .
3639
37- Once sha256 are verified against every whl files on Dropbox, perform
38- a sanity check, then upload all of the whl files
39- (2.7, 3.4, 3.5, 3.6, 3.7 on Linux and 2.7 on macOS) to PyPI.org:
40+ Once sha256 are verified against every whl files in artifact, perform
41+ a sanity check locally to make sure at least the getting started example
42+ on [ README.md] ( https://github.com/tensorflow/io/blob/master/README.md )
43+ is able to run correctly. After that upload all of the whl files
44+ (3.5, 3.6, 3.7 on Linux, macOS, and Windows) to PyPI.org:
4045
4146```
4247twine upload *.whl
0 commit comments