handle scope correctly for @isdefined #22281
Conversation
this allows testing whether a variable is defined, and may be especially useful for emitting from a macro (which may want to conditionally initialize a variable)
vtjnash
force-pushed
the
jn/isdefined-local
branch
from
e2acbaf
to
ef030bd
Compare
| @@ -357,6 +357,7 @@ static Function *jlboundserrorv_func; | |||
| static Function *jlcheckassign_func; | |||
| static Function *jldeclareconst_func; | |||
| static Function *jlgetbindingorerror_func; | |||
| static Function *jlboundp_func; | |||
There was a problem hiding this comment.
what is "boundp" supposed to mean?
There was a problem hiding this comment.
I don't name the functions, I just use them 😆
There was a problem hiding this comment.
wow that's old 691b8be and used in a grand total of 3 places before this PR - maybe it should get a new name that says what it does?
There was a problem hiding this comment.
boundp is a common lisp function for testing whether a variable is bound: http://www.ai.mit.edu/projects/iiip/doc/CommonLISP/HyperSpec/Body/fun_boundp.html
There was a problem hiding this comment.
that explains where the name came from, but doesn't help anyone reading the code who doesn't already know that figure out at a glance what it does
There was a problem hiding this comment.
The intent of my comment was to explain where the name came from.
There was a problem hiding this comment.
What's the conclusion, do I need to add renaming of this function to the PR, or is this gtg?
There was a problem hiding this comment.
not urgent, we should do a naming-cleanup-pass of the exported API before 1.0 and the unexported APIs after
There was a problem hiding this comment.
I think the renaming can be done separately.
This allows testing whether a variable is defined, and may be especially useful for emitting from a macro (which may want to conditionally initialize a variable).
I expect this will also be very useful when the behavior of dispatch is changed is changed to make static parameters conditionally undefined (#21026 (comment)).