gh-131878: Fix input of unicode characters with two or more code points in new pyrepl on Windows

[sergey-miryanov]

If unicode characters with two or more codepoints (ñ or é) typed or pasted, then it should be converted to bytes before passing to eventqueue.push.

Also fixed handling of exceptions while decoding buffer. If exception occurred then buffer should be flushed to prevent mixing it with following control commands (for example "\x1b[201").

• Issue: New REPL on Windows exits when accented character is pasted/typed #131878

[python-cla-bot]

All commit authors signed the Contributor License Agreement.

[chris-eibl]

LGTM. I cannot do a thorough review, but at least confirm, that with this PR I can enter äÄöÖüÜ, etc, again 🚀

[tomasr8]

Tested with a Czech keyboard and everything works. Thanks!

[chris-eibl]

LGTM. I cannot do a thorough review, but at least confirm, that with this PR I can enter äÄöÖüÜ, etc, again 🚀

Had some time now: this needs fixes for Linux. Otherwise looks good, thanks @sergey-miryanov!

[tomasr8]

This breaks Linux, because only one char is inserted in UnixConsole.get_event() [...]

Since this wasn't caught by the tests, we should also probably add a test for that.

[chris-eibl]

I have suggested the def test_push_unicode_character_two_bytes(self): which tests exactly that.

[sergey-miryanov]

@chris-eibl Please take a look! I would like to keep tests for paste mode to be sure that escape sequences not mixed with input if error occurred. But I'm OK to remove them if you think it is not necessary.

[chris-eibl]

LGTM. Sure, more tests are always better. Though, AFAIU, there is nothing special about paste mode here:

• test_push_unicode_character_two_bytes_in_paste_mode just asserts that a valid "two byte char" within a sequence of "one byte chars" is working as expected.
• likewise, test_push_unicode_character_as_str_in_paste_mode asserts that when an exception during a sequence of feeding bytes via push is caught, we can continue pushing afterwards.

IMHO, both are reasonable, but "paste mode" makes them look too special about pasting?

So I'd rename them and use arbitrary "one byte chars" in there.

[sergey-miryanov]

Sequence of "one byte chars" is a control command to enable "paste mode". Originally in gh-131878 problem aroise from case where the wrongly passed unicode string puts to the buffer and mixed with following control command that disables paste mode and asserted in those code path. So, I want to check that we don't "corrupt" chars buffer.

[chris-eibl]

Yupp, I know. But as said, there is nothing special about bracketed pasting here. From the perspective of the eventqueue test, these are just arbitrary bytes. Whether bracketed pasting is working correctly, is tested in

cpython/Lib/test/test_pyrepl/test_pyrepl.py

Line 1201 in 1b7470f

deftest_bracketed_paste(self):

If you'd like to showcase that sequences of "one byte chars" interleaved with "multi byte chars" work correctly via "paste mode", then maybe just add a comment?

And small nit: ~ is missing in the escape sequences, see

cpython/Lib/test/test_pyrepl/test_pyrepl.py

Lines 1227 to 1228 in 1b7470f

	paste_start="\x1b[200~"
	paste_end="\x1b[201~"

To me, those two tests have merit, but are too special about "paste mode".

[sergey-miryanov]

@chris-eibl It seems that I did not correctly assume how paste-mode works (shame on me!). I removed one test because the same code path is covered by test_push_unicode_character_two_bytes. And renamed and simplified a bit test_push_single_chars_and_unicode_character_as_str (not sure the name is ok though).

[chris-eibl]

I don't have a better name - the comment in there clarifies what we intend to test (not much tbh).
Just a small nit on the comment.

Otherwise LGTM. Thanks for bearing with me!

[sergey-miryanov]

@chris-eibl Thanks for the review and patience!

[chris-eibl]

Thank you @sergey-miryanov!

[tomasr8]

Tested again after the simplifications. WFM on both Windows and Linux :)

[ambv]

Gentlemen, excellent work. Not only does it fix the bug, but it makes the codebase simpler and more elegant. If not for the new test, the net number of lines added here would be negative. Impressive!

[miss-islington-app]

Thanks @sergey-miryanov for the PR, and @ambv for merging it 🌮🎉.. I'm working now to backport this PR to: 3.13.
🐍🍒⛏🤖

[miss-islington-app]

Sorry, @sergey-miryanov and @ambv, I could not cleanly backport this to 3.13 due to a conflict.
Please backport using cherry_picker on command line.

cherry_picker 0c5151bc81ec8e8588bef4389df12a9ab50e9fa0 3.13

[sergey-miryanov]

@ambv Please take a look - this is the same problem with backport as in earlier PR - #130805 (comment)

[bedevere-app]

GH-133468 is a backport of this pull request to the 3.13 branch.

[bedevere-bot]

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Hi! The buildbot AMD64 Debian root 3.13 (tier-1) has failed when building commit 891232f.

What do you need to do:

1. Don't panic.
2. Check the buildbot page in the devguide if you don't know what the buildbots are or how they work.
3. Go to the page of the buildbot that failed (https://buildbot.python.org/#/builders/1441/builds/1169) and take a look at the build logs.
4. Check if the failure is related to this commit (891232f) or if it is a false positive.
5. If the failure is related to this commit, please, reflect that on the issue and make a new Pull Request with a fix.

You can take a look at the buildbot page here:

https://buildbot.python.org/#/builders/1441/builds/1169

Summary of the results of the build (if available):

Click to see traceback logs

remote: Enumerating objects: 26, done. remote: Counting objects: 3% (1/26) remote: Counting objects: 7% (2/26) remote: Counting objects: 11% (3/26) remote: Counting objects: 15% (4/26) remote: Counting objects: 19% (5/26) remote: Counting objects: 23% (6/26) remote: Counting objects: 26% (7/26) remote: Counting objects: 30% (8/26) remote: Counting objects: 34% (9/26) remote: Counting objects: 38% (10/26) remote: Counting objects: 42% (11/26) remote: Counting objects: 46% (12/26) remote: Counting objects: 50% (13/26) remote: Counting objects: 53% (14/26) remote: Counting objects: 57% (15/26) remote: Counting objects: 61% (16/26) remote: Counting objects: 65% (17/26) remote: Counting objects: 69% (18/26) remote: Counting objects: 73% (19/26) remote: Counting objects: 76% (20/26) remote: Counting objects: 80% (21/26) remote: Counting objects: 84% (22/26) remote: Counting objects: 88% (23/26) remote: Counting objects: 92% (24/26) remote: Counting objects: 96% (25/26) remote: Counting objects: 100% (26/26) remote: Counting objects: 100% (26/26), done. remote: Compressing objects: 33% (1/3) remote: Compressing objects: 66% (2/3) remote: Compressing objects: 100% (3/3) remote: Compressing objects: 100% (3/3), done. remote: Total 14 (delta 12), reused 12 (delta 11), pack-reused 0 (from 0) 
From https://github.com/python/cpython
* branch 3.13 -> FETCH_HEAD
Note: switching to '891232f3386dd8b20a216a473954c1b01cede7ec'.
You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by switching back to a branch.
If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -c with the switch command. Example:
git switch -c <new-branch-name>
Or undo this operation with:
git switch -
Turn off this advice by setting config variable advice.detachedHead to false
HEAD is now at 891232f3386 [3.13] gh-131878: Fix input of unicode characters with two or more code points in new pyrepl on Windows (gh-131901) (gh-133468)
Switched to and reset branch '3.13'
configure: WARNING: no system libmpdecimal found; falling back to bundled libmpdecimal (deprecated and scheduled for removal in Python 3.15)configure: WARNING: pkg-config is missing. Some dependencies may not be detected correctly.find: ‘build’: No such file or directoryfind: ‘build’: No such file or directoryfind: ‘build’: No such file or directoryfind: ‘build’: No such file or directorymake: [Makefile:3116: clean-retain-profile] Error 1 (ignored)