[incr TSDB()]

There is code that generates output for the profiling and test suite management tool [incr TSDB()] (http://www.coli.uni-sb.de/itsdb/) developed by Stephan Oepen. If you want to implement larger grammar fragments it is recommended to use this tool. It provides the following functionality (and much more):

• storing time, number of passive edges, memory requirements, errors for every parsed item
• compare test runs (performance, coverage, overgeneration)
• detailed comparison on item basis of
  • number of readings/edges
  • the derivations (i.e. tree structure with rule names)
  • the MRSes
• test parts of test suites on a phenomenon basis or some other selection from your test items (restrictions may be formulated in SQL queries)

The [incr TSDB()] uses the Parallel Virtual Machine. If you have several CPUs idle you may distribute the processing of your test suite over several machines which enormously shortens the time needed for testing and speeds up grammar development.