[Pyunit-interest] Automatically generated tests

[Magnus L. H. <ma...@he...>]

Hi!

I've been using the default test loader, writing several tests in the
same TestCase (using the method naming convention method). This seems
fine for hard-coded tests, but I keep running into the need to
auto-generate tests, by trying out various combinations of parameters,
for example (i.e., having a few loops in my test method). The problem
is that I'd like to run each case (i.e., combination of parameters)
with the same fixture. In other words, I'd like to have a separate
test (with a separate run of setUp/tearDown) for each case -- but
without having to write separate test methods for each of them.

I guess I could have separate code for creating individual TestCase
objects separately and add them to a common suite etc. -- but using
the standard unittest.main() method is quite convenient... :]

What is the "best practice" for automatically generated tests in
PyUnit?

-- 
Magnus Lie Hetland               Time flies like the wind. Fruit flies
http://hetland.org               like bananas.         -- Groucho Marx