Sunday, March 29, 2009

What a great EclipseCon!

As anyone on this planet knows, last week was EclipseCon time. As always, it was a great experience. Seriously... How many conferences can one discuss, during lunch, the double-checked locking pattern or how EMF could support facets? (thanks Eike and Kenn)

All talks I've attended were awesome. It's reinvigorating to see all the energy and technology happening behind Eclipse: from known technologies to hot topics like E4 and XText.

And it gets better! The people at the conference are simply fantastic. Just to name a few, it was great to chat with Nick, Chris, and Boris and to finally meet in person Tom, Kevin, Peter, and Paul Webster. There are many others that, although not listed here, have contributed to make last week a memory that will not fade away from my mind.

A bit more personal, I like to think that the EMF tutorial was well received. We chose to go deeper into several subjects instead of mechanically read all the slides. Unfortunately, because of that, we didn't cover all the material. The offer to be available to answer individual questions after the conference is still valid ;-)

I got some positive reviews for the modeled UI in e4 talk, which definitely doesn't seem to agree with the official numbers. I wonder if we were too "slide driven" while the folks attending were expecting to see more action. Or perhaps my delusional-self is correct and people did mistake the '+1' for the '-1' bucket. Oh well... I am sure we'll do better the next time.

Finally, a special thanks to Steve: I was peacefully seated at the closing event when he handed me a winning deck for Thursday's poker game. My son loved the RC boat and will enjoy playing with it on summer time.

Another finally... The EMF book was the first one to be sold out at EclipseCon! I had the pleasure to autograph a few copies. On the subject, EMF and modeling in general are everywhere. It is good to see something I helped built making others more productive ;-)

Saturday, September 27, 2008

I am lazy!

It is annoying to feel compelled to start every post with "Wow... It has been a long time since I've posted something here". Of course one answer is "So don't". I'd rather think that some people would prefer to say "So post more". I am known for having an occasional delusional moment here and there, though ;-)

I believe one of the reasons that I don't update this blog as often as I wanted to, besides being insanely busy, is that, bottom line, I am a lazy person. Fortunately, a characteristic that would put me out of any work back 50 years ago, is actually something that should be rewarded on the software development world. Let me illustrate this with an example.

When I am heads down, working 15+ hours per day to develop an entire product or a complicated feature, I get extremely paranoid (way more than normal) about introducing regressions by entering a, or better, any new line of code. And testing everything all the time is too tiring (I started by saying that I am lazy, remember? ;-). So I usually choose to automate as many tests I can. Sure I have to invest some time putting together an automation strategy (usually based on JUnit). But after that, I just need to click a button and watch the JUnit progress bar go while I exercise the divine right of not doing anything for a few minutes - OK, I am lying: usually I am working on another portion of the code while the tests are running.

On my last endeavor, for example, testing the code consisted on executing 11 Java examples that invoke an API written to identify security issues on web applications by looking at their code and by interacting with them as a hacker would do. Each one of these Java applications run for about 2 to 10 minutes and produces a long log that has to be verified. After doing this by hand 3 times, my laziness kicked in and forced me to find an easier way to test my code. After one or two hours of JUnit fun, I had a class that, for a test method like the one below, executes the XYZ.class on a separate VM, collects the generated log, and compares it with a baseline I know is correct.

@Test
public void codeCoverageExample() throws Exception {
runExampleAndCompareOutput(XYZ.class);
}

The comparison part was a bit tricky to implement because the logs have things like timestamps and temporary file paths that change from execution to execution. With some creativity and several regular expressions, I managed to circumvent all the issues and, like a famous black-turtle-neck-guy says, "BOOM" : the annoying tests were out of my way ;-)

I will write at least one more post pointing out cases in which being lazy was extremely valuable. I am already thinking about EMF related examples...

Actually, on a minor digression, is my passion for EMF the ultimate proof that I am lazy? After all, it does generate a bunch of the code that I would need to code by hand ;-)

Wednesday, June 18, 2008

"Ch-ch-ch-ch-Changes" and "I am still around!"

I couldn't really decide what would be the best title for this post. So I chose to use both ideas that I had in mind. Fortunately, as this title, there are several opportunities in which we can actually choose all the options that are presented to us. For example, three years ago, I've chosen to be both Brazilian and Canadian. More recently, I've also chosen to change a bit my professional life and do some product work while keeping my duties as an EMF committer.

After four years dedicating myself exclusively to EMF, I am now working on the Rational AppScan Developer Edition team (our beta is available here). A minor digression, the security-related technology behind this product is quite amazing. I really don't want to sound like a salesman, though. I know that it wouldn't go really well given my "amazing" commercial skills ;-)

The good news is that both Ed and my new manager (Jeff Turnham) have been extremely supportive in my decision of work both internally and on open source. Hopefully most of you agree with their attitude. I certainly do ;-)

To me it will be a bliss to employ the technology that I help build. Although the EMF team has always eaten its own dog food caviar ;-), it is quite different to face the problems that consumers of our code have to deal with. On the other hand, I am really expecting that the product experience will make me a better EMF developer. EMF has been part of my life for quite a while and it certainly deserves the best I can be.

Thursday, March 20, 2008

Platform scheme URI

Wow... It has been more than six months since the last time I took a few minutes to write something here. Things have been a bit crazy lately. A lot has kept me occupied, like, for example, writing the 2nd version of the EMF book (which is "already" available as a rough cut), working on a few bugzillas, and supporting the EMF community. On a more personal note, in October we moved to our new house right when our son was entering the "terrible twos" days. Actually, since then, my wife and I have been working as maniacs to finish the basement and prepare the kids' rooms for the arrival of the new member of the Paternostro family, who is supposed to be here by the end of April.

But I am sure you are not reading this to hear about what I've been doing ;-) So, let's move on to the subject of this post. Quite frequently people ask me how to "locate" files in Eclipse. In 80% of the cases, the conversation is about things like IFile and IWorkspace. Sometimes it digresses to files available inside bundles and, rarely, it involves the "state location" of a bundle (see the Plugin.getStateLocation() javadoc for more details). Obviously you can use IResources or java.io.File to work with files in these places. There is another way, though: platform URI. Although this scheme has been around since the beginning of time, I decided to ramble about it here because I've never seem its uses described in one single place.

There are a few ways to work with the "platform" scheme:

platform:/resourceIt is used to identify a resource located in the workspace. The next path segment after "resource" should be the name of a project, which can be followed by the folder and/or file we want to locate.
platform:/pluginIt is used to locate a resource available in a plug-in (I know, I know, bundle). One really cool thing about this one is that it doesn't really matter if this resource is available in a directory or in a jar file. It also doesn't matter if the bundle is installed in a link folder or in the default directory.

The path segment after "plugin" should be the identifier of the bundle, which can be followed by the path of the resource in the bundle.
platform:/fragmentThis one is quite similar to "platform:/plugin", being to used to locate fragment resources instead of bundle resources. As you are probably guessing, the segment after "fragment" should be the fragment's identifier.
platform:/metaWe can use this to access a bundle's stage location. The path segment after "meta" should be the bundle's identifier, followed by the path of the resource we want to refer to.
platform:/configThe "config" segment causes the platform URI to refer to the configuration area of the running Eclipse (usually the eclipse/configuration directory). This can be useful to read the config.ini file, for example.
platform:/baseThis always refers to the directory of the Eclipse being executed.

It is interesting to note that, for example, platform:/base/plugins/org.eclipse.emf/plugin.xml and platform:/plugin/org.eclipse.emf/plugin.xml don't necessarily refer to the same resource. The former is a "pointer" to a plugin.xml file located in a directory plugins/org.eclipse.emf under the directory that Eclipse is installed. The latter points to the plugin.xml of the "org.eclipse.emf.ecore" bundle regardless of where it is installed and whether it is jarred or not.

For the URI-savvy people, you should see "resource", "plugin", "fragment", "meta", "config", and "base" as authorities. Perhaps they could become authorities in e4 (yey! Now this is an e4 related post ;-)

Since we all like actual code...
IProject project = 
ResourcesPlugin.getWorkspace().getRoot().getProject("myproject");
if (!project.exists())
{
project.create(new NullProgressMonitor());
}

System.out.println("\n==== platform:/resource ====");
{
URI uri = URI.createPlatformResourceURI("myproject", true);
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));

uri = uri.appendSegments(new String[]{"folder", "file.txt"});
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));
}

System.out.println("\n==== platform:/plugin ====");
{
// Just for fun, choose a bundle that is not in the default location
URI uri = URI.createPlatformPluginURI("org.eclipse.emf.ecore", true);
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));

uri = uri.appendSegments(new String[]{"model", "Ecore.ecore"});
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));

// Choose a bundle that is in the default location (<eclipse-dir>/plugins)
uri = URI.createPlatformPluginURI("org.eclipse.core.resources", true);
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));

uri = uri.appendSegments(new String[]{"META-INF", "MANIFEST.MF"});
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));
}

System.out.println("\n==== platform:/fragment ====");
{
URI uri = URI.createURI("platform:/fragment/org.eclipse.swt");
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));

uri = uri.appendSegments(new String[]{"META-INF", "MANIFEST.MF"});
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));
}

System.out.println("\n==== platform:/config ====");
{
URI uri = URI.createURI("platform:/config/");
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));
}

System.out.println("\n==== platform:/base ====");
{
URI uri = URI.createURI("platform:/base/");
System.out.println(uri);
System.out.println(CommonPlugin.resolve(uri));
}
The call to CommonPlugin.resolve(URI) returns a URI that uses a protocol which is native to the Java class library (file, jar, http, etc).

I will leave it to you to run this code and see the results :-P Don't forget that it must be executed in an Eclipse shell. Probably the simplest way to do so is to paste these lines into a JUnit test located in a bundle and execute it as a "JUnit Plug-in Test".

So what can we do with platform URIs? For one, read the contents of the resources pointed by them. We may also be able to write to such resources or even delete or create them. A tip for EMF 2.4 users: URIConverter.INSTANCE allows easy access to methods that are extremely handy when dealing with URIs (createOutputStream(URI), createInputStream(URI), delete(URI), and exists(URI)).

Personally I consider URIs a good fit for APIs that would normally use "plain" paths. Take as an example the icon attribute of the extension point below. Because its value is handled as a URI, we are allowed to refer to an image located in a different bundle.
<extension point="org.eclipse.ui.editorActions">
<editorContribution ...>
<action
icon="platform:/plugin/com.myplugin/icons/me.gif"
...
/>
</editorContribution>
</extension>
BTW, if I were to implement the code to process such an extension point, I would probably do something like this:
IConfigurationElement configurationElement = ...
URI iconURI = URI.createURI(configurationElement.getAttribute("icon"));
if (iconURI.isRelative())
{
URI pluginURI =
URI.createPlatformPluginURI(
configurationElement.getContributor().getName() + "/", true);
iconURI = iconURI.resolve(pluginURI);
}

try
{
ImageDescriptor imageDescriptor =
ImageDescriptor.createFromURL(new URL(iconURI.toString()));
descriptorImpl.setIcon(imageDescriptor.createImage());
}
catch (Exception e)
{
e.printStackTrace();
}
This code assumes that when the value of the icon attribute is a relative URI (like icon/me.gif for example), the developer is indicating that the image is contained in the bundle that uses the extension point.

For obvious reasons, I wrote these examples using EMF APIs, including our URI class. It shouldn't be terribly difficult to rewrite the code in this post to use basic Eclipse and Java code (like, java.net.URI).

Friday, August 31, 2007

Almost there, mom...

Yet another month has gone and my perfect plan to blog about something cool in EMF every week didn't work as intended. If you check the date of my last post, you will see that it was written right on the last day of July. And here I am, last day of August...

Being honest, I can't even say that I've been busy doing development work. Other than a few bug fixes and time assisting users, most of my (and Dave's) days have been spent on the second edition of the EMF book. The good news is that we are really almost done writing the manuscript. And the material is looking great! We took the time to go deep on several topics and also to add much more content than what is available in the first edition. Just to give you an idea, the second edition will be at least as thick as the first one and it doesn't have the not so popular "javadoc" chapter.

Btw, if you were intrigued by the title... I was naive enough to tell my family and close friends about the book when we've started working on it, a looooong time ago. So now I am often hearing the inevitable question "when will it be available" regardless of where I am: at work, conferences, and, unfortunately ;-), BBQs, bars, ...

Tuesday, July 31, 2007

No more NPE when generating models

For most users, EMF 2.3 is all about Java 5.0. Although most of our time was certainly spent on Generics and other "5isms", we sure managed to implement a few nice features as listed in our release notes. BTW, I know that this report is not as beautiful as a "New and Noteworthy" page, but that's the best we can offer now. Note to Wassim: I promise we will try to improve for this release ;-)

Of these nice features, the one I think I like the most is the complete validation of Ecore models. Basically this work consisted on defining and implementing all the constraints that characterize a well-formed model using the EMF Validation framework. Without an automated way to check these constraints, aka "before 2.3", any poor soul manipulating Ecore models would probably end up getting all sorts of exceptions when generating the code for the models. In face of these exceptions, the users would do the simplest thing available at the time: post a question on the newsgroup and wait the excruciating 10 minutes delay that usually takes Ed to answer.

Now this has changed! To check if your model is correct, hence capable of being used for code generation, all you need to do is to either open the .genmodel file editor or, on the .ecore file editor, select a package and run the Validation action available in the context popup-menu. Any existing problem in your model will be then presented in a tree viewer, with (hopefully ;-) all the information you need to have it fixed.

I should mention that EMF 2.3 can validate any Ecore model. So even if you are using an older version of EMF and usually edit your Ecore models directly, I believe the model validation justifies having the 2.3 release installed somewhere in your machine.

Thursday, June 28, 2007

I want my feature request implemented yesterday!

"What do you get when you put together a bunch of passionate developers and a code freeze policy due to the end of a release?" This is probably a good start for a Dilbert cartoon. Certainly Scott Adams' answer would be brazilian times funnier than mine, which is "crazy people implementing the cool things they want regardless of how beautiful the day is outside".

The "cool thing" I worked on yesterday was to leverage the pattern filter widget you see on Eclipse's New Project Wizard to some EMF dialogs. After some digging I was able to get the results I was looking for by simply using new FilteredTree(...).getViewer() instead of new TreeViewer(...). BTW, I would recommend reading the PatternFilter's Javadoc if you are considering doing the same in your code.

Everything was going smooth up to the moment that I've opened a subclass of org.eclipse.ui.dialogs.ElementTreeSelectionDialog. I couldn't believe what I was seeing: there was no way to replace the TreeViewer used by this class. Profoundly disappointed, I began to write a feature request for the UI team. As I was doing it, I thought that I would be a much better Eclipse citizen by providing a patch instead of simply describing what to do. So I checked out the "org.eclipse.ui.workbench" bundle, opened ElementTreeSelectionDialog, and added a new protected method that instantiates the TreeViewer. Before changing the line of code that would use my new method, I looked around to see how the Javadocs on the file are written. To my surprise I discovered this new method: ElementTreeSelectionDialog.doCreateTreeViewer(Composite, int). After blinking my eyes for a few seconds I realize that the method I need is already available in Eclipse 3.4. Amazing!

Kudos to the UI team.