-
Notifications
You must be signed in to change notification settings - Fork 54
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Use non-lazy platform specific artifacts for unicode-cldr #439
Conversation
There is an argument to be made for testing on Windows 32-bit to ensure this artifact is present and working. I'll temporarily enable this environment as part of this PR but I don't believe it's necessary to test against this environment for other PRs. The exception being additional changes to the unicode-cldr artifact generation code. |
Codecov Report
❗ Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the Github App Integration for your organization. Read more. @@ Coverage Diff @@
## master #439 +/- ##
=======================================
Coverage 95.57% 95.57%
=======================================
Files 38 38
Lines 1762 1762
=======================================
Hits 1684 1684
Misses 78 78
|
We could configure it so that the Windows 32-bit CI job runs on PRs that touch said code, but otherwise is automatically skipped. |
An additional future improvement that can be made would be to create a new artifact to replace our use of unicode-cldr such that we can use a tiny XML file as the artifact instead of the 50MB artifact we currently fetch. |
The path filtering with GHA workflows happens at a workflow level so we'd either need a separate workflow, generate our matrix, or introduce conditions into our steps. If I end up making the reduced sized artifact I'll probably use a separate data pkg and remove the Artifacts.toml here so I'll punt on the CI changes for now. |
This reverts commit dee8df5.
Validated that Windows 32-bit works in dee8df5. Reverted the commit and will merge after CI completes |
As brought up in #438 we were previously using lazy artifacts for the multiple unicode-cldr artifacts which are only needed on Windows for translating the current OS time zone name into the IANA time zone name. This approach was problematic because PackageCompiler 1.0 would always download all of the lazy artifacts and PackageCompiler 2.0 would not download the lazy artifacts (unless you use
include_lazy_artifacts=true
) resulting in the used unicode-cldr artifact being downloaded at runtime. Additionally, we only require the latest unicode-cldr artifact and the only reason we included multiple of these artifacts was due to us failing to clean them up in thedev/generate_artifacts.jl
script which generates this file.This PR changes the
dev/generate_artifacts.jl
to update theArtifacts.toml
to only include the latest unicode-cldr. We now specify the platform for this architecture so that the artifact is only used on Windows which also allows us to specify it as non-lazy. Finally, there was a newer unicode-cldr release so I've also updated to use release-43-1 here.Supercedes #438