Skip to content

FAQ

Jump to bottom
Akash Mozumdar edited this page Jun 17, 2021 · 26 revisions

Frequently Asked Questions

Why didn't you answer my request/issue (whether on GitHub, YouTube, or email)?

As of April 2020 I've pretty much gotten tired of working on Textractor, and don't want or plan to do new development beyond minor bug fixes. If your issue can be resolved simply using the existing features of Textractor, I'll tell you how to do so, but if it would require me to develop Textractor significantly or in another way require significant effort from me to troubleshoot, I probably just won't reply: you're on your own. Sorry.

Textractor is extracting text mostly correctly, but there's some extra characters as markup/garbage (e.g. a \n in place of every line break). Is there a way to clean the text?

Yup, use the 'Regex Filter' or 'Replacer' extension. Remember to put the extension near the top of the list so the other extensions see the cleaned text. Some useful regex filters:

  • \s (filters all whitespace)
  • [\u0021-\u00ff] (filters all european language and most special characters)
  • [\u0100-\uffff] (filters all non european language characters)
  • [\u0000-\u2fff\ua000-\uffff] (filters all non Chinese/Japanese/Korean characters)
  • <.+?> (filters all HTML tags like <p id="some_guid"> or </span>)

Textractor is extracting text with some characters missing or is unable to extract any text remotely close to what I need. How do I extract the correct text?

First, remove the 'Remove Repeated Phrases' and 'Remove Repeated Characters' extensions and in the settings turn off the repetition filter. These are known to sometimes filter out text even when they shouldn't. If text still isn't being extracted, unfortunately it looks like you found a game with an engine that Textractor doesn't natively support. There's two things you should try:

  1. Search for the text itself in memory. To do this, click Search for hooks and Search for specific text. Type in the first few characters of the text that you're trying to find. So if the game is displaying the text Hey dude, what's up? you should type Hey du. Now click Ok: with any luck, Textractor will be able to find where the game stores its text and one of the resulting hooks should be able to continuously extract the game's current text. This entire process can be seen in the tutorial video.
  2. Brute force search for hooks. To do this, click Search for hooks and Start hook search (you should also check the box if searching for Chinese/Japanese/Korean). Once Textractor says it's initialized the hook search, click through the game to make it display new text. Once the search has finished, Textractor will display a list of hook codes along with the corresponding text they extracted. Go through the list to find the text shown by the game, and copy the hook code. Then click Add hook and paste that hook code and click Ok.

If you find yourself needing to use either of those techniques, please post an issue. It's very useful for me to keep track of which games are not currently working with Textractor, so that I can figure out which game engines they use and add native support. However I probably won't fix it unless and until I find some other games using the same engine.

I'm getting massive performance issues.

The most common culprit of this is the 'Remove Repeated Phrases' extension and the repetition filter in settings, so try removing those (it's pretty expensive computationally to detect and remove repetition, this won't be improved unless some genius out there knows a better algorithm). Performance issues may also be caused by the Unity/Mono game engine. This is a very big game engine so Textractor has to insert a huge number of hooks to reliably automatically extract correct text from it. However you can manually configure Textractor to only insert specific hooks - see https://github.com/Artikash/Textractor/issues/307 for instructions on doing this.

I'm getting white boxes when extracting non-English text.

Try changing the font (right click the text output box and click Font) to something that supports the language of the text you're trying to extract.

Can I pre-translate names before they're sent to Bing/Google Translate?

Yup. Use the 'Replacer' extension. Remember to move it to a place in the extension list above the translation extension.

The text I'm trying to capture is split into multiple text threads. Can I merge them somehow?

Yup. Use the 'Thread Linker' extension.

Can I use Textractor to extract text and some other program (e.g. Translation Aggregator) to process it?

Probably. If that program can receive text via the clipboard then use the 'Copy to Clipboard' extension and enable clipboard monitoring in the other program.

I'm getting an "error while translating" error message!

The translation APIs temporarily ban your IP if you use them too much without an API key. The most reliable solutions are to either supply an API key or use a VPN so your IP gets masked. Using an API key supports the translation provider (for whom the infrastructure to have an API for public use, let alone one that embeds a high quality neural network, is not cheap) while using a VPN is basically stealing...anyway it's your choice.

If you don't want to do either of those, try restarting Textractor, that might fix it. If not, wait (or use a different translation API) for a few hours then try again. If you're still getting that error message post an issue.

Clone this wiki locally