This repository is hosting the source code and scripts implementing FastDiag and Parallel FastDiag, the scripts allows the following paramenter:

• model: A CNF file depicting the constraints representing the background knowledge for the algorithm (e.g. the constraints in a feature model)
• product: A CNF file representing the product selection or user requirements to look for the confict set.
• lmax: The number of look aheads to perform per step in the algorithm
• solver: The backend solver to use, we support at this time, Glucose3, Sat4j and Choco.
• difficulty: A parameter to inject more difficult consistency checks, any integer representing secods can be used (dafaullt to zero)

Also, the following scripts are provided:

• fastdiag.py: The implementation of the traditional FastDiag algorithm.
• fastdiag_parallel_mp.py: An implementation o the Parallel FastDiag relying on the multiprocessing python package.

Finally, also provide:

• A set of models with a fixed set of conflict sets.
• The results shown in the associated paper.