On Monday, 17 June 2019 14:18:33 PDT Mitch Curtis wrote: > The thing is that I have other data in the file that doesn't need to be > translated, so copying them would require keeping each file in sync. The > file itself is edited often fairly regularly to add new data (both > translatable and non-translatable).
Sounds like a design flaw. I suggest you split that up. A simple solution is to *not* have the translations in the .json files, but only US-ASCII English versions. Then you extract those strings into a proper .ts or .pot file, translate them using your preferred tool, and *keep* the translations there. At runtime, you translate using: // .toUtf8() would be strictly speaking more correct, but if you restrict // to US-ASCII, you gain a bit in performance tr(stringFromJson.toLatin1()); > I do like the idea of generating .ts files from JSON... You can play with the examples/corelib/serialization/convert tool and an XSL-T stylesheet. Or write your own extract tool, of course. Or you may want to generate a dummy .cpp file with your extracted strings so you can pass it to lupdate or msgfmt, so it handles fuzzying and merging of new strings into an already-translated file. -- Thiago Macieira - thiago.macieira (AT) intel.com Software Architect - Intel System Software Products _______________________________________________ Interest mailing list Interest@qt-project.org https://lists.qt-project.org/listinfo/interest