Patch CPython to use a type reflection trampoline if possible #3964
Conversation
There was a problem hiding this comment.
Thanks, @hoodmane, it looks generally good to me, Let's see what CPython devs respond in the upstream PR.
Personally, I am okay with merging this even if the upstream PR is rejected for some reason, so this does not block other JSPI related features.
| + case 3: | ||
| + return ((three_arg)func)(self, args, kw); | ||
| + default: | ||
| + PyErr_SetString(PyExc_SystemError, "Handler takes too many arguments"); |
There was a problem hiding this comment.
What happens if a function falls into here? Is this a "should not happen" condition?
There was a problem hiding this comment.
Yes. Maybe we could fail more dramatically to make it more obvious. But I have never seen a function in the wild which takes too many arguments. The issue is always with functions that take too few arguments because they don't care about some of them.
|
Looks like they are likely willing to merge it, I got an approval. I'm actually pretty surprised that Brett rejected your |
Yeah, I think I'll try it again when they include ctypes module in their WASM build. |
This is a replacement for pyodide#3954 and pyodide#3955 structured as an upstream patch
|
Okay I'm going to merge this since I think the upstream PR is ready. Hopefully upstream will merge this soon too. |
This is a replacement for #3954 and #3955 structured as an upstream patch.
CF upstream PR:
python/cpython#106219