[Python-Dev] doctest, unicode repr, and 2to3

Michael Foord fuzzyman at voidspace.org.uk
Fri Mar 5 17:11:59 CET 2010

On 05/03/2010 15:56, Antoine Pitrou wrote:
> Le Thu, 4 Mar 2010 23:30:12 -0500,
> Barry Warsaw<barry at python.org>  a écrit :
>> If you really want to test that it's a unicode, shouldn't you
>> actually test its type?  (I'm not sure what would happen with that
>> under 2to3.)  Besides, the type of the string is very rarely
>> important, so I think the u-prefix and quotes is mostly just noise.
> String type is actually very important, if you don't want your
> application/library to fail in the face of non-ASCII data.
> That's why we did all this thing in py3k, after all :)

Well, I'd like to see a 'unicode agnostic' mode for different reasons - 
although given that it all changes in Python 3 and it is almost 
certainly too late to get changes to doctest before the 2.7 release the 
point is probably moot.

IronPython has very different unicode behaviour to Python 2. IronPython 
is like Python 3 in this regard, all strings are unicode (str is 
unicode). This means that it rarely puts the u prefix on strings. So 
doctests that use string repr also fail on IronPython - the django 
doctest suite is a culprit in this regard too.

Dictionary iteration order is also different in IronPython - so dict 
reprs also fail. For dict reprs it would be nice if doctest compared in 
an order agnostic way. (Internally reconstruct the dict as a dict of 
strings from the repr and then compare using dict equality.)

I know it is easily *possible* to construct doctests that are robust 
against these differences, but the fact that the "obvious-thing" often 
fails in subtle ways on other platforms are *one* of the reasons I don't 
consider doctest to be a suitable tool for unit testing [1]. :-)


[1] It isn't the main reason and I realise that is entirely orthoganal 
anway... but doctest makes every line an assertion, and you have to jump 
through hoops if you don't want that behaviour. All besides the point. 
Doctest simply *rocks* for testing documentation examples - especially 
in conjunction with Sphinx.

> Regards
> Antoine.
> _______________________________________________
> Python-Dev mailing list
> Python-Dev at python.org
> http://mail.python.org/mailman/listinfo/python-dev
> Unsubscribe: http://mail.python.org/mailman/options/python-dev/fuzzyman%40voidspace.org.uk


READ CAREFULLY. By accepting and reading this email you agree, on behalf of your employer, to release me from all obligations and waivers arising from any and all NON-NEGOTIATED agreements, licenses, terms-of-service, shrinkwrap, clickwrap, browsewrap, confidentiality, non-disclosure, non-compete and acceptable use policies (”BOGUS AGREEMENTS”) that I have entered into with your employer, its partners, licensors, agents and assigns, in perpetuity, without prejudice to my ongoing rights and privileges. You further represent that you have the authority to release me from any BOGUS AGREEMENTS on behalf of your employer.

More information about the Python-Dev mailing list