bpo-37335: Fix unexpected ASCII aliases in locale coercion tests #14443
Conversation
| "print(codecs.lookup(sys.getfilesystemencoding()).name)", | ||
| "print(codecs.lookup(sys.stdin.encoding).name + ':' + sys.stdin.errors)", | ||
| "print(codecs.lookup(sys.stdout.encoding).name + ':' + sys.stdout.errors)", | ||
| "print(codecs.lookup(sys.stderr.encoding).name + ':' + sys.stderr.errors)", |
There was a problem hiding this comment.
Oh wait, I added _PyUnicode_InitEncodings() in Python 3.8: filesystem and stdio encodings are now always normalized. https://bugs.python.org/issue36775 :
commit 43fc3bb7cf0278735eb0010d7b3043775a120cb5
Author: Victor Stinner <[email protected]>
Date: Thu May 2 11:54:20 2019 -0400
bpo-36775: Add _PyUnicode_InitEncodings() (GH-13057)
Move get_codec_name() and initfsencoding() from pylifecycle.c to
unicodeobject.c.
Rename also "init" functions in pylifecycle.c.
So _handle_output_variations() can simply be removed from 3.8, without having to add codecs.lookup().name here.
This change makes sense to Python 3.7. Would you mind to first create a new PR for master, backport it to 3.8, and then retarget this PR to 3.7?
There was a problem hiding this comment.
Yes, that will be the reason why Python 3.8 no longer has this problem. I will look into it.
Is there a way how to retarget this to 3.7? When we tried it previously, Github tried to backport the whole master into 3.7... Or do I have to create new PR for that as well?
There was a problem hiding this comment.
Well, maybe creating a new PR for 3.7 is simpler :-)
|
A Python core developer has requested some changes be made to your pull request before we can consider merging it. If you could please address their requests along with any other requests in other reviews from core developers that would be appreciated. Once you have made the requested changes, please leave a comment on this pull request containing the phrase |
|
I opened new other PRs for this change - closing this one. |
Solaris uses '646' as an alias for ASCII encoding and thus all the locale coercion tests are failing. This patch fixes the problem by using codecs lookup to normalize encoding names and thus prevent these problems.
https://bugs.python.org/issue37335