Simon Wright <[email protected]> writes:

> Working at revision 78668d09a4e4fa2361c5c513b91586e98f77a49c, the tests fail 
> with
>
> --- ../../test/ada_mode-conditional_expressions.adb   2016-06-20 
> 18:10:21.000000000 +0100
> +++ ada_mode-conditional_expressions.adb.tmp  2016-06-20 18:23:47.000000000 
> +0100
> @@ -75,7 +75,7 @@
>           when 0  => 41,
>           when 1  =>
>             (if J > 42
> -              -- comment indent matching GNAT style check
> +           -- comment indent matching GNAT style check
>              then 44
>              else 45));
>  end Ada_Mode.Conditional_Expressions;
>
> If I put a blank line before the comment the comment remains aligned
> in a neat-looking place (rather than with the '(' of '(if').

The point is to test ada-indent-comment-gnat: t.

The original position is where ada-mode puts it with
ada-indent-comment-gnat: nil, but that's not what should be committed.
So I didn't do the commit right when I implemented
ada-indent-comment-gnat. Sigh.

>
>
> I also was puzzled by this at line 1331 of ada-wisi.el:
>
>          (t
>           (or
>            prev-indent
>            next-indent
>            (floor indent ada-indent)))

Yes, I recently changed that to:

           (t
            ;; prev-indent and next-indent are both set here;
            ;; could add more checks to decide which one to use.
             prev-indent)

changing this to next-indent changes the above to align with "then";
that's an example the kind of "more checks" that could be added.

-- 
-- Stephe

_______________________________________________
Emacs-ada-mode mailing list
[email protected]
http://host114.hostmonster.com/mailman/listinfo/emacs-ada-mode_stephe-leake.org

Reply via email to