lexical-binding breaks decl-scan #25
Comments
Hi-Angel
added a commit
to Hi-Angel/purescript-mode
that referenced
this issue
Nov 5, 2024
The function has implemented a poor man's hash table by manually enlisting keys and then jumping through the hoops by fetching them from one place, converting values to symbols to values… In particular, it was using `symbol-value` to map a symbol to its value, however the symbol was a local variable, and `symbol-value` doesn't work with them under lexical-binding, which the file was converted to since commit 9a9f550. So fix the regression and simplify the code at the same time. Fixes: purescript-emacs#25
|
Thank you for the report! Please, in the future when you notice someone introduced a regression (disregarding the project), please CC the author, so they'd be able to get involved. I fixed the bug in a PR, it's waiting for review. I'm not sure how long it's gonna take though because @kritzcreek seems to have disappeared, and from what I understand there isn't other maintainers… In the meanwhile, you can apply the changes locally if you want by copying from the raw version. I tested the output of the function before and after the change on one PS file and the hashsums match, the output didn't change. So presumably everything should be okay. |
Hi-Angel
added a commit
to Hi-Angel/purescript-mode
that referenced
this issue
Nov 11, 2024
The function has implemented a poor man's hash table by manually enlisting keys and then jumping through the hoops by fetching them from one place, converting values to symbols to values… In particular, it was using `symbol-value` to map a symbol to its value, however the symbol was a local variable, and `symbol-value` doesn't work with them under lexical-binding, which the file was converted to since commit 9a9f550. So fix the regression and simplify the code at the same time. Fixes: purescript-emacs#25
Hi-Angel
added a commit
to Hi-Angel/purescript-mode
that referenced
this issue
Nov 11, 2024
The function has implemented a poor man's hash table by manually enlisting keys and then jumping through the hoops by fetching them from one place, converting values to symbols to values… In particular, it was using `symbol-value` to map a symbol to its value, however the symbol was a local variable, and `symbol-value` doesn't work with them under lexical-binding, which the file was converted to since commit 9a9f550. So fix the regression and simplify the code at the same time. Fixes: purescript-emacs#25
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hemlo!
Since 9a9f550 the function
purescript-ds-create-imenu-indexthrows an error:With some creative insertion of
(message ..)I managed to find the guilty expression:(symbol-value sym)in line 497 (https://github.com/purescript-emacs/purescript-mode/blob/master/purescript-decl-scan.el#L497).I am still quite a elisp-noob, so I cannot explain why the
index-imp-alistvariable is void, because I see it initialized in the(let* ..)in line 463ff.When I remove the
lexical-binding: tin the first line of the file, the error disappears and my helm/imenu declaration list is properly populated.Emacs Versions tested: 26.X, 29.4, 31.0
The text was updated successfully, but these errors were encountered: