Re: [cdt-dev] Re: [cdt-patch] FW: c/c++ refactoring contribution

[John Camelon <jcamelon@xxxxxxxxxx>]

I concur that this is a great start
Markus.  
I look forward to helping you tune this in M6 & M7 ... in my testing
I've encountered some UI spinning away mercilessly on files where there
are lots of inclusions. 
Once you get back we should set up a
call/netmeeting to discuss where we go next.

Cheers,
JohnC
www.eclipse.org/cdt

cdt-dev-admin@xxxxxxxxxxx wrote on 03/09/2005 05:00:39
PM:

> ***********************
> Warning: Your file, initial-refactoring-contrib.zip, contains more
> than 32 files after decompression and cannot be scanned.
> ***********************
> 
> 
> ***********************
> Warning: Your file, initial-refactoring-contrib.zip, contains more
> than 32 files after decompression and cannot be scanned.
> ***********************
> 
> 
> Great work, Markus! This is definitely going in the direction I was
hoping 
> to see, i.e., maximize reuse of the LTK and the CDT DOM.
> 
> I have committed this with one minor change (i.e. the removal of the
> assert since we haven't moved to Java 1.5 yet). It can be found in
the 
> org.eclipse.cdt-core folder in /home/tools. If anyone feels that this
is 
> the wrong place for it, please step forward and we can have that debate
> and a proper vote on it.
> 
> I have not added this to the cdt feature since the LTK still has the
> dependency on JDT. Once that is removed we can have a debate and vote
on 
> whether to add it in so all CDT users get it. From what I've seen
so far, 
> it is going to be a lot better than the current rename refactoring.
> 
> One thing to watch out for, Markus and everyone working on this feature,
> is scalability. An interesting exercise would be to try this out on
our 
> standard large project, i.e., Mozilla, and see what happens to memory
> consumption. I noticed that you are caching the ASTs during the 
> refactoring. This could get large if you find a lot of files that
match.
> 
> Cheers,
> Doug Schaefer
> Ottawa Lab, IBM Rational Software Division
> 
> 
> 
> "Schorn, Markus" <Markus.Schorn@xxxxxxxxxxxxx> 
> Sent by: cdt-patch-admin@xxxxxxxxxxx
> 03/09/2005 09:16 AM
> Please respond to
> cdt-patch
> 
> 
> To
> <cdt-patch@xxxxxxxxxxx>
> cc
> 
> Subject
> [cdt-patch] FW: c/c++ refactoring contribution
> 
> 
> 
> 
> 
> 
> ***********************
> Warning: Your file, initial-refactoring-contrib.zip, contains more
than 32 
> files after decompression and cannot be scanned.
> ***********************
> 
> 
> here is my initial contribution for c/c++ refactoring. You may want
to
> pass on some of the info to the community when putting the code into
> the CVS repository.
> This does not come as CVS patch, as this is not a change to an
> 
> existing plugin.
> 
> 
> Plugins:
> ========
>   + org.eclipse.cdt.refactoring
>      + src
>         + org.eclipse.cdt.internal.refactoring
>         + org.eclipse.cdt.internal.refactoring.scanner
>         + org.eclipse.cdt.internal.refactoring.ui
>         + org.eclipse.cdt.refactoring
>   + org.eclipse.cdt.refactoring.tests
>      + src
>         + org.eclipse.cdt.core.tests
> 
> 
>         + org.eclipse.cdt.refactoring.tests
> 
> As the cdt.refactoring plugin depends on ltk.ui.refactoring there
is
> an implicit dependency on jdt.core. See
> 
> https://bugs.eclipse.org/bugs/show_bug.cgi?id=61312
> 
> 
> Licensing:
> ==========
> The code is licensed under the EPL, if that is not apropriate I could
> change that.
> I use a header slightly different to the one used inside CDT. It is
the
> one suggested at
> http://www.eclipse.org/legal/copyrightandlicensenotice.html
> (and the one approved by my legal department).
> 
> 
> Test cases:
> ===========
> I have ported the existing test cases for renaming to the new
> 
> implementation and started to extend them with further tests.
> 
> There will be more to do for me.
> 
> 
> Status:
> =======
> The core of the rename-refactoring is implemented. Regression tests
that
> passed with the old version also pass with the new version. There
is one
> exception to that for which I'll file a bug today.
> 
> 
> The new version additionally fixes the bugs: 72612, 72717, 72605,
79231,
> 
> 79282, 79295, 72620, 72617.
> 
> It is possible to register renaming participants with the same extension
> point jdt is using. One of the regression tests shows how to do this.
> This could be of interest to the debugger (renaming a function with
a
> 
> breakpoint sitting on it?)
> 
> Undo/Redo is working by means of the refactoring framework. To check
it
> out you have to enable the 'Java Coding' Action Set in your perspective.
> 
> 
> Effects on the GUI:
> ===================
> The plugin adds a menu item to the context menu of the CDT-Editor.
> (Refactoring - Rename (work in progress)).
> 
> 
> What's next:
> ============
> Starting after Easter I plan to
> 
> * finish the checks for shadowing/redeclaring vars/funcs/types.
> * extend test cases.
> * handle macros.
> * extend test cases.
> * integrate new rename, undo and redo into the GUI.
> * remove old renaming facility.
> 
> Markus.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> [attachment "initial-refactoring-contrib.zip" deleted by
John 
> Camelon/Ottawa/IBM]