Allow lazy implementations to raise from intrinsically eager functions #652
Conversation
There was a problem hiding this comment.
Thanks @cbourjau. This looks like a good start. I added a few minor comments. Then, I think __complex__ should get the same note, right?
My main suggestion for more context is perhaps something that could go in a short design document on lazy/eager, that could be put under https://data-apis.org/array-api/latest/design_topics/index.html. That also gives us a place to add other future lazy evaluation related topics. I am happy to draft that as a follow-up to this PR; we can get this merged first to clarify things.
|
Thanks for the review! I addressed your comments as suggested. |
|
Is there anything missing in this PR before it could be merged? |
No, there isn't. My array API review bandwidth disappeared for a bit. In it goes, thanks again @cbourjau! |
I'll need to do this now. |
Follow-up to data-apisgh-652, which added notes to the specifications for `__bool__` & co on this topic.
|
Follow-up in gh-708. |
Follow-up to data-apisgh-652, which added notes to the specifications for `__bool__` & co on this topic.
Follow-up to gh-652, which added notes to the specifications for `__bool__` & co on this topic.
Allow intrinsically eager functions (such as
array.__bool__) to raise in lazy implementations. Closes #642