Closed jackfirth closed 3 years ago
This looks cool!
But it looks like there's a problem. I tried building with this patch, and I got need-result?: undefined; cannot use before initialization
when raco setup
tried to run the first document. I see that the revised code has (define need-result? (and need-result? (not here-result)))
, probably where the original code was intentionally shadowing need-result?
with let
.
Oops. There's a bit of logic to catch those cases with let
, but not let*
. I'll fix the tool and repush.
@mflatt I've fixed the bug in the tool that produced the buggy need-result?
definition and repushed the generated changes.
Thanks!
This pull request was generated by an automated tool, with some by-hand postprocessing to preserve comments. The tool replaces uses of the various
let
forms with equivalentdefine
forms, when doing so would reduce nesting and when it can prove that the binding structure remains unchanged. Parts of this change involved changing some code from usingif
to usingcond
, as well as other similar changes needed to introduce a definition context.The tool produces moderately well-formatted code, but it makes a conscious effort to avoid changing the formatting of any existing subforms that were not refactored.
Note: the tool is called
resyntax
, and is implemented here.