Re: rnrs records: shadows previous definition of `dummy-1a78708d3c9406a3'

Thanks for the patch Tomas.  I will try to review/merge it this week.

On Sun, 24 May 2026, Tomas Volf <[email protected]> wrote:
> Hello,
>
> Am=C3=A9lie Devos <[email protected]> writes:
>
>> On 23/05/2026 3:21, Olivier Dion wrote:
>>> I have no idea why `dummy' is necessary here.  I'm sure someone that
>>
>> To keep things in a definition context, or whatever the technical term i=
s, I'd
>> assume.
>>
>> Going only by the code mentioned in the mail, It seems eliminatable
>> though.  [..]
>
> I took the liberty to implement your suggestion and attach it to this
> email.  Seems to work fine and no warnings are produced.
>
> This type of change hopefully is fine even for patch release, since
> relying on dummy-1a78708d3c9406a3 being defined by define-record-type
> has to be in the "not supported" territory, right?
>
> From 9db5c76701cecab240bf0eee0fea384b8e80c67d Mon Sep 17 00:00:00 2001
> From: =3D?UTF-8?q?Am=3DC3=3DA9lie=3D20Devos?=3D <[email protected]>
> Date: Sun, 24 May 2026 16:12:54 +0000
> Subject: [PATCH] Adjust R6RS `define-record-type' to suppress shadowing
>  warning.
>
> Given
>
>     (use-modules (rnrs records syntactic))
>
>     (define-record-type a)
>     (define-record-type b)
>
> shadowing warning was produced
>
>     $ guild compile -Wshadowed-toplevel -o test.go test.scm
>     test.scm:4:0: warning: shadows previous definition of `dummy-1a78708d=
3c9406a3' at test.scm:3:0
>     wrote `test.go'
>
> Since the `dummy' binding is not used by anything, solution is to not just
> produce it.
> ---
>  module/rnrs/records/syntactic.scm | 9 ++++-----
>  1 file changed, 4 insertions(+), 5 deletions(-)
>
> diff --git a/module/rnrs/records/syntactic.scm b/module/rnrs/records/synt=
actic.scm
> index bde6f9348..a4c7c1f29 100644
> --- a/module/rnrs/records/syntactic.scm
> +++ b/module/rnrs/records/syntactic.scm
> @@ -173,14 +173,13 @@
>                        (record-constructor
>                         (make-record-constructor-descriptor
>                          record-name #,parent-cd #,protocol)))
> -                    (define dummy
> +                    (define predicate-name
>                        (let ()
> -                        (register-record-type=20
> +                        (register-record-type
>                           (quote record-name)
> -                         record-name (make-record-constructor-descriptor=
=20
> +                         record-name (make-record-constructor-descriptor
>                                        record-name #,parent-cd #,protocol=
))
> -                        'dummy))
> -                    (define predicate-name (record-predicate record-name=
))
> +                        (record-predicate record-name)))
>                      #,@field-accessors
>                      #,@field-mutators))]
>               [((fields record-fields ...) . rest)
> --=20
> 2.54.0
>
>
> --=20
> There are only two hard things in Computer Science:
> cache invalidation, naming things and off-by-one errors.
--=20
Olivier Dion