Where is localization done in the Eclipse code?

Question

I have a contributed view in an Eclipse RCP application. This view has a localized name %view.name, with the translation on a plugin.properties file to "My View Part".

I am looking into the source code of Eclipse to understand how the internationalization works. So far, I haven't been able to find at what point the value from the properties file is read and assigned as the name of the view.

If I start from the resource reading part, I see that the class ManifestLocalization reads the properties file, but I can't find where it is used.

If I start from the name assignment, I see that the TableReader class gets the name of the view part, but it is already localized. I also saw that Eclipse converts this plugin.xml into some serialized object at some point. Maybe it is done for performance reasons?

So my question is, at what time is this "%view.name" string converted into "My View Part"

Answer 1

There is at least two mechanisms supporting localization in eclipse: the NLS, and the org.eclipse.core.runtime.registry.IConfigurationElement. The part you are interested in occurs in the IConfigurationElement.

When the IConfigurationElement.getAttribute(String key) is called, the ResourceTranslator.getResourceString() method uses the Bundle-Localization header in the OSGi manifest to find the correct properties file (usually this is plugin${locale}.properties) and entry in the properties file.

You can get what is actually in the plugin.xml file by calling the getAttributeAsIs(String key) method instead.