In nose.plugins.errorclass.ErrorClassPlugin.patchResult, we monkey patch a
non-standard TestResult instance so that it emulates the error handling
functionality of Nose's TextTestResult.
When we do this, we store the original methods on the result instance, named
(for example) _orig_addError. We also have a couple of unit tests (in
test_skip_plugin and test_deprecated_plugin) to verify that these original
methods exist.
My question is, why do we do this? They do not appear to be used (apart from
the unit tests), and creating them makes ErrorClassPlugin harder to understand.
Original issue reported on code.google.com by gazi...@gmail.com on 24 Nov 2011 at 6:57
Original issue reported on code.google.com by
gazi...@gmail.com
on 24 Nov 2011 at 6:57