On Mon, 2011-10-31 at 10:01 +0100, Zdeněk Pavlas wrote: > Under some locales, python creates UTF8 encoded non-unicode > exceptions. For such, exception2msg() returns values that > fail to convert to unicode.
In general we can't have generic functions (like exception2msg()) returning str and unicode in different code paths ... it almost guarantees unicode errors. Also changing the return type is probably bad, even though it's relatively new. I think I originally had it return str() objects because it was replacing code like: "str(e)" with "exception2msg(e)" ... although not it's used in a few places where it'd be better as unicode now. So maybe another function (uexception2msg?) or just lots of to_unicode calls? _______________________________________________ Yum-devel mailing list Yum-devel@lists.baseurl.org http://lists.baseurl.org/mailman/listinfo/yum-devel
