From: Bryce H. <br...@os...> - 2005-12-19 17:02:33
|
Hi all, A lot of us are running test harnesses. Often we're running the same sorts of tests. However, every test seems to have its own unique output format for its results. A common challenge we share is the ability to share these results between different tools for analysis, reporting, and data mining needs. Test::Parser is a new project aimed at addressing this problem by gathering together parsers for different tests. Results are stored into a standard data structure that provides your harness with a standard API to access test results. For example, this would enable you to create a general purpose test result browser for your test harness that can display results from any supported test. As people add new parsers for more tests, your browser becomes more broadly useful. You can download release 1.00 of Test::Parser from CPAN: http://search.cpan.org/~bryce/Test-Parser/ This is the first official release of this codebase. In addition to the base code, it includes the following parsers: * Kernel build results * Linux Test Project (LTP) test runs This package also includes Test::Metadata, a module for collecting multiple parsed test runs and attaching metadata to them including name, path, report info, expected results, and suite type. This module is able to print out SpikeSource's TRPI XML format for describing test runs. More information about TRPI is available from SpikeSource's website. Future plans include: * Parser for PyNFS * Parser for Connectathon * Parser for Iozone * Ability to export to tslogparser (see tslogparser.sf.net) Adding a new parser is not difficult, and takes on the order of an afternoon and under 100 lines of Perl. If you'd like to contribute a new parser for a test, I'll be happy to answer any questions you have! Thanks, Bryce |