[m-rev.] for pre-review: id of only-mode procedures

[Julien Fischer]

On Thu, 9 Jan 2025 at 15:10, Zoltan Somogyi <zoltan.somogyi at runbox.com> wrote:
>
> I am working on a change on the wording of diagnostics
> about procedures that are the only procedure in their predicate
> or function. We current identify procedures in some error messages
> as "name(argmode1, ..., argmodeN)" or as "name(argmode1, ... argmodeN) =
> returnmode". In many cases, the arity is high enough that this description
> takes up more than one line. This verbiage is necessary if the predicate
> or function has some other procedures, but is unnecessary if it doesn't.
> In the latter case, my diff-in-progress replaces the ids of the forms above
> with just "pred name/N" or "func name/N".
>
> Technically, the affected diagnostics are about a procedure, while e.g.
> "pred name/N" names a predicate, so this is a category error, but adding
> "the only mode of" in front of "pred name/N", which would resolve
> this category error, would also be clutter to many readers (though maybe not all).
>
> My questions are: is the category error easy enough to see past that
> it is ok for the diagnostic's text to have it, or should I fix that category error?
> And if the latter, should this be done by adding the "the only mode of" prefix,
> or by some other means?

I think the shorter version (as in the attached file) is fine and you
don't need to add
any extra prefixes.

Julien.