Nah, I just switched to using the web version which seems to work better, though with a lot of issues. The workflow that seemed to work best for me was
- Click ++ circle icon in top right -> translator * language (click > to expand) -> DeepL -> check "Show interceptor window", also I recommend changing Code Escaping Algorithm to Hex Placeholder
- Export all game scripts into .xslx (I tried a bunch of other formats like .csv, .odt, .xls, but this one was the only one without issues for me)
- Use the "split" function of the attached script (requires Python3, make sure to install the required dependency first with pip3 install pyexcel-xlsx and check the top of the file for more info)
- Import the new .xlsx scripts from "output_split" back into Translator++
- Close Translator++, find your autosave.json from the project you just created (should be in Translator++\www\php\cache\ in the newest folder there) and use the "fixjson" function of the script there on it to fix the line breaks (I had issues with linebreaks being messed up on the translation tab after importing spreadsheets of all formats until I set the linebreak char in this way, not sure if it's a problem for everyone but might as well to be safe)
- Reload Translator++ and reopen the most recent project, translate your scripts a few at a time (I generally aim for about x/30~35 total to avoid getting blocked)
- Export your scripts again from the top menubar, making sure to clear your selection on the left first (otherwise Translator++ won't save your translations)
- Hit win+r and paste/type %LocalAppData% to open your local app data folder, delete the translator++ folder here (to clear the built in web browser's cookies/cache)
- Repeat 6/7/8 until all files are translated
- Now you may think you're done but... unfortunately the DeepL addon has a tendency to dump untranslated text into your scripts. You can use the following regex find/replace to remove them: Find: [\s\S]*([一-龠]|[ぁ-ゔ]|[ァ-ヴー])[\s\S]* Replace: (empty) But you might want to check and verify it's not just speaker tags or something first because you can probably pretty easily fix those without trashing otherwise perfectly good translations.
- Export again to save after removing untranslated lines
- Copy the root folder with your original xlsx files and output_split folder (now filled with the exported xlsx files you just translated) into a new copy, naming it something like "Pass 1" or something
- Use the "merge" function from the provided script on the new folder you just copied
- Copy the contents of output_merge back into the root folder, overwriting when prompted. Once you've copied all the merged files out, you can delete output_split and output_merge from the newly copied folder.
- Use the "split" function from the provided script to split out only the untranslated lines from your newly merged .xlsx files. (This is needed because Translator++'s "ignore already translated" function just straight up doesn't work with DeepL as far as I know, so we need to handle it ourselves in the input file.)
- Import your .xlsx files from output_split into Translator++ as a new project. There may be less of them since any files that were fully translated in the last pass won't appear here.
- Use fixjson again to fix the line break setting on your newly created project
- Now go back to step 6, repeat for as many passes until a fairly small amount of untranslated text remains. Usually there's a few lines that DeepL stubbornly refuses to translate no matter what, though you may be able to get them individually after setting your default translator engine in options to DeepL and just using ctrl+g on them. That or just run a pass with Google which should ignore the already translated lines unlike DeepL and just fill in the gaps.
- Export and merge your final .xlsx output files, you should now hopefully have a set of fully translated .xlsx files... But unfortunately importing these back into our original RPG maker project isn't as easy as I'd like so I added the somewhat hacky solution into my script to simply write your translation directly into the autosave.json which we will do next, so reopen Translator++ and create a new project with the final merged .xlsx sheet. Find the autosave.json for this new project, as well as for your original RPG maker project. (If you can't find it it's ok to go back and just recreate that one too since we didn't translate anything)
- Run translator_tools.py inject "[EXCEL autosave.json path]" "[RPGMAKER autosave.json path]"
- Hopefully if all goes well it will finish without errors and you can reopen translator++ to your RPG maker project and see your fully translated scripts populating the 2nd column. Wasn't that easy! If the previous step fails you can try modifying the CHAR_REPLACEMENTS or PREFIX_EXTENSION variables at the top of the script to get better results since occasionally some characters don't exactly match up between the exported and original RPG maker files and need to be accounted for there.
Now after you did all this you should hopefully be able to export your project back into your RPG maker game and be ready to play it. (Besides line wrapping, which I've posted about some solutions to help with that some time before since Translator++'s built in word wrapping is still kind of bad as far as I know.) Though with older VX/Ace/whatever games I have noticed that occasionally the automated RPG Maker Trans step will fail and you will need to export into RPTrans patch first, and then open all files with notepad++ or something, Find: > [0-9].*\r\n Replace: (empty) and then save all and patch yourself with RPG Maker Trans and it should (hopefully) work, but be careful because sometimes random lines in Scripts.txt also get matched by this but then again you probably don't want to translate Scripts.txt anyway because it tends to break stuff. I haven't had any issues exporting directly into MV games yet.