Metadata-Version: 2.4
Name: para-pytest
Version: 0.1.3
Summary: Run pytest tests in parallel chunks for faster test execution
Author: ALHelton
License: MIT
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: pytest>=7.0.0
Requires-Dist: pytest-json-report>=1.5.0
Dynamic: license-file

# Parallel Pytest (para-pytest)

Run pytest tests in parallel chunks for significantly faster test execution.


## Installation

```bash
pip install para-pytest
```


## Quick Start

```bash
# Run with default settings (4 chunks)
para-pytest

# Specify number of chunks
para-pytest --chunks 8

# Specify test path
para-pytest --path tests/
```


## CLI Options

```bash
para-pytest [OPTIONS]

Options:
  --chunks N      Number of parallel chunks (default: 4)
  --path PATH     Path to tests (default: current directory)
  --debug         Show detailed chunking and pattern matching info
```

Useful for debugging:
```bash
# See which tests match serial patterns and how tests are chunked
para-pytest --path tests/ --debug
```


## Use Cases

### Local Development

Speed up your test suite during development:

```bash
# Before (sequential execution)
pytest tests/

# After (parallel execution)
para-pytest --chunks 4
```

### Continuous Integration

Add to your GitHub Actions workflow:
```yaml
name: Tests

on: [push, pull_request]

jobs: 
	test:
    runs-on: ubuntu-latest

    steps:
      - uses: actions/checkout@v4

      - uses: actions/setup-python@v5
        with:
          python-version: '3.12.3'

      - name: Install dependencies
        run: |
          pip install -r requirements.txt
          pip install para-pytest

      - name: Run tests
        run: para-pytest --chunks 4 --path tests/
```

Or use the Github Action:
```yaml
- uses: ALHelton/parallel-pytest@v1
  with:
    chunks: 8
    path: tests/
```

### Python API

```python
from para_pytest import ParaPytestRunner

# Serial patterns are automatically loaded from pyproject.toml
runner = ParaPytestRunner(chunks=4, pytest_args=['tests/'])
exit_code = runner.run()

# Or explicitly pass serial patterns (overrides pyproject.toml)
runner = ParaPytestRunner(
    chunks=4,
    pytest_args=['tests/'],
    serial_patterns=[
        'tests/unit/test_database.py',  # Simple file path
        '**/test_integration_*'          # Wildcard pattern
    ]
)
exit_code = runner.run()
```


## Configuration (Optional)

### Serial Test Patterns

Some tests can't run in parallel (e.g., database tests, integration tests with shared state). Configure these to run serially by adding to your existing `pyproject.toml`:

```toml
[tool.para-pytest]
serial_patterns = [
    # Simple file paths (automatically matches all tests in the file)
    "app/tests/unit/test_database.py",
    "app/tests/unit/routes/test_organizations.py",
    
    # Wildcard patterns
    "**/test_database_*",
    "**/test_migration_*",
    "tests/e2e/**"
]
```

**Pattern Matching:**

Para-pytest automatically handles test node IDs (like `test_file.py::test_name`). You can use:

1. **Simple file paths** - No wildcards needed! 
   - `"app/tests/unit/test_database.py"` matches all tests in that file
   - Automatically expands to match `test_database.py::test_function`

2. **Wildcard patterns** - For matching multiple files
   - `"**/test_database_*"` - Any test file starting with `test_database_`
   - `"tests/e2e/**"` - All tests in `tests/e2e/` directory and subdirectories
   - `"**/test_*_integration.py"` - Any test file ending with `_integration.py`

That's it! No external dependencies required - the configuration is parsed using built-in Python modules.


## Performance

Real-world example with 2000+ tests:

| Method | Time | Speedup |
|--------|------|---------|
| `pytest tests/` | 50s | 1x |
| `para-pytest --chunks 4` | 15s | 3.3x |
| `para-pytest --chunks 8` | 10s | 5x |

*Performance varies based on test suite characteristics*


## How It Works

1. Loads serial patterns from `pyproject.toml` (if configured)
2. Collects all tests from pytest
3. Separates tests into parallel and serial groups based on patterns
4. Splits parallel tests into equal chunks
5. Runs parallel chunks concurrently using asyncio
6. Runs serial tests sequentially (if any)
7. Aggregates and displays results


## Troubleshooting

### Tests Fail in Parallel But Pass Normally?

Some tests can't run in parallel due to shared state (databases, files, etc.). Configure them to run serially:

```toml
[tool.para-pytest]
serial_patterns = [
    "**/test_database_*",
    "**/test_integration_*",
]
```

Run with `--debug` to see which tests are being run serially:
```bash
para-pytest --path tests/ --debug
```

### Common Patterns for Serial Tests

- Database tests: 
  - `"tests/unit/test_database.py"` (specific file)
  - `"**/test_database_*"` (any file starting with `test_database_`)
- Integration tests: 
  - `"**/test_integration_*"`, `"tests/integration/**"`
- E2E tests: 
  - `"tests/e2e/**"`, `"tests/functional/**"`
- Any test modifying shared resources

**Note:** For specific files, you don't need wildcards - just use the file path!


## Requirements

- Python 3.8+
- pytest 7.0+


## License 

MIT
