bug#81308: 32.0.50; The compilation buffer is over-eager about GCC 16 diagnostics

Arsen Arsenović [15/Aug 11:52am +02] wrote:
> This one *should* be correct, but it has one issue I've not been able to
> fix.

I compared your changes to compilation.txt with and without the fix, and
now the handling is (more) correct, so I've installed the patch.
Thanks!  Please take a quick look at my revised commit message, though.

> You'll see compilation.txt, it has the following example on L351.  The
> "candidate is" line is the relevant one.
>
> test.cc:8:6: error: no declaration matches ‘void foo::test(int, int, const void*, int)’
>     8 | void foo::test(int i, int j, const void *ptr, int k)
>       |      ^~~
>   • there is 1 candidate
>     • candidate is: ‘void foo::test(int, int, void*, int)’
>       test.cc:4:10:
>
> compilation-mode seems to, sometimes, interpret that latter location as:
>
>   #s(compilation--message
>      (10 4
>          (("test.cc" nil) nil (8 (42 8 #2 nil nil) (6 8 #2 nil nil))
>           (4 (35 4 #2 nil nil) #1) (1 (7 1 #2 nil nil)))
>          nil nil)
>      2 nil gnu)
>
> ... but not if I isolate it into its own file, or if I
> revert-buffer-quick, or if I just reopen the file, sometimes!
>
> I am not sure what's going on here.

Let's leave the bug open for now, then.

-- 
Sean Whitton