Just for transparency, (as I think it is useful to know what modifications are being made to adt.jar if you are going to just use a modified adt.jar without knowing what mods are being done)...
If I were to guess, I'd image Anton is taking the adt.jar, unzipping it and decompiling the Java .class file "com.adobe.air.ApplicationDescriptor"
This class file contains a hard-coded array of strings in source:
public static final List ALL_SUPPORTED_LANGUAGES = Arrays.asList(new String[] { "cs", "de", "en", "es", "fr", "it", "ja", "ko", "nl", "pl", "pt", "ru", "sv", "tr", "zh" });Modify this list and recompile the class as needed with its dependencies, and inject this modified compiled .class into a new adt.jar
I personally think this is poor validation design by Adobe to have all these rules and strings hard coded into compiled Java classes when there are perfectly good XML Schema .xsd files in the SDK folder under: templates/air/Descriptor.*.xsd that would perform all the same validations without the heavy handed maintenance of Java coding by Adobe, and then also give us developers the ability to fix or modify the validation rules in the .xsds to pass the validation process and adjust valid elements of the app descriptor without waiting for a new SDK build as features like image sizes or valid languages are introduced on the platform OS.
