-
Notifications
You must be signed in to change notification settings - Fork 113
Missing call to FunctionDeclarationInstantion in EvaluateBody #38
Comments
Cool, thanks for clarifying. Yeah this is an interesting one. Instantiating a new function to essentially contain the static field initializations certainly seems like the least drastic approach, e.g.
becoming kind of like
I'd figure that it should just be around the final new steps that prep the environment and call Edit: The rest of this was me misreading the spec, woops!
|
Computed field keys are still evaluated once, in the same scope and at the same time where the computed method keys are evaluated. |
For example, |
Oh shoot, you are totally right, I misread. Disregard the second half of that then. |
To answer the first part, I thought this would be better to add in EvaluateBody, which would match now normal functions do it. Please, keep the clarifying questions and bug reports coming; I really appreciate the work you're doing here. |
Could you clarify what you mean about that? Not quite making the jump from |
I think the outer context is captured by the FunctionCreate call from ClassFieldDefinitionEvaluation . |
Sorry, looking at all this again, I think the current spec already has the mechanics for the sequence of events, which is:
There are a couple editorial issues in this code path which I'll fix up. |
Thanks for looking again, this does seem to cover it, so we should be all set. Following the chain of events through the calls was tripping me up a bunch. I think I had just missed that there was that |
This is all tricky stuff; you can see that I didn't understand it immediately either! |
We need this call, or an analogous lower-level thing, to set up scopes. See this thread for context. Thanks @loganfsmyth for pointing out the issue.
The text was updated successfully, but these errors were encountered: