Benchmark driver for incremental modular experiments

This is a driver for performing tests of intermod driver.

The test performed consists of analyzing different states of the sets of modules of a give directory. This is done by adding a number of random clauses each iteration of the test and performing the analysis afterwards.

Results of the analysis are left in the directory from which the test was runned.

Important!: Each time a benchmark is going to be rerunned the previous analysis directory has to be removed.

This module is thought to be used as an executable:

 ./exec <bench_name> <#changes/iteration> <edition_type> <domain> Opts

The seed is used to be able to repeat the same 'random' sequence in several tests.

For example:

./exec qsort 5 add gr incremental modular under_all --rand 1

will perform a test of adding 5 random clauses in the files of benchmark qsort with an incremental modular analysis.

The benchmarks that can be currently performed are present in test_dirs.pl. .