I am working on an Eclipse Java project which uses Pydev to interpret Python scripts. After the 6.3.0 version of Pydev, there has been significant changes on initializing plugin programatically and on Preference access objects. I couldn't find a relevant documentation for the new version.
For example, starting from v6.3.0 there is no PydevPlugin.setPythonInterpreterManager(IInterpreterManager im) method in the PydevPlugin class. However, I couldn't find an example or a document on initializing the interpreter manager with the new version. In earlier versions, the task is performed similar to the below 3 lines:
IInterpreterManager im = new PythonInterpreterManager(PydevPlugin.getDefault().getPreferenceStore());
PydevPlugin.getDefault().getPreferenceStore();
PydevPlugin.setPythonInterpreterManager(im);
Has anyone dealt with this issue?
The many changes related to the interpreter configuration have been done to better separate the core and ui related plugins.
You can see the tests (i.e.: org.python.pydev.ast.codecompletion.revisited.CodeCompletionTestsBase.setInterpreterManager(String)) as a reference in how to initialize it now (it now uses org.python.pydev.ast.interpreter_managers.InterpreterManagersAPI to set the interpreters).
Related
I am currently trying to modify the behavior of an existing open-source plug-in for Eclipse.
I'm trying to understand how the plugin works. To do so, I opened it in "Debug" mode as an Eclipse application and I am testing it out after having inserted many breakpoints.
However, sine Eclipse plugins do not have a Main method, it's still really difficult for me to keep track of everything that is going on. The calls seem to jump arbitrarily (which I quickly realized was happening through all the interfaces and superclasses the plugin is inheriting) and I can't see exactly what's doing what.
What is the proper (read: BEST) way to debug a program with no Main method? How can I test, tweak, and explore and program - in this case a plugin - whose modus operandi I'm uncertain of?
You need to run your plugin in a runtime workbench. This (simplistically) starts a new instance of Eclipse with all existing plugins installed, plus the plugin you want to debug. Make sure that you have the PDE tools installed in your Eclipse instance and then in the debug configurations area, double-click on Eclipse Application to generate a default runtime workbench launch config.
I'd also recommend that you read up on PDE (plugin development environment), and you can get an overview here: http://wiki.eclipse.org/PDE/FAQ. And you can read up on plugins in general here: http://eclipsepluginsite.com/. There are many tutorials and lots of information all over the web. So, google is your friend.
Along with using breakpoints in Debug method you can even try printing stack trace using Thread.currentThread().getStackTrace() method to know the starting point of a Thread.
I am new to sakai, i want to add my own tool in sakai source code like announcement,syllabus...etc That tool having some my own functionality for this what i have to do, how can i develop my own tool in sakai.
Thanks.
Use one of the archetypes to get you started, then customise it. When you are comfortable you can change the UI layer to be whatever you like.
https://confluence.sakaiproject.org/display/BOOT/Developer+Tools
If you are looking for a solid "base" tool shell to begin with tool development my recommendation would be to start here: "Sakai Wicket Maven Archetype". There are a few other very good archetypes available but are often outdated and may cause you some problems based on my experience. This has been the one I've had the most success with.
This uses Wicket but does not mean you have to use Wicket to develop further with it. Make sure to pull the 1.2 version as last check the binaries for the others were not available yet so replace -DarchetypeVersion=1.4.1 with -DarchetypeVersion=1.2 It will compile and deploy successfully as a tool "as-is" that you can modify as needed. It also allows you to code in Java to directly access the sakai API and pull from its database or even add your own tables. I've successfully used it with NetBeans 8.0.1 and MySQL 5.6 under Windows 7 with Maven 3.2.3 and Sakai 10.1.
The code is well-documented and is no-frills but with enough varied functional use examples to allow you to do "almost" anything you want. The wicket manual is 600+ pages but in my experience you should be able to accomplish a lot just by looking at the code in the archetype.
I apologise for the very amateur question about E4 but I am a bit confused about a couple of things about RCP development using the new Eclipse 4 framework. I read in this tutorial that one can no longer use any default commands the way we could in 3.x especially for common things like Save, Save As... in the File menu. It says that in E4 we have to write our own commands. The reason I am confused is because the thing I liked about Eclipse previously is that a lot of things are already implemented and we can just extend that to our own needs. But it feels like now everything must be written from scratch.
That led me to considering reusing the command code already written for Eclipse Juno. I had the Live Editor open so I could see the list of commands etc but I don't really see any Handlers implemented for any of them. Then I used the Spy on Eclipse Juno and checked out some of the menu items and they all seem to point to Actions. That really confused me as I thought Juno was based on E4.
I could be completely wrong so I am sorry for asking such a silly question. I only just started using E4 and need to decide whether one of fairly young Eclipse 3.x projects should be migrated to 4.x.
AfaIk, in Juno the compatibility layer translates the 3.x based implementation of the IDE into E4 concepts. This is also the reason why the reusable commands are not yet available.
If you have an 3.x based RCP it should also run using the compatibility layer (and you can reuse the commands mentioned in your question) if no incompatible APIs have been used.
Here is more information (also a tutorial from Lars Vogel): http://www.vogella.com/articles/Eclipse4CompatibilityLayer/article.html
Groovy seems to fix a lot of the things I dislike about Java, and I was wondering if it would be possible to actually write an Eclipse plugin in Groovy instead of Java.
Does anyone know if this is possible, and if so how to go about it?
I've just found a blog entry which says it's not officially supported but is actually possible.
Not yet tested to see if it works, but it seems promising:
Writing Eclipse plugins with Groovy, by Jörn Dinkla
#Peter, I do not think that the blog post you linked to is complete or if it will really work. It is pointing to the old version of Groovy-Eclipse, which is no longer supported and is out of date.
Yes. It is possible to create your own plugins in Groovy.
First, install the Groovy-Eclipse plugin from here:
http://dist.codehaus.org/groovy/distributions/greclipse/snapshot/e3.7/
Then you can create a new plugin project and add the Groovy Nature.
Remove the Groovy Libraries classpath container
Instead, add the org.codehause.groovy as a required bundle
Create your Groovy code as normal
Now, the tricky part is exporting the plugin using PDE. See this blog post for how to do that: http://contraptionsforprogramming.blogspot.com/2010/08/groovy-pde-redux.html
One important thing to note is that you will need at least one Java file in your project for PDE to compile anything, It can be a dummy, empty file (this is a bug that has not yet been fixed).
Rejoice!
As an example, here is the codenarc Eclipse plugin that was written completely in Groovy:
http://sourceforge.net/projects/codenarceclipse/
You can also use JRuby, or Javascript ...
JAM Circle is a great example showing how to make great use of a scripting language in an Eclipse plugin, by allowing the end user to write his own actions and load them at runtime.
There's a proxy-like plugin that allows you to implement the plugin virtually in any language that supports JSR223 (javax.scripting)
We are developing a RAP(Rich Ajax Platform) application using Helios 3.6 Version of Eclipse.Here, we are using Command Framework which uses Eclipse Modeling Frame Work.
When a command gets executed ,getMostAffectedObject() returns us the most recently affected Model Object that got modified due to the execution of command.
My question is "How's the most recently changed changed Model object identified by command framework " ?
Please help me regarding this.
Which model objects are affected is implemented in the command itself.
e.g. take a look at the following classes:
org.eclipse.emf.edit.command.AddCommand.doGetAffectedObjects()
org.eclipse.emf.edit.command.SetCommand.doGetAffectedObjects()