DESCRIPTION

EGit operation / action refactoring

Replace IWorkspaceRunnable in EGit operations (e.g. BranchOperation)
by a new interface IEGitOperation.
IWorkspaceRunnable is no good interface for an operation. The user 
thinks he has to execute the operation using Workspace.run. But that 
is not right for all EGit operations. Currently many operations are 
used by just calling op.run. Thus a new interface with an execute
method was introduced. Those operations that need a workspace runnable
and a scheduling rule do the workspace runnable in their execute 
method (e.g. branch). It's an internal of an operation how it needs
to execute.
BranchOperation now uses a workspace runnable to lock the workspace
while the working tree is modified.
BranchAction executes BranchOperation in a Job. The other actions
will be adopted to this execution pattern in further commits.

Change-Id: Ia3c1bf157d17632a17b7c678ee45717fb7944f0c
Signed-off-by: Marlie Beasley xxx@xxx.xxx

COMMENTS

author: Marlie Beasley
date: 2010-04-28 16:23:26.000000000

Uploaded patch set 2.

-------------------------------------
author: Malia Sampson
date: 2010-04-28 16:35:28.000000000

Patch Set 2: No score; No score

Any reason we don't use the commands/handlers/operations stuff from the Eclipse workbench? That is, IUndoableOperation/AbstractOperation?

-------------------------------------
author: Jaylon Avila
date: 2010-04-28 16:50:05.000000000

Patch Set 2: No score

Jens, I noticed you didn't change the code in BranchOperation or respond to my original question. Was this an oversight?

-------------------------------------
author: Marlie Beasley
date: 2010-04-29 07:16:27.000000000

Patch Set 2: No score

Hi Remy,

did you see my comment on your comment? It was:

We do not know which projects will be affected by the branch operation. It's e.g. possible that changes in a project are performed that resides in an EGit repo but the project is not shared with EGit. Thus I think it's safer to lock the complete WS.

Best regards

Jens

-------------------------------------
author: Marlie Beasley
date: 2010-04-29 07:19:17.000000000

Patch Set 2:

Hi Chris,

I do not expect there will be undo functionality for team operations in the near future. Thus I kept it simple and defined a minimal interface to get rid of the not fitting IWorkspaceRunnable.

Best regards

Jens

-------------------------------------
author: Jaylon Avila
date: 2010-04-29 09:44:28.000000000

Patch Set 2:

Jens, when I expand the first patch set I only see "1 comment" next to BranchOperation. Do you see "2 comments"? Maybe there is a bug in Gerrit?

-------------------------------------
author: Jaylon Avila
date: 2010-04-29 11:50:49.000000000

Patch Set 2:

I see the inlined comments now on the first patch now, Jens, thanks.

I think I may have some other concerns about this whole locking thing but let's get this out of the way first before we do more polishing. :)

-------------------------------------
author: Willow Cantrell
date: 2010-05-04 08:15:21.000000000

Patch Set 2: Looks good to me, but someone else must approve; IP review completed



-------------------------------------
author: Malia Sampson
date: 2010-05-04 14:49:06.000000000

Patch Set 2:

Can you resubmit a patch and I'll merge.

And we will work from there :)

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