testsAndMisc/.github/copilot-instructions.md
Krzysztof kuhy Rudnicki 21b7b8107b fix: resolve all pre-commit hook failures after file splits
- Remove all # type: ignore and # noqa comments (banned by no-noqa hook)
- Add mypy --disable-error-code flags to pre-commit config for error
  codes previously suppressed by inline comments
- Fix broken imports after ruff auto-removed re-exports:
  steam_backlog_enforcer, stockfish_analysis, word_frequency, lichess_bot
- Re-add re-exports with __all__ in translator.py, screen_lock.py
- Split _process_epc_fc.py (524 lines) into _process_epc_fc.py + _process_fc.py
- Fix test failures: keyboard_coop, stockfish_analysis, tag_divider
- Add per-file-ignores for PLC0415 (deferred imports) in 7 files
- Mark shebang scripts as executable
- Add __init__.py for generate_images and repo_explorer packages
- Fix codespell, eslint, ruff-format, prettier issues
- Update copilot-instructions.md with --no-verify ban
2026-03-18 22:20:05 +01:00

127 lines
4.5 KiB
Markdown

# Copilot Instructions for testsAndMisc
## Project Overview
A mixed-language monorepo containing Python packages, C programs, TypeScript apps, and misc scripts. The primary actively-developed component is `python_pkg/lichess_bot/` - a Lichess chess bot.
## Architecture
### Python Packages (`python_pkg/`)
- **lichess_bot/** - Main project: Lichess bot with 100% test coverage requirement
- `main.py` - Bot entry point, game handling, event loop
- `lichess_api.py` - Lichess API client (NDJSON streaming)
- `engine.py` - Wraps C engine at `C/lichess_random_engine/`
- `tests/` - Comprehensive pytest tests using `MagicMock`, `PropertyMock`, `patch`
- **stockfish_analysis/** - Post-game analysis using Stockfish engine
- Called by lichess_bot after games to analyze move quality
- Parses PGN from game logs, outputs per-move evaluations
- **keyboard_coop/** - 2-player cooperative word game (pygame)
- Players form words using adjacent QWERTY keys
- **mock_server/** - mitmproxy-based connection failure simulator
- **screen_locker/** - Tkinter screen lock with systemd integration
- Other standalone scripts: download_cats, extract_links, random_jpg, etc.
### C Engine (`C/lichess_random_engine/`)
Chess move selection engine called by Python via subprocess:
```bash
# Build the engine
cd C/lichess_random_engine && make
# Binary usage (called by python_pkg/lichess_bot/engine.py)
./random_engine --fen "<FEN>" move1 move2 move3... # Returns chosen UCI move
./random_engine --fen "<FEN>" --explain move1... # Returns JSON with analysis
```
**Key files:**
- `main.c` - CLI interface, argument parsing
- `movegen.c/h` - Legal move generation
- `search.c/h` - Alpha-beta search with scoring
- `Makefile` - Build with `make`, clean with `make clean`
### Go Projects (`robotgo_demo/`)
- **robotgo_demo/** - Desktop automation using [go-vgo/robotgo](https://github.com/go-vgo/robotgo)
- `main.go` - Demo: mouse, keyboard, screen, clipboard, window control
- Requires X11 (Arch Linux), built with `go build -o robotgo_demo .`
- See `robotgo_demo/.github/copilot-instructions.md` for full API reference
### Cross-Language Integration
- Python `engine.py` → calls C binary via `subprocess.Popen`
- Python `stockfish_analysis/` → post-game analysis subprocess
## Development Workflow
### Testing (Critical - 100% coverage enforced)
```bash
# Run lichess_bot tests with coverage
python -m pytest python_pkg/lichess_bot/tests/ --cov=python_pkg.lichess_bot --cov-branch --cov-fail-under=100
# Quick test run
python -m pytest python_pkg/lichess_bot/tests/ -x -v
```
### Pre-commit Hooks (Always run before commits)
```bash
pre-commit run --all-files # Full check
pre-commit run --files <file1> <file2> # Specific files
```
**Hook order**: ruff (lint+fix) → ruff-format → mypy → pylint → bandit
**CRITICAL: NEVER use `--no-verify`** on `git commit` or `git push`, even when pre-commit failures are pre-existing and unrelated to your changes. Fix the failures or ask the user how to proceed — never bypass hooks.
### Linting Tools Configured in `pyproject.toml`
- **ruff**: `select = ["ALL"]` - all rules enabled, Google docstrings
- **mypy**: `strict = true` with full type checking
- **pylint**: all checks enabled
- **coverage**: `fail_under = 100`, branch coverage required
## Code Conventions
### Python Style
- Use `from __future__ import annotations` for forward references
- Google docstring convention
- Absolute imports only (`ban-relative-imports = "all"`)
- Type hints required on all functions
- Private functions prefixed with `_` (e.g., `_process_game_event`)
### Test Patterns (`python_pkg/lichess_bot/tests/`)
```python
# Type aliases for test dicts (keeps mypy happy)
Event = dict[str, Any]
GameThreads = dict[str, threading.Thread]
# Mock external calls, never hit real APIs
with patch("python_pkg.lichess_bot.main.threading.Thread") as mock:
...
# Use PropertyMock for property exceptions
type(mock_obj).property_name = PropertyMock(side_effect=TypeError())
```
### Branch Coverage Tips
- Use explicit `while True` + `try/except StopIteration` instead of `for` loops when iterator exhaustion needs coverage
- Mock threads/subprocesses to avoid slow tests
## Key Files
- `pyproject.toml` - All tool configs (ruff, mypy, pylint, pytest, coverage)
- `.pre-commit-config.yaml` - Pre-commit hook definitions
- `requirements.txt` - Runtime dependencies
- `.github/workflows/python-tests.yml` - CI pipeline
## Per-File Ignores (in `pyproject.toml`)
Test files allow: `S101` (assert), `PLR2004` (magic values), `S310`, `S607`, `PLC0415`