bpo-20182: AC convert remaining functions/methods in _hashopenssl.c#9213
Conversation
|
@serhiy-storchaka, thanks for the review! I've addressed your (excellent) comments and would be happy for another review. |
Sorry, something went wrong.
|
I removed the skipnews label. Only trivial PRs are allowed to skip news. @vstinner, @encukou, @serhiy-storchaka something is wrong with CODEOWNERS for the file. Any change to the hash module should be gated by @gpshead, Alex, or me. |
Sorry, something went wrong.
How is Or perhaps |
Sorry, something went wrong.
|
It's a mystery. @python/crypto-team exists and I'm a member. |
Sorry, something went wrong.
|
Can confirm it works well enough that I got an email :D |
Sorry, something went wrong.
|
I think that the NEWS file should contain only information that is interested for end users. Refactorings and conversions to Argument Clinic are not mentioned in it. |
Sorry, something went wrong.
gpshead
left a comment
There was a problem hiding this comment.
Thanks for the AC conversions and thanks everyone else for the already great reviews. This looks good.
Sorry, something went wrong.
|
@tiran, I'd like to re-apply the "skip news" label and merge this. Are you okay with that? |
Sorry, something went wrong.
|
I've added the skip news label for you (nothing news worthy about this internal refactoring of _hashopenssl.c). This PR likely needs syncing with merge conflicts addressed before it can go in. |
Sorry, something went wrong.
|
I'd gladly resolve the conflicts and merge this in. I was waiting for @tiran to approve this; but given that @serhiy-storchaka and @gpshead have approved I guess this is good to go. |
Sorry, something went wrong.
Notes:
_hashlib.HASHmethods, e.g._hashlib.HASH.copy, despite them not usingPyArg_Parse*argument parsing code.#include "clinic/_hashopenssl.c.h"line had to be moved down a bit in the file to resolve a compilation issue.https://bugs.python.org/issue20182