Pex is a parameterised unit testing tool based on dynamic symbolic execution. To generate test data it uses the Z3 constraint solver. Sometimes constraints involving floating point variables (and decimal types) prevent Pex from exploring certain execution paths through your program. This Pex extension contains a collection of meta-heuristic search algorithms which can be used to increase Pex's code coverage for floating point rich applications. It is open source and can be downloaded from codeplex.

Click here to download.

Structural evolutionary testing degenerates to random testing in the presence of flag variables. This tool implements a testability transformation along with a prototype tool that transforms programs with loop assigned, simple and function assigned flags into flag free equivalents, so that existing test data generation approaches can be applied successfully.

Click here to download.

AUgmented Search based TestING (AUSTIN) combines a simple hill climb with a set of constraint solving rules for pointer type inputs. The algorithm has been implemented in a prototype tool named AUSTIN. Similar to dynamic symbolic execution based testing, AUSTIN instruments the program under test to perform a kind of pseudo symbolic execution of the program along the concrete path of execution for a particular input vector. When execution diverges away from the selected target at a critical branching node, AUSTIN either uses a custom procedure to solve constraints over physical memory locations, or employs an alternating variable method to find values for input parameters that lead successive executions of the program closer to the target.

Click here to download Austin.