> Well since unfortunately the encoding of a property file is set as ISO-8859-1 > in it spec, At least for JDK 11 the API says the opposite : https://docs.oracle.com/en/java/javase/11/docs/api/java.base/java/util/PropertyResourceBundle.html, see the API paragraph. And indeed it's what I experiment: if I use ResourceBundle.getBundle(Class, Module),getString(key), it does automatically read UTF-8 or ISO strings (I made a test with both encodings in a single.properties file). If I use NbBundle.getMessage() then it always expects ISO, that's why I don't use it. The problem is that I can't alter the behavior of the Netbeans internal stuff which still works the "old way". This impacts branding localization or action displayname declared via @ActionRegistration(lazy=true).
> I'd use a generator step in the build which would translate the UTF-8 > characters into \uxxxx of the final Bundle.properties. Yes, this is what I should try in the meantime... I'm not familiar with the And build-script, which target should I start with ? Le lundi 21 décembre 2020 à 16:14:33 UTC+1, Laszlo Kishalmi <laszlo.kisha...@gmail.com> a écrit : Well since unfortunately the encoding of a property file is set as ISO-8859-1 in it spec, I'd use a generator step in the build which would translate the UTF-8 characters into \uxxxx of the final Bundle.properties. On 12/21/20 6:56 AM, Jerome Lelasseux wrote: I use the crowdin.com platform (free for open-source!) to get the translations done for my Netbeans RCP application. It gets synchronized automatically to GitHub nicely. The problem is that I pull UFT-8 encoded Bundle*.properties, and the RCP platform expects ISO-8859-1 encoding, so special characters got messed up. Do you guys have experience how to integrate this seamlessly ? (ideally by programmatically alter the behavior of the RCP framework so that UFT-8 is used to read .properties) Thanks. Jerome