karpik-testing

name	karpik-testing
description	KarpikEngine testing standards. Use when Codex adds or reviews unit tests, integration tests, edge-case tests, property-based tests, BenchmarkDotNet benchmarks, allocation checks, or module verification strategy.

Karpik Testing

Required Coverage

For each module or meaningful function, consider:

• Unit tests for isolated behavior.
• Integration tests for interactions between systems.
• Edge-case tests for boundaries and invalid inputs.
• Property-based tests for invariants over generated data.
• Benchmarks or allocation checks for hot paths.

Edge Cases

Cover:

• normal values;
• zero and one;
• negative values;
• int.MaxValue, int.MinValue, float.MaxValue when relevant;
• null and empty collections;
• exact boundaries.

Example:

[Theory]
[InlineData(0)]
[InlineData(1)]
[InlineData(-1)]
[InlineData(int.MaxValue)]
[InlineData(int.MinValue)]
public void Rectangle_Contains_EdgeCases(int value)
{
    var rect = new Rectangle(0, 0, 100, 100);
    var point = new Vector2(value, 50);

    var result = rect.Contains(point);

    Assert.Equal(value >= 0 && value <= 100, result);
}

Property-Based Testing

• Generate 100-1000 random inputs for non-trivial pure logic.
• Check invariants, not only fixed expected values.
• Use FsCheck or the repo's existing property testing stack.
• Keep generated tests deterministic enough for CI diagnostics.

Performance Tests

• Use BenchmarkDotNet for performance-sensitive code.
• Enable allocation measurements for hot paths.
• Treat unexpected allocations in frame-loop code as failures unless justified.

Integration Tests

Test realistic subsystem flow, for example:

• input -> hit test -> widget events;
• network serialization -> deserialization -> handler;
• ECS setup -> system run -> component state.