Software Agitation of a Dynamically Typed Language

Much research has been conducted on automated testing tools. Recently, Agitar Technologies developed Agitator™, which combines input generation and invariant detection into \u27software agitation\u27. Agitator analyzes code to automatically detect and report likely program invariants. However, Agitator only operates on Java, a statically typed language. I apply Agitator\u27s techniques to Python, a dynamically typed language, in a tool called PyStick. PyStick implements some alternative techniques to Agitator due to the differences between Java and Python. Since PyStick is a proof of concept, it does not cover the entirety of Python. It has nonetheless generated very promising results, and demonstrates the power and viability of software agitation on a dynamically typed language