refactor: fix add-path / GITHUB_PATH commands

[ChristopherHX]

The existing code applies prepend path for all actions, while it shouldn't do it for docker actions.
Additionally GITHUB_PATH shouldn't update the env directly, instead add to extrapath

Seems like the old ::add-path:: never worked correctly, because the first add-path always wins e.g.

• add node12 as node
• use node ver 12
• add node16
• still in version 12, node16 added to a later path component after node12

Duplicated path are now eliminated.

Resolves #1384

Still to complex for get it ready.

[github-actions]

🦙 MegaLinter status: ✅ SUCCESS

Descriptor	Linter	Files	Fixed	Errors	Elapsed time
✅ EDITORCONFIG	editorconfig-checker	12		0	0.04s
✅ REPOSITORY	gitleaks	yes		no	2.44s
✅ REPOSITORY	git_diff	yes		no	0.0s
✅ REPOSITORY	secretlint	yes		no	0.94s

See errors details in artifact MegaLinter reports on CI Job page
Set VALIDATE_ALL_CODEBASE: true in mega-linter.yml to validate all sources, not only the diff

MegaLinter is graciously provided by

[codecov]

Codecov Report

Merging #1472 (03b61e9) into master (4f8da0a) will increase coverage by 3.76%.
The diff coverage is 67.67%.

@@            Coverage Diff             @@
##           master    #1472      +/-   ##
==========================================
+ Coverage   57.50%   61.27%   +3.76%     
==========================================
  Files          32       45      +13     
  Lines        4594     6990    +2396     
==========================================
+ Hits         2642     4283    +1641     
- Misses       1729     2403     +674     
- Partials      223      304      +81     

Impacted Files	Coverage Δ
pkg/common/file.go	0.00% <0.00%> (ø)
pkg/container/docker_logger.go	52.08% <ø> (ø)
pkg/container/host_environment.go	0.00% <0.00%> (ø)
pkg/container/parse_env_file.go	0.00% <0.00%> (ø)
pkg/container/util.go	0.00% <0.00%> (ø)
pkg/model/action.go	0.00% <0.00%> (ø)
pkg/model/step_result.go	0.00% <ø> (ø)
pkg/container/docker_run.go	13.58% <11.48%> (+8.04%)	⬆️
...ontainer/linux_container_environment_extensions.go	24.32% <24.32%> (ø)
pkg/model/workflow.go	46.43% <25.25%> (-4.48%)	⬇️
... and 38 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[ChristopherHX]

I finally got the tests all green.

I modified tests, read carefully.

[catthehacker]

Looks good

[mergify]

@ChristopherHX this pull request has failed checks 🛠

[mergify]

@ChristopherHX this pull request has failed checks 🛠

[mergify]

@ChristopherHX this pull request has failed checks 🛠

[mergify]

@ChristopherHX this pull request has failed checks 🛠

[mergify]

@ChristopherHX this pull request has failed checks 🛠

[mergify]

@ChristopherHX this pull request has failed checks 🛠

[ChristopherHX]

...I am doing something wrong..., ...now the increased 15min timeout isn't sufficient after merging my other PRs...

[mergify]

@ChristopherHX this pull request has failed checks 🛠

[mergify]

@ChristopherHX this pull request has failed checks 🛠

[ChristopherHX]

I will need to debug this test execution regression, github actions also fails with 30min timeout.

Converting to draft

Thank you @catthehacker, @KnisterPeter for reviewing my PR wave.

[KnisterPeter]

I will need to debug this test execution regression, github actions also fails with 30min timeout.

I think we have this issue in master already. So it might be something we merged recently.

[ChristopherHX]

Yes you are right, it is on master as a flaky test.

It's freezes in TestMaskValues while waiting for the output. Removing the test make the ci pass.