Fix run skipped/ignored tests with --run-ignored

[OmChillure]

Fixes #548

Summary

This PR adds run-ignored support through a dedicated CLI option and fixes skip-handling behavior while preserving existing filter semantics.

You can now run:

karva test --run-ignored only
karva test --run-ignored all

What Changed

Added a new test CLI option: --run-ignored WHICH

Supported modes:
default
only
all

Wired run-ignored mode through CLI, runner orchestration, worker settings, and semantic test execution.

Updated skip handling so:

• default: skipped tests stay skipped
• only: only skipped tests are executed
• all: skipped and non-skipped tests are both executed
• Added skip-marker presence handling in tag evaluation to support only mode correctly.

Added Integration Coverage For

1. run-ignored only mode
2. run-ignored all mode
3. no-skipped-tests behavior under only mode
4. conditional skip behavior with skipif(False, ...)
5. module wiring for the new run_ignored integration test suite

Behavior

1. Without --run-ignored: skipped tests remain skipped.
2. With --run-ignored only: only skipped tests execute.
3. With --run-ignored all: skipped and non-skipped tests execute together.

Validation

1. Ran targeted integration tests for run_ignored scenarios.
2. Verified expected outcomes for both only and all modes.
3. Confirmed compile and test consistency for updated semantic runner flow.

Screenshots

Before

--run-ignored only

--run-ignored all

Notes
This PR introduces run-ignored behavior via a dedicated CLI option, not via filter expressions.

[MatthewMckee4]

Thanks!

[MatthewMckee4]

Why do we not check should_skip here?

[OmChillure]

I used has_skip_tag rather than should_skip so that tests with a skip decorator are included even when the skip condition evaluates to False (e.g. @karva.tags.skip(False,=reason=...)). should_skip would exclude those tests because the condition is inactive, but --run-ignored only targets any test decorated with skip, regardless of the condition.

[MatthewMckee4]

I'm not sure this should be the case. If a test is not skipped normally (if it has a conditional skip tag) then it runs when we run skipped tests, that doesn't seem right.

Another way I think of this is that if you run tests normally, then you run tests with --run-ignored only, all tests should have run once and only once.

[OmChillure]

ahh yeah , I reconsidered the thinking using has_skip_tag would cause tests with a False condition to run in both normal and --run-ignored passes, breaking the once-and-only-once . will update to 'should_skip'

[OmChillure]

fixed now

PS : I have used .0 extracts the bool from (bool, Option<String>); we don't need the reason here since we are overriding the skip. thats fine right?

[codspeed-hq]

Merging this PR will not alter performance

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

✅ 1 untouched benchmark

---

_{Comparing OmChillure:fix-run-skipped-test-with-run-ignored (b43db8c) with main (094c588)}