Update tasks documentation

This commit is contained in:
2026-06-21 12:28:35 +02:00
parent ad1aace0f5
commit 107158eb04

View File

@@ -1,30 +1,3 @@
## Global / Infrastructure Issues
### Task INFRA-1: Fix Server Startup in Robot Framework Suite Setup
- **Failing Step**: `Suite Setup → Start Aniworld Server → Wait For Server`
- **Failure Message**: `Server did not become healthy within 30 seconds`
- **Files to Check**:
- `tests/robot/resources/common.resource` (keywords: `Start Aniworld Server`, `Wait For Server`)
- `tests/robot/run.sh`
- `tests/robot/__init__.robot`
- **Reference Docs**:
- `Docs/TESTING.md`
- `Docs/API.md` §1 (Base URL and Versioning)
- **Expected Behavior**:
- The suite setup should start the FastAPI uvicorn server on `127.0.0.1:8765`.
- The `Wait For Server` keyword should poll `/health` and receive HTTP 200 within 30 seconds.
- The server process should remain running for the duration of the test suite.
- **Actual Behavior**:
- The server process likely fails to start because the active Python environment (`base` conda env, Python 3.13.5) does not have the project's dependencies installed.
- `run.sh` invokes plain `python`, which resolves to the base environment instead of the project's `AniWorld` conda environment.
- **Suggested Fix**:
- Update `tests/robot/resources/common.resource` to start the server with `conda run -n AniWorld python -m uvicorn ...` instead of plain `python -m uvicorn`.
- Alternatively, ensure the `AniWorld` conda environment is activated before running `tests/robot/run.sh`.
---
### Task INFRA-2: Fix Suite Teardown — Close Browser Keyword Argument Mismatch
- **Failing Step**: `Suite Teardown → Close Browser`
@@ -2072,4 +2045,4 @@
1. **INFRA-1** — Fix the server startup environment issue. This is the blocker for all 129 tests.
2. **INFRA-2** — Fix the suite teardown `Close Browser` keyword argument mismatch.
3. Once the server starts correctly, re-run the full Robot Framework suite to identify any individual test logic failures.
4. Address any remaining per-test failures based on the new run results.
4. Address any remaining per-test failures based on the new run results.