On Jun 13, 2012, at 06:26 PM, Colin Watson wrote: >> Meta-question: do you think it makes sense to turn on `from __future__ >> import unicode_literals`? > >We've talked about this on a few occasions. :-)
I know. I just can't seem to let this go. :) >This is more or less the poster child for a case where I think >unicode_literals is difficult. There's a complex API which is quite >sensitive to whether things are bytes or unicode in places, and often >handles both. The tests need to check whether both are handled. It >interacts with several other interfaces that prefer to work with bytes >in Python 2 but Unicode strings in Python 3. > >In short, I did try, but I found it to be much more complicated to try >to shift to unicode_literals than it was worth, and I really wasn't >confident enough in absolute 100% test coverage to be certain I hadn't >broken anything. Agreed, thanks for the explanation! Everything looks great; just one last suggestion. diff --git a/lib/debian/deb822.py b/lib/debian/deb822.py index 4c5b74e..7e8d0a6 100644 --- a/lib/debian/deb822.py +++ b/lib/debian/deb822.py > @@ -246,6 +246,8 @@ class Deb822Dict(object, UserDict.DictMixin): > # If we got here, everything matched > return True > > + __hash__ = None > + I think it would useful to include a comment for why this is here. It's a relatively obscure corner of the language, so it'll be helpful for the next person reading the code.
signature.asc
Description: PGP signature
