Closed 0dminnimda closed 1 year ago
Is it reasonable to first try to merge it to cpython?
Generally we do it the other way: we merge stuff here and I sync them from time to time in batches.
Ok, I see. Also can I ask you for the launch of the workflow, please?
Probably a good time to say, that two tests are failing on the main
. The same tests are failing here. Although I think fixing those tests is out of the scope of this pr, unfortunately
Getting #66 in first would make sense to get CIs to test on more versions.
Does #66 fixes broken tests? If yes, then we probably should try merging those fixes first independently of the #66.
Generally we do it the other way: we merge stuff here and I sync them from time to time in batches
Well, then what's about distributing credit/blame for the changes? Will those changes be cherry-picked into cpython?
Generally we do it the other way: we merge stuff here and I sync them from time to time in batches
Well, then what's about distributing credit/blame for the changes? Will those changes be cherry-picked into cpython?
No, they will be directly included in one batch and blame/credit is lost.
No, they will be directly included in one batch and blame/credit is lost.
Aww, isn't this a reason to make a direct pr to cpython?
Well, not like it's that important, but it would be nice if the credit/blame was preserved!
Aww, isn't this a reason to make a direct pr to cpython?
No, we don't generally merge individual PRs in CPython for pegen because it makes syncing much more complicated.
Well, not like it's that important, but it would be nice if the credit/blame was preserved!
Well, the canonical code for pegen
is here. The CPython version doesn't match this code because it has a bunch of Cpython-isms and the C generator which is is own beast. Notice that the only python parser being generated in CPython is the grammar parser, but the main output is the C version.
@0dminnimda Can you rebase on main so that we get an extended test run ?
Oh yeah, workflow activation, what a lovely github feature ♥
Bye, I am leaving this repo now.
oh .. bye
well, we need to get back to the work
3.10 didn't pass, it seems like versoin of python that tox runs is somehow different from the other ones
Because if I try replicate this I see a string without '
I don't quite understand why it happens
Closing and opening to triger the ci
oh yeah, forgot to merge
Is there any additional comments or this can be merged?
I find even more concerning that no test found the issue
Should I add the tests for it here? Tho seems appropriate for another pr
Maybe @pablogsal have other comments?
Maybe @pablogsal have other comments?
I will try to review this soon but currently, I am very busy with the 3.11.0b5 release :(
@MatthieuDartiailh can you allow the workflow?
Done
@pablogsal would you have time to look at this ? or do you want to defer the final review to me maybe (since you seem quite busy ATM).
@pablogsal would you have time to look at this ? or do you want to defer the final review to me maybe (since you seem quite busy ATM).
I apologize for the late reply @MatthieuDartiailh. Unfortunately the release of 3.11 is being ridiculously challenging and I'm expending all my OSS time on that :(
Please, go ahead and merge it if you feel confident. I trust your criteria :+1:
I wanted to add a couple small comments but apparently I cannot push to your repo. S I will merge and add the comment directly on main.
Weird usually there's no problem in that Maybe i accidentally checked out Allow edits by maintainers..
Is it reasonable to first try to merge it to cpython?