CEDET

Arun writes:
> Cool . Completion works thus , which is great. Just one more question . Is
> there any way to implement javascript jump . According to me most of the
> javascript code will be either a JSON or a prototype inheritence. Currently I
> have ctags installed and jump works on php. Do you know how to add a new
> language to ctags . I can write some regexp but I just need to know where to
> start. 

See for example

http://weblogs.asp.net/george_v_reilly/archive/2009/03/24/exuberant-ctags-and-javascript.aspx

-David

On 06/12/2012 04:18 PM, D.Y Feng wrote:
> I'm using the cmake builder,and I want to use the Generic project.
>
> M-x ede-enable-generic-projects
> M-x ede-customize-project
>
>
> I get the error:
>>
>> eieio-oref: Wrong type argument: (or eieio-object-p class-p), nil
>
>
> and I found the ede-generic-new-autoloader in ede-generic.el
>
>>   (ede-generic-new-autoloader "generic-cmake" "CMake"
>>       "CMakeLists" 'ede-generic-cmake-project)
>
>
> Is "CMakeLists" means the file name?But in cmake system,it should be
> "CMakeLists.txt".I get the same error after I changed "CMakeLists" to
> "CMakeLists.txt".
>
> I have been tried the ede-cmake in
> https://code.launchpad.net/~arankine/+junk/ede-cmake.When I M-x
> ede-new,I get the unbound slot error.I have sent several e-mails to
> the author.however,no reply.

Hi,

Your error just means that there was no project type discovered for your 
file / directory.  It seems likely that is related to the CMakeLists 
file name problem you mention.

EDE's project detector is quite slow, so a cache went in several years 
ago to improve the performance, but it can get in your when when you 
develop Emacs projects, and just assume your change will take effect.

Thus, if you visit a file, and discover no project there, and then 
change some ede code, and try again, it will still miss.  You need to 
call ede-flush-directory-hash to fix it.  Once you do that, I'll guess 
your change to look for the .txt version will start to work.

Eric

On 06/12/2012 12:14 PM, Pierre Lorenzon wrote:
> Hi Eric,
>
> Here is the little hack I found to let ede work with emacs
> 24.1.50 I first replaced the file ede.el embedded in emacs with
> that provided by cedet-1.1. Up to very small modifications (in
> require and autoload calls) it works in particular with the
> embedded eieio implementation. But it does not solve the
> problem. Hence I replaced the file ede/files.el by that
> provided by cedet-1.1 (named ede-files.el). Up to renaming and
> suitably changing the provide directive at the end of the file,
> it works and it solves the problem. I cannot precisely locate
> the function in which the problem lies but I am almost sure
> that it is in files.el. I am impatient now that a good merge
> allows us to use ede fomfortably without this hack but I now
> that may take time before things are enough tested to be
> securely merged !
>
> Regards
>
> Pierre
Hi Pierre,

I'm glad you were able to verify a fix, and I can understand your 
impatience.  I'm not sure when such a merge might happen though.

Thanks!
Eric

On 06/14/2012 11:39 PM, David Ventimiglia wrote:
> Hi Eric,
>
> This is great!  Thanks!  I tried it, and it seems to work as expected.
> I used it to define in my .emacs file a new java-root project with a
> name, a project file (just a README), a srcroot, and a classpath.   It
> seemed to work, insofar as all of the Java imports (i.e., "includes")
> were identified, and I could navigate to the ones that are file-based.
> semantic-ia-complete-symbol still doesn't work on Java local variables,
> but I didn't expect this EDE project to fix that.  A couple of
> additional comments:
>
> * Yes, this is very useful.
> * Could this be abstracted into a base class that ede/java-root and
> ede/cpp-root could extend from?

Sure.  I generally like to make sure I have fully understood my classes 
before extracting a baseclass.  I think that will be a next step here.

> * A single srcroot won't be sufficient for many Java projects.
> Alternatively, perhaps it could take a list, like classpath does.

I contemplated that, and doing so will be an easy fix.  I didn't know if 
that was at all common.

> * srcroot resolves relative to the EDE project root, but I think I had
> to put in fully-qualified absolute paths into the classpath list.  It
> would be nice if the elements of the classpath list were resolved
> relative to the EDE project root.

The classpath is unmodified from the value passed in by EDE before 
passing it on.  Can you make a recommendation on the items that should 
be returned based on inputs?  I could make it similar to the include 
path in cpp-root.

Eric

On 06/12/2012 06:28 AM, Arne Schmitz wrote:
> I have a problem with semantic-analyze-proto-impl-toggle. Consider this very simple example:
>
> hello.h:
> ==============
> class Hello
> {
> public:
>     void sayHi();
> };
> ==============
>
> hello.cc:
> ==============
> #include<iostream>
>
> void Hello::sayHi()
> {
>     std::cout<<  "Hi"<<  std::endl;
> }
> ==============
>
> semantic-analyze-proto-impl-toggle will jump from the declaration to the definition with no problem. But the other way around, I only get an error "semantic-analyze-proto-impl-toggle: Could not find suitable prototype". Why?


Hi,

I inserted:

#include "hello.h"

into hello.cpp and this would work for me.

You might ask... why?

Well, the method you are trying to match is in "Hello", and if your 
local scope cannot find the definition of "Hello" (from Hello.h which I 
needed to add an include for) then the local scope gives up, and claims 
that sayHi has no parent.   If it has no parent, it now assumes it is a 
tag not in a class, and the basic tag search misses sayHi in the header 
because it is at the wrong level.

Does that make sense?

Anyway, I included a short patch below that makes the 'simple' tag 
hunter do a deep search, but you will likely be disappointed, since it 
will be quite confused if you have multiple classes with the same method 
name in it.  As such, I won't be checking that in.

Enjoy
Eric

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

=== modified file 'lisp/cedet/semantic/analyze/refs.el'
*** lisp/cedet/semantic/analyze/refs.el	2012-04-21 11:39:55 +0000
--- lisp/cedet/semantic/analyze/refs.el	2012-06-16 03:04:30 +0000
***************
*** 279,285 ****
     (let* ((name (semantic-tag-name tag))
   	 (brute (semanticdb-find-tags-collector
   		 (lambda (table tags)
! 		   (semanticdb-find-tags-by-name-method table name tags)
   		   )
   		 nil ;; This may need to be the entire project??
   		 nil t))
--- 279,285 ----
     (let* ((name (semantic-tag-name tag))
   	 (brute (semanticdb-find-tags-collector
   		 (lambda (table tags)
! 		   (semanticdb-deep-find-tags-by-name-method table name tags)
   		   )
   		 nil ;; This may need to be the entire project??
   		 nil t))

On 06/08/2012 03:22 AM, Fredrik Malmros wrote:
> Hi,
>
> There seems to be an issue with visibility when completing members with
> default visibility i.e. members or classes without access modifier in Java.
> When calling semantic-analyze-possible-completion with point at ^ in the
> class below, I get:
> void baz(), but I expect to get both void baz() and void bar() since bar
> is visible in Foo (and all other classes in the same package as Foo).
>
> public class Foo {
>       void bar () {}
>
>       private void baz () {
> b
>            ^
>      }
> }

Hi,

Sorry for the late reply.

The java utilities return the protection of 'bar' as being 'nil', and 
'baz' as being 'private'.  I think all that is needed is for java to 
override semantic-tag-protection to return private if nothing was 
specified.  (If private is the right thing.  I'm not that familiar with 
java.

This patch should solve the problem if 'private is the right answer. 
Please let me know if unspecified protection means private.

Thanks
Eric

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

=== modified file 'lisp/cedet/semantic/java.el'
*** lisp/cedet/semantic/java.el	2011-11-07 20:57:12 +0000
--- lisp/cedet/semantic/java.el	2012-06-16 02:14:07 +0000
***************
*** 173,178 ****
--- 173,187 ----
             (semantic-find-tags-by-class
              'type (semantic-find-tag-by-overlay point))))

+ ;; Tag Protection
+ ;;
+ (define-mode-local-override semantic-tag-protection
+   java-mode (tag &optional parent)
+   "Return the protection of TAG in PARENT.
+ Override function for `semantic-tag-protection'."
+   (let ((prot (semantic-tag-protection-default tag parent)))
+     (or prot 'private)))
+
   ;; Prototype handler
   ;;
   (defun semantic-java-prototype-function (tag &optional parent color)