ROX-28901: Derive EARLIER_CHART_VERSION in scanner-v4-install tests

[mclasmeier]

Description

Don't hard-code an older version, instead derive a previous version from the existing git tags.

User-facing documentation

Just a change in tests, nothing user-facing.

Testing and quality

• the change is production ready: the change is GA, or otherwise the functionality is gated by a feature flag
• CI results are inspected

Automated testing

• modified existing tests

How I validated my change

• Tested the function locally.
• CI results.

From the CI log:

INFO: Thu Jun 12 11:05:52 UTC 2025: [setup-file] Using MAIN_IMAGE_TAG=4.9.x-47-gf4e58726af
INFO: Thu Jun 12 11:05:52 UTC 2025: [setup-file] Cloning released Helm charts into /tmp/helm-charts.hEJf2a...
INFO: Thu Jun 12 11:05:56 UTC 2025: [setup-file] Using EARLIER_CHART_VERSION=4.7.0

which is correct (given that Helm charts fro 4.8.x have not yet been released).

[sourcery-ai]

Hey @mclasmeier - I've reviewed your changes - here's some feedback:

• Adjust the grep and sed regex to support multi-digit patch versions (e.g. use '[0-9]+' for the z component) so tags like '1.10.12' aren’t dropped.
• Use a version-aware sort (e.g. sort -V) instead of numeric sort on the x.y string to correctly order multi-digit versions.
• Add explicit error handling or a clear fallback when no previous x.y.0 tag exists to avoid silent failures in the test.

Here's what I looked at during the review

• 🟡 General issues: 3 issues found
• 🟢 Security: all looks good
• 🟢 Testing: all looks good
• 🟢 Complexity: all looks good
• 🟢 Documentation: all looks good

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}

[rhacs-bot]

Images are ready for the commit at f4e5872.

To use with deploy scripts, first export MAIN_IMAGE_TAG=4.9.x-47-gf4e58726af.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 48.79%. Comparing base (3864927) to head (f4e5872).
Report is 2 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master   #15594      +/-   ##
==========================================
- Coverage   48.80%   48.79%   -0.01%     
==========================================
  Files        2590     2590              
  Lines      190492   190492              
==========================================
- Hits        92963    92960       -3     
- Misses      90231    90233       +2     
- Partials     7298     7299       +1     

Flag	Coverage Δ
go-unit-tests	48.79% <ø> (-0.01%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[porridge]

Compared to the way we figure out the previous version in the operator Makefile, this purely git-based method has the advantage that it's mostly offline, but it also has the disadvantage that it will generate an unusable version in the time frame between the x.y.0 tag is pushed to the git repo, to the moment the installation artifacts (images and perhaps also helm charts) are built and pushed.

Given that we need network connectivity anyway for this test to work, I think it would be preferable to use the image registry to check for previous version existence, than have a way which is offline but fails every now and then...

[mclasmeier]

@porridge, see my last commit, I think I was able to increase reliability while also retaining the general approach.

Instead of using git tags and then hoping that this will allow the test suite to resolve the computed tag to resolve to a Helm chart version, I figured it would make much more sense to begin with Helm chart versions, because that is what is needed here.

So, the code now inspects the helm-charts repo, looks for the versions in that repo and then figures out what the 'previous' (x.y.0) version should be -- similar to what I previously did with git tags.

[mclasmeier]

/retest

[vladbologa]

@sourcery-ai review

[sourcery-ai]

Hey @mclasmeier - I've reviewed your changes - here's some feedback:

• Consider extracting the new version‐resolution helpers (resolve_previous_x_y_0_version, list_sub_directories, etc.) into a shared test‐utility file to keep the main .bats script concise.
• To speed up CI and reduce overhead, consider caching the cloned helm‐charts repo between runs or using a more focused shallow clone (e.g. limit tags) instead of pulling the full repo each time.

Here's what I looked at during the review

• 🟢 General issues: all looks good
• 🟢 Security: all looks good
• 🟢 Testing: all looks good
• 🟢 Complexity: all looks good
• 🟢 Documentation: all looks good

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}

[porridge]

The new functions are rather complicated, so it would be great if there were some unit tests covering corner cases like when there exist previous patch versions or no patch version for the given minor version.

[vladbologa]

Looks good!

[openshift-ci]

@mclasmeier: The following test failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
ci/prow/gke-qa-e2e-tests	f4e5872	link	false	/test gke-qa-e2e-tests

Full PR test history. Your PR dashboard.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here.