Closed palkan closed 2 years ago
And then we should re-open https://github.com/whitequark/parser/issues/845, right? I guess correct fix is to change p_kwnorest
rule to be reduced into [val[0], val[1]]
and move s(:match_nil_pattern)
construction to the place that needs it. From what I understand low-level p_kwnorest
derivation constructs a node that is too high-level.
But I think it's ok to merge this PR and re-open a tracking issue.
Also, enable testing for 3.2 (that's why we haven't caught this yet).
my bad 👀
And then we should re-open https://github.com/whitequark/parser/issues/845, right?
Yep.
I think it's ok to merge this PR and re-open a tracking issue.
Let's wait a bit; I will try to provide a proper fix (as you suggested).
@iliabylich Done!
@palkan Thank you
Also, enable testing for 3.2 (that's why we haven't caught this yet).
p_kwnorest
is resolved intos(:match_nil_pattern)
, which we can not use inf_no_kwarg
.Ref https://github.com/whitequark/parser/commit/820cc04b8d176c6f46ccc9a010c55f6113f4b2a