Arity checks should be changed to use Null instead of dynamic #674
Comments
natebosch
added a commit
that referenced
this issue
Aug 22, 2017
Fixes #674 dynamic is treated as the bottom type for function arguments, but this will soon change. Null will continue to be the bottom type. The question we are asking is "Is this a function that takes one argument", when dynamic is no longer the bottom type the question will become "Is this a function that takes one dynamic argument", and a function which take one argument of a specific type does *not* take a dynamic argument.
natebosch
added a commit
that referenced
this issue
Aug 22, 2017
Fixes #674 dynamic is treated as the bottom type for function arguments, but this will soon change. Null will continue to be the bottom type. The question we are asking is "Is this a function that takes one argument", when dynamic is no longer the bottom type the question will become "Is this a function that takes one dynamic argument", and a function which take one argument of a specific type does *not* take a dynamic argument.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This package (including at least code in expect_async) uses the pattern of defining a function type taking
dynamicarguments and using this to test arity via anischeck:This currently works in DDC because of the special treatment of function types with dynamic arguments, but this treatment is going away now that we have moved
Nullto the bottom of the hierarchy and can use that. This code needs to be updated to use the following essentially almost equivalent pattern instead:Note that this is only mostly equivalent, in that it changed what
fgets promoted to. Iffis called in the scope of the promotion toFunc1, this call must be made explicitly dynamic to recover the previous semantics:(f as dynamic)(args).The text was updated successfully, but these errors were encountered: