Hello, On Thu, Apr 23, 2009 at 10:55 PM, spir <denis.s...@free.fr> wrote:
> > My app is about parsing, which input and output usually both are big and > complicated *strings*. So that I find the command line model really > unappropriate for expressing test cases and their expected results. > Interesting - haven't really used Python's test frameworks here - and perhaps my suggestion may be too basic - and if so I apologise - not sure what you mean by command line model in the context. Could you not have a directory with input files and expected output files, then write a test runner going through input -> apply_operation -> assert(output) cycles? The test runner can be as simple or as complicated as you want of course, so you can configure tests say by pieces of parsing functionality, but they need not deal with the tests' input themselves, which will mostly be preconfigured (or partially dynamic with plugged in values) etc. Heck, you could have configuration files for the test runner! Have I completely missed your question? -- K -- There is more to life than increasing its speed. -- Mahatma Gandhi
_______________________________________________ Tutor maillist - Tutor@python.org http://mail.python.org/mailman/listinfo/tutor