GeezerVison

I renamed utility functions for testing to conform with PEP8.

Also I added URLs in comments to documentation on COM specifications for developers who may have limited experience with COM and may review or maintain this code in the future.

In #126686, COMError became public, but I would like this test to be backported to 3.13 and 3.12, where COMError remains private.

If possible, I would appreciate it if we could merge this as is, and after the backport PR is submitted by the bot, I will prepare a follow-up PR reflecting the fact that COMError is now public.

[Edit: I deleted a misguided suggestion]

!buildbot Windows

🤖 New build scheduled with the buildbot fleet by @encukou for commit cc07c44 🤖

The command will test the builders whose names match following regular expression: Windows

The builders matched are:

• ARM64 Windows Non-Debug PR
• AMD64 Windows11 Bigmem PR
• AMD64 Windows10 PR
• AMD64 Windows11 Non-Debug PR
• AMD64 Windows Server 2022 NoGIL PR
• AMD64 Windows11 Refleaks PR
• ARM64 Windows PR

I don’t have a strong preference for the order of actual == expected or expected == actual.
If it’s mergeable as is, please go ahead and merge it.

In my native language, Japanese, the equivalent concept is expressed as “予実” (yojitsu), with the characters for “expected” and “actual” written in that order.
That might be why I wrote it in this order.

I added the block.

Ah thank you. I wasn't sure whether it was needed or not. But one build bot is failing I think?

I’m not sure which bot you’re referring to, but all the CI checks have passed, and there’s a green ✔ as well.

!buildbot Windows

🤖 New build scheduled with the buildbot fleet by @picnixz for commit 6c60512 🤖

The command will test the builders whose names match following regular expression: Windows

The builders matched are:

• ARM64 Windows Non-Debug PR
• AMD64 Windows11 Bigmem PR
• AMD64 Windows10 PR
• AMD64 Windows11 Non-Debug PR
• AMD64 Windows Server 2022 NoGIL PR
• AMD64 Windows11 Refleaks PR
• ARM64 Windows PR

I misunderstood the type of bot running in CI.
I thought all the bots specified with !buildbot Windows were running in CI.

I’m not sure why the bot is failing.
There are no changes to production code this time.
Looking at the bot’s traceback, it seems that something related to f-strings isn’t working correctly, but the test code I added this time doesn’t use any f-strings.

!buildbot Windows

You don't have write permissions to trigger a build

Sorry, I had to step away yesterday.

By the way, do we need a if name == 'main' block executing the unittest as is?

No, that's not necessary in new tests. But it doesn't hurt.

I misunderstood the type of bot running in CI.

The Buildbot tests run after merge, as they can be slow or unreliable. And they run on volunteer-maintained hardware; for security we only run them on reviewed code.
I triggered them in order to check the test on several versions of Windows, e.g. 32-bit or ARM.
Sometimes they find issues in uncommon setups. Usually, some fail for unrelated reasons; those can be ignored. (GitHub unfortunately shows a scary-looking red message if any reported tests fail.)

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

Thanks @junkmd for the PR, and @encukou for merging it 🌮🎉.. I'm working now to backport this PR to: 3.12.
🐍🍒⛏🤖

GH-127159 is a backport of this pull request to the 3.13 branch.

GH-127160 is a backport of this pull request to the 3.12 branch.