Open Uzukii opened 2 years ago
Have you tried a different translator extension? I use deepl with textractor and it mostly gets the numbers right. It sometimes says "six" instead of "6", but not too often for me. If that doesn't help, you could try using the Replacer extension. Open the SavedReplacements text file and enter this:
|ORIG|100|BECOMES|100|END| or |ORIG|100¥|BECOMES|100 Yen|END|
It might force Textractor to show the numbers. You would have to do it each time one of those roman numbers show up though. In the extension window, make sure Replacer is after Copy to Clipboard and before your chosen translator (DeepL, Google, etc.). I haven't tried it, but it may work.
Good Luck.
I'll try that, thanks.
Do you pay for the DeepL API? Because I don't. My experience with using the free DeepL extension in Textractor has been inconsistent and limited at best. After a dozen or so lines of translating it stops working and the site blocks me for several minutes. So I've been using Google Translate in Textractor + DeepL's desktop app side by side. That way I get translations from both and thus more context because sometimes one might miss what the other doesn't.
No, I use the DevTools DeepL extension. I think it's a bit slower than the free or paid DeepL, but I don't mind. I don't think there's a limit to the DevTools version. I've never come across the translation/time limit. These are my settings for it (these are probably the default settings). If I understand the numbers, it means I can get 30 translations a minute.
I also use a second program, but I use Translation Aggregator on a 2nd monitor just in case DeepL misses something.
What does translation cache do? I keep it turned off since from what I can see, it just logs everything into a bloated text file. Making hundreds of unnecessary writes to my SSD I don't need. shrug
I haven't tried the DevTool version of DeepL but I'll give it a shot. It's not like I'm spamming their servers with the free version. There just seems to be a hard cap on the number of lines you can translate until it blocks you. Which is weird because that doesn't happen when translating the same amount of text using the DeepL app.
Cache isn't necessary. The only advantage I can think of is that if you play a vn with mutiple branches that lead to a lot of repeated scenes, the previously translated phrases will load almost immediately instead of seaching for a translation.
I think what the devtools version does is that it copies the clipboard into a deepl website translate search and copies the results back into textractor. So, I guess that's why there's no limit to translation.
As I'm sure we all know, sometimes roman numeric characters appear in Japanese text when things are quantified. (ie: 100¥) And Textractor doesn't pick up on it because it's only set to one language. Thus it doesn't get translated properly. Is there a way around this?