Yicong-Huang opened a new pull request, #4871: URL: https://github.com/apache/texera/pull/4871
### What changes were proposed in this PR? Introduce a class-level ScalaTest tag annotation and a dedicated CI job for Scala tests that need a Python runtime, and start the migration with the reconfiguration specs. Other e2e specs (`DataProcessingSpec`, `PauseSpec`, `BatchSizePropagationSpec`, `PythonWorkflowWorkerSpec`) can move in follow-up PRs. **Tag annotation** — `amber/src/test/scala/.../tags/IntegrationTest.scala`. A Scala `StaticAnnotation` marked with `@TagAnnotation` + `@Retention(RUNTIME)` + `@Target(TYPE | METHOD)`. ScalaTest's `-n` / `-l` flags treat the FQN `org.apache.texera.amber.tags.IntegrationTest` as the tag name; applying the annotation at the class level tags every test in the suite. **Test split** — `ReconfigurationSpec` (5 tests) split by Python dependency: - `ReconfigurationSpec` keeps the two pure-Scala tests (Java operator reconfiguration + source-operator guardrail). - New `ReconfigurationIntegrationSpec` carries the three tests that spawn Python UDF workers, annotated with `@IntegrationTest`. **`build.yml`** - `amber` job drops `setup-python` + `uv` install steps. Test step prepends a `sbt 'set ... testOptions += Tests.Argument(... "-l" tag)'` so jacoco excludes integration-tagged tests. - New `amber-integration` job mirrors amber's JDK + sbt + Postgres setup, adds Python install, runs only the `-n`-filtered tagged subset under jacoco. Uploads coverage with `flags: amber-integration`. Gated by a new `run_amber_integration` input. **`required-checks.yml`** — `precheck` emits `run_amber_integration`. `LABEL_STACKS` now treats `amber-integration` as its own stack: | Label | frontend | amber | amber-integration | platform | python | agent-service | |---|---|---|---|---|---|---| | `frontend` | ✓ | | | | | | | `python` | | | ✓ | | ✓ | | | `engine` | | ✓ | ✓ | | ✓ | | | `platform` | | | | ✓ | | | | `agent-service` | | | | | | ✓ | | `common` | | ✓ | ✓ | ✓ | | | | `ddl-change` | | ✓ | ✓ | ✓ | | | | `ci` | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | A bare `python` label fires only `amber-integration` + `python` — the lighter `amber` job is no longer paid for. ### Any related issues, documentation, discussions? Closes #4870. Follow-ups will move `DataProcessingSpec`, `PauseSpec`, `BatchSizePropagationSpec`, `PythonWorkflowWorkerSpec` into the new job (or split them similarly when only some tests need Python). ### How was this PR tested? YAML parses locally for `build.yml` and `required-checks.yml`. The PR's own CI run will exercise both jobs end to end; this PR carries the `engine` + `python` + `ci` labels, so all five Scala/Python stacks should fire. ### Was this PR authored or co-authored using generative AI tooling? Generated-by: Claude Code (claude-opus-4-7) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
