Propagate foast & past typing information to ITIR #1199
Conversation
tehrengruber
changed the title
| ) | ||
| inferred = ti.infer(testee) | ||
| assert inferred == expected | ||
| assert ti.pformat(inferred) == "(It[T₁, T₂, float⁰]) → It[T₁, T₂, float⁰]" |
There was a problem hiding this comment.
What is being tested here that hasn't been in the first assert?
There was a problem hiding this comment.
I think the idea of @fthaler was to test the type pretty printer like this. Additionally this gives a more readable representation of the inferred type.
| if node.kind: | ||
| kind = {"Iterator": Iterator(), "Value": Value()}[node.kind] |
There was a problem hiding this comment.
It's not very important, but I think it would better to use explicit pattern matching:
| if node.kind: | |
| kind = {"Iterator": Iterator(), "Value": Value()}[node.kind] | |
| match node.kind: | |
| case "Iterator": | |
| kind = Iterator() | |
| case "Value": | |
| kind = Value() | |
| case _: | |
| raise AssertionError(...) |
There was a problem hiding this comment.
I see the point, but I like the less verbose version more.
There was a problem hiding this comment.
Honestly, for just two cases, I don't think the dict version brings any value. Even a simple if-else would be better.
|
@tehrengruber What is the current state of this PR? Is it blocked by something else? |
|
@egparedes you never clicked approve on this ;) we're waiting for your review. |
Adds support for constraining the type of symbols in ITIR (only
dtypeandkindright now) and propagates the information from FOAST and PAST to ITIR.