Danek Duvall wrote: > Danek Duvall wrote: > > > Saurabh Vyas wrote: > > > > > https://cr.opensolaris.org/action/browse/pkg/saurabhv/CR-7127117-rev1/webrev/ > > > > Please put commonize this into a function. That function should take the > > input str object and return a unicode object after doing the three tests, > > and the final ascii/replace fallback. Also, please use double quotes. > > And it actually might make better sense to simply find the encoding of the > output stream, and attempt only that before trying the replacement version. > If you were going to sys.stdout, then you could use sys.stdout.encoding. > I'm not sure how to get to that through the logging interfaces, but I'm > sure there is one, if you dig in the right place. But if you can find > that, then you should use that, and only that. Otherwise, your current > fallback is fine.
Sorry for the stream of consciousness, but it occurred to me ask what the encoding was if the output were simply redirected to a file, and that turns out to be None, just as if it went to a pipe, and that seems annoying. So then it occurred to me whether (and if so, why) we should settle for ascii output to either a file or a pipe, and frankly, I don't know the answer to that, never mind what to do in that case. Perhaps we can use your fix as a temporary workaround, but I don't think this is the final answer. Danek _______________________________________________ pkg-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
