You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right now categorize simply gets stuck in a loop when we try to categorize a mutually-recursive function.
I think the problem is tied to inlining, so it would be great if we could keep track of what we've inlined, and error if we try to inline a function that we're in the middle of converting already.
The problem with this approach currently is that we usually inline the original expression within a larger unconverted expression, then convert the larger expression, so it's possible that if we just track it in the larger expression that we're noticing parallel invocations of the same function, rather than nested ones.
Perhaps we could manually introduce a Tick node whenever we inline, so then we can notice that we hit the inlined identifier within the larger expression.
Right now
categorize
simply gets stuck in a loop when we try to categorize a mutually-recursive function.I think the problem is tied to inlining, so it would be great if we could keep track of what we've inlined, and error if we try to inline a function that we're in the middle of converting already.
The problem with this approach currently is that we usually inline the original expression within a larger unconverted expression, then convert the larger expression, so it's possible that if we just track it in the larger expression that we're noticing parallel invocations of the same function, rather than nested ones.
Perhaps we could manually introduce a
Tick
node whenever we inline, so then we can notice that we hit the inlined identifier within the larger expression.(Extricated from https://kitty-hawk.atlassian.net/browse/SW-2733)
The text was updated successfully, but these errors were encountered: