[3.13] fix: remove ast.MatchStar.name class-level default value (gh-134674)

[hunterhogan]

I inappropriately submitted this to the main branch in pull request #134676, which is now closed.

Fixes gh-134674

Problem

ast.MatchStar.name incorrectly appears to have a class-level default value when using ast.dump(). The issue causes ast.MatchStar(name=None) to display as MatchStar() instead of the expected MatchStar(name=None).

Solution

Change the ASDL definition of MatchStar, and add MatchStar to the special cases list in ast.dump() alongside Constant and MatchSingleton.

[picnixz]

• Why was this opened against 3.13? is it 3.13 only?
• If not, please checkout from main and open a PR against main. We will make the backports if needed.

[hunterhogan]

• Why was this opened against 3.13? is it 3.13 only?

• If not, please checkout from main and open a PR against main. We will make the backports if needed.

I inappropriately submitted this to the main branch in pull request #134676, which is now closed.

[picnixz]

Yes, but why was it submitted incorrectly? all PRs should be opened against main. AFAICT, the other PR was based from 3.13 and not main. Or is the issue only in 3.13?

[hunterhogan]

Yes, but why was it submitted incorrectly? all PRs must be opened against main. afaict, the other PR was based from 3.13 and not main. Or is the issue only in 3.13?

#134676 (comment)

Why? I am diligently trying to follow instructions.

I am a little frustrated that you are asking me questions that are certainly answered in the first pull request or in the Issue.

I cannot do this the right way if I don't understand what the right way is.

[picnixz]

that are certainly answered in the first pull request or in the Issue

I can't find in the issue whether the issue is 3.13 only or not. If it's a 3.13 issue, then it's fine, but AFAICT, the current behaviors between 3.13 and 3.15 look the same. For that reason, 3.13, 3.14 and 3.15 are affected by this issue and thus we should fix 3.15 (which is main) and then we backport the changes to 3.14 and 3.13 (this is how it's done).

[picnixz]

#134676 (comment)

This comment was perhaps incomplete but the idea was to:

• Checkout from main
• Make the changes.
• Open a PR against main.

It would have been fine to checkout from 3.13, make the changes and open a PR against 3.13 only if the bug is present in 3.13 but not in 3.14 or 3.15 (but AFAICT, this is not the case here)

[picnixz]

And I apologize if I was rude, that wasn't my intention!

[hunterhogan]

And I apologize if I was rude, that wasn't my intention!

Thank you very much for the additional information! I was confused. I sincerely hope I understand things well enough now to submit a well-formed pull request.

[picnixz]

Don't worry, we'll go through together! It's 3:24 AM here so I'll be away but I'll definitely have a look at your PR. I'm sorry if I sounded rude with my questions that were maybe a bit too direct

[hunterhogan]

It's all good. You believed I had as much knowledge as you have, which is a compliment to me because I barely understand git or this process for contributing to CPython.