v1.16.0

✨ New Things

@felixfontein added a new codecov-token input, letting the end-users pass their Codecov token when using this action via #82.

🛠️ Internal Dependencies

• codecov/codecov-action was bumped from 4.3.1 to 4.5.0 in #84
• A number of pre-commit.ci checks have been auto-updated in #83

🪞 Full Diff: v1.15.0...v1.16.0

v1.15.0

✨ What's New

It is now possible to specify which sanity tests to run / skip / enable. Courtesy of @felixfontein: #73.

📝 What's Documented

@felixfontein added a note to README advising disabling coverage for roles-only collections via #78

💅 Internal Refactoring

@felixfontein improved the code layering by moving composing of all ansible-test arguments into Python code via #75.

🛠️ Internal Dependencies

• actions/checkout was bumped from 3 to 4 in #69
• actions/setup-python was bumped from 4 to 5 in #71
• A number of pre-commit.ci checks have been auto-updated in #70, #72 and #76

🪞 Full Diff: v1.14.1...v1.15.0

v1.14.1

🐛 What's Fixed

Stopped restricting PyYAML version in #65, which unbreaks sanity checks under Python 3.11, thanks to @felixfontein.

Diff Changelog: v1.14.0...v1.14.1

v1.14.0

💅 New stuff

@felixfontein added origin Python auto-detection support for ansible-core v2.15 in #60. He also bumped the default Python runtime fallback to v3.11.

🛠️ Internal thingies

• @felixfontein also introduced a Dependabot config in #61, targeted at updating GitHub Action versions in the smoke test workflow.
• Some pre-commit based linters have been upgraded and @webknjaz integrated gitlint; he also cut this release.

🐛 Bugfixes

• @webknjaz fixed displaying the markdown coverage table in GHA job summary in #62 with some guidance from @mattclay via #55.

🪞 Full Diff: v1.13.0...v1.14.0

v1.13.0

For this release, @felixfontein made a series of pull requests with the changes listed below.

What's Changed

New action inputs were added (default is bold):

• coverage with the possible values of always, never, and auto to allow controlling code coverage (#50)
• integration-continue-on-error, boolean true, let's ansible-test continue with other integration tests after encountering an error (#51)
• integration-diff, boolean true, for showing the diff in integration tests (#51)
• integration-retry-on-error, boolean true, instructs ansible-test to reattempt a failed integration test target instead of giving up immediately (#51)

Internal things

• Switched to using Python 3.8 instead of 3.6 as environment Python version (#54)

Full Diff: v1.12.0...v1.13.0

v1.12.0

Something New

• Implemented an optional ansible-test's change detection mode for Pull Requests by @felixfontein in #49

Internal Updates

• Use yaml.safe_load() instead of yaml.load() by @felixfontein in #47

Full Diff: v1.11.2...v1.12.0

v1.11.2

What's Changed

Only the internals of the action have been updated to produce job step outputs through environment files¹²³. No UX changes were made. Additionally, a third party action that we use has been fixed to do the same⁴.

Full Diff: v1.11.1...v1.11.2

1. https://hynek.me/til/set-output-deprecation-github-actions/ ↩

2. https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/#examples ↩

3. https://docs.github.com/en/actions/using-workflows/workflow-commands-for-github-actions#setting-an-output-parameter ↩

4. #41 ↩

v1.11.1

What's Fixed

The coverage summary and reporting logic attempted to interpret uncombined coverage reports. This is unnecessary and causes problems. It has been updated to only look at the combined coverage.

Full Diff: v1.11.0...v1.11.1

v1.11.0

What's New

This release integrates displaying job coverage summary reports on the GitHub Actions Workflow Summary page.

Full Diff: v1.10.0...v1.11.0

v1.10.0

What's New

This version of action started posted per-job test summaries inferred from JUnit reports on the GitHub Actions Workflow Summary pages.

Here's an example of that: https://github.com/ansible-community/ansible-test-gh-action/actions/runs/3139054007#summary-8582766319.

Full Diff: v1.9.0...v1.10.0