Motivation

Testing log files has shown to be a very demanding and time consuming task which usually ends up in enormous amounts of test code which can hardly be read and maintained. Why not better choose a semi-programmatically approach that allows for writing even complex tests which are still easy to understand, easy to maintain and very reliable? This tool allows for finite automata model based testing of log files.

Usage scenario

This tool can be used to test if a log file contains the expected output.

Usually, continuous development has two phases where tests are performed: Unit testing and integration testing. The most advanced of us may also use acceptance testing. While for unit testing it is quite easy to evaluate the outcome as the results can be verified programmatically, it is usually not that easy if the software runs integrated into its target environment. A very common technique is to write log files and analyze them later on. This is where this tool comes into play.

The following image gives an overview on where this tool integrates into the continuous delivery pipeline.

References

There are a lot of scientific papers dealing with this topic. If you need some more background on what is going on here, the following references are recommended.

• http://www.cs.ubc.ca/labs/isd/FormalWare/abstracts/logfiles.pdf
• http://sqa.stackexchange.com/questions/3594/how-common-is-testing-based-on-log-files-analysis
• http://www.ijcse.com/docs/INDJCSE12-03-05-082.pdf
• https://en.wikipedia.org/wiki/Automata-based_programming