Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.python > #45836

Re: subclassing from unittest

From Terry Jan Reedy <tjreedy@udel.edu>
Subject Re: subclassing from unittest
Date 2013-05-23 15:42 -0400
References <07786a18-bb9f-4434-823c-d0fe00f858b0@cl9g2000vbb.googlegroups.com> <bar07a-pca.ln1@satorlaser.homedns.org>
Newsgroups comp.lang.python
Message-ID <mailman.2031.1369338168.3114.python-list@python.org> (permalink)

Show all headers | View raw


On 5/23/2013 2:58 AM, Ulrich Eckhardt wrote:

> Well, per PEP 8, classes use CamelCaps, so your naming might break
> automatic test discovery. Then, there might be another thing that could
> cause this, and that is that if you have an intermediate class derived
> from unittest.TestCase, that class on its own will be considered as test
> case! If this is not what you want but you still want common
> functionality in a baseclass, create a mixin and then derive from both
> the mixin and unittest.TestCase for the actual test cases.

This is now standard practice, gradually being implemented everywhere in 
the CPython test suite, for testing C and Py versions of a module.

class TestXyz():
   mod = None
   <test_a, etc, methods>

class TestXyz_C(TestXyz, TextCase):  # Test C version
   mod = support.import_fresh_module('_xyz')  # approximately right

class TestXyz_Py(TestXyz, TextCase):  # Test Python version
   mod = support.import_fresh('xyz')

This minimizes duplication and ensures that both implementations get 
exactly the same tests.

tjr

Back to comp.lang.python | Previous | NextPrevious in thread | Find similar | Unroll thread


Thread

subclassing from unittest Charles Smith <cts.private.yahoo@gmail.com> - 2013-05-22 08:32 -0700
  Re: subclassing from unittest Charles Smith <cts.private.yahoo@gmail.com> - 2013-05-22 08:47 -0700
  Re: subclassing from unittest Terry Jan Reedy <tjreedy@udel.edu> - 2013-05-22 17:29 -0400
  Re: subclassing from unittest Ulrich Eckhardt <ulrich.eckhardt@dominolaser.com> - 2013-05-23 08:58 +0200
    Re: subclassing from unittest Roy Smith <roy@panix.com> - 2013-05-23 08:43 -0400
    Re: subclassing from unittest Terry Jan Reedy <tjreedy@udel.edu> - 2013-05-23 15:42 -0400

csiph-web