Re: No completion candidates for a template case

On Mon, Oct 19, 2015 at 5:47 PM, Eric Ludlam <[email protected]> wrote:
> On 10/19/2015 02:00 PM, Liang Wang wrote:
>>>
>>> >Hi,
>>> >
>>> >I pushed an updated c parser for CEDET that should enable this case to
>>> > work.
>>> >Thanks for the nice simplified example.
>>> >
>>
>> Thanks for the fix. I used cedet/semantic within emacs before, but it
>> seems that fix is only in cedet git repo.
>>
>> When I tried to integrate cedet into emacs, I got the following error
>> message during startup.
>>
>>    require: Autoloading failed to define function debug.
>>
>> I'm using emacs trunk and will try to debug this problem later.
>
>
> Did you use cedet-devel-load.el as the mechanism for pulling in CEDET from
> the git repo?

No, I didn't. By loading this file first, cedet starts successfully.
But I need the following patch to build it. Otherwise, make-string
complains that the first argument is a negative number.

>
> Alternately, are you using emacs 25?

Yes.

> The git repo hasn't caught up to Emacs 25 changes to eieio yet.

I got error when calling semantic-analyze-current-context, like
  Symbol's function definition is void: eieio--object-class

I don't find eieio source in cedet git repo. Could you tell me how I
can get a version of eieio for current cedet git repo?

>
> Eric
>

diff --git a/cedet-build.el b/cedet-build.el
index 2c60b3e..06cb3d6 100644
--- a/cedet-build.el
+++ b/cedet-build.el
@@ -227,7 +227,7 @@ OVERRIDE-CHECK to override cedet short-cicuit."

       (let ((ns (object-name-string targ)))
  (cedet-build-msg "   Target %s...%s" ns
- (make-string (- 20 (length ns)) ? )))
+ (make-string (- 40 (length ns)) ? )))

       ;; If it is an elisp target, then do that work here.
       (let ((ans (save-excursion

------------------------------------------------------------------------------