Testing Guide¶

Testing practices and guidelines for mockapi-server.

Quick Start¶

make test          # Tests with coverage
make test-fast     # Fast (no coverage)
make benchmark     # Performance benchmarks

Test Structure¶

tests/
├── conftest.py              # Shared fixtures
├── fixtures/                # Test data models
├── test_*.py                # Unit tests
└── benchmarks/              # Performance tests

Test Naming¶

test_<component>_<scenario>_<expected>

Examples:

def test_parser_with_valid_models_extracts_all_fields():
    """Parser should extract all fields from valid Pydantic models."""
    pass

def test_generator_with_email_field_creates_valid_email():
    """Generator should create valid email for fields named 'email'."""
    pass

Writing Tests¶

Basic Structure¶

def test_feature():
    """Test description."""
    # Arrange
    parser = SchemaParser()

    # Act
    result = parser.parse_file("test.py")

    # Assert
    assert result is not None
    assert "User" in result

Parametrized Tests¶

import pytest

@pytest.mark.parametrize("field_name,field_type", [
    ("email", str),
    ("age", int),
    ("active", bool),
])
def test_field_types(field_name, field_type):
    """Test field type extraction."""
    # Test implementation
    pass

Testing Exceptions¶

def test_parser_with_missing_file_raises_error():
    """Parser should raise FileNotFoundError."""
    parser = SchemaParser()

    with pytest.raises(FileNotFoundError):
        parser.parse_file("nonexistent.py")

Running Tests¶

Make Commands¶

make test          # Tests with coverage
make test-fast     # Fast (no coverage)
make test-unit     # Unit tests only
make test-cov      # View HTML coverage report
make benchmark     # Run benchmarks

Coverage¶

Targets¶

Component	Target
Parser	95%
Generator	90%
Store	100%
Router	95%
Overall	90%

Checking Coverage¶

pytest --cov=mock_api --cov-report=term-missing
pytest --cov=mock_api --cov-report=html
open htmlcov/index.html

Benchmarks¶

We use pytest-benchmark to track performance across 53 tests covering parser, generator, store, router, and server components.

make benchmark           # Run all benchmarks (~45s)
make benchmark-compare   # Compare with baseline (fails if >20% slower)
make benchmark-save      # Save new baseline after optimizations
make benchmark-report    # Generate histogram in .benchmarks/

Performance Targets: - Parser: <10ms per model - Generator: <20ms for 100 records - Store CRUD: <10μs per operation - API requests: <2ms

CI Integration: Benchmarks run on develop and main branches, posting results as commit comments and failing PRs with >20% performance regression.

Writing Benchmarks¶

def test_parser_performance(benchmark):
    """Benchmark: parser-simple."""
    parser = SchemaParser()
    result = benchmark(parser.parse_file, "tests/fixtures/models.py")
    assert len(result) == 10

Best Practices¶

DO¶

Write tests first (TDD for bugs)
Test one thing per test
Use descriptive names
Keep tests fast (< 100ms)
Test edge cases

DON'T¶

Test implementation details
Create test dependencies
Skip tests without reason
Use sleep() for timing

CI Integration¶

Tests run on every push and PR. CI runs:

make ci

Includes:

Format check
Linting
Type checking
Tests with coverage
Coverage threshold (90%)