Skip to content

Mining from visual novels

Visual novels are possibly the best resource to learn from. The combination of written Japanese and voiced text and dialogue makes them a good resource both for beginners and advanced learners and most importantly, they are fun. In this section I will show you how to extract the text from a visual novel and use Yomichan and ShareX to 'mine' them.


Textractor

Textractor is a text hooking program that allows you to extract the text out of most visual novels. You can grab it here. (Get the English only version.)

  1. For most visual novels you will need the x86 so start that up first.
  2. Remove the following extensions by clicking on them and pressing delete:
    • Bing Translate
    • Extra Window

You shouldn't use machine translation tools no matter at what stage in your learning you are. They are useless and will ruin your experience. If you want you can remove the Regex plugin as well (it spawns an annoying extra window every time). There are cases where it is useful however. Basic usage of it is detailed below.


HTML Page and Clipboard inserter

To make use of the extracted text to do lookups and create flashcards with Yomichan we utilize an empty html page with a plugin that pastes our clipboard into it.

Get the plugin here:

You turn it on by clicking on the icon inside the browser task bar. If you are using Chrome you may need to allow it to access local files. Otherwise it is incapable of accessing the html page we download in the next step.

  • chrome://extensions/ → Details → Allow access to file URLs

Html pages:

  • With removable lines (but incompatible with the mpv script introduced in the anime section):

    Download

  • Without removable lines (but compatible with the script):

    Download


Hooking a visual novel and mining

  1. Launch your visual novel and start the game so some text is displayed on screen
  2. Launch the html page and activate the plugin
  3. Launch Textractor → 'Attach to game' → Find and select the process for your game → OK
  4. Go back to your visual novel and advance by one line
  5. Go back to Textractor and scroll through all threads until you find one that matches the text

Now the visual novels text will be copied to our clipboard whenever we advance the text. → You can directly paste it into Anki and it appears in the html page!

The final text hooking result.

You can now mine from visual novels:

  1. Identify a word you want to mine and create a card with Yomichan
  2. Paste the text that is inside your clipboard into Anki
  3. Record the audio and paste it into the audio field
  4. Take a screenshot and paste it into the picture field

For taking visual novel screenshots I prefer the 'capture active window' ShareX function. How you replay audio differs by visual novel. For some you do it with a hotkey, sometimes you need to go into the text log and click on the line and sometimes it's not even possible. In that case you can try the method I suggest for video games. (TBA)

How this whole process generally looks like is shown here:

The card making process in the video is somewhat outdated so don't sweat the details.


Saving a hook

After you successfully set up a hook you should click on 'Save hook(s)'. This will cause Textractor to remember the hook and activate it automatically the next time you launch the game.


Textractor Advanced

Although it will work with 90% of VNs ouf of the box Textractor has some advanced functions for finding hooks and working with text which I'll show here.

Hook codes

Textractor should be able to find a hook code for most visual novels automatically. If it can't you can try taking a look here: https://vn-hooking.fandom.com/wiki/H-Code

If you are lucky and it has a code for the game you are looking for click on 'Add hook', paste the code and press OK. After advancing the text the hook should appear in the list.

Searching for hooks

You can also search for hooks. While this rarely works it's still worth a shot.

Click on 'Search for hooks', tick 'Search for C/J/K' and start the hook search. Advance the text once and wait for the result. You will be given a large list of potential hooks, one of which hopefully contains what you were searching for which you can then add.

Using Regex with Textractor

Regular expressions are an extremely powerful tool for finding certain text or matching text to patterns. The Regex extension for Textractor lets you use regular expressions to filter the output for your clipboard. Let's say for example that a visual novel is giving you both English and Japanese. You could write a Regex to filter out the English text. Things like Furigana may also appear weird, take this for example:

耕司こうじ would render as something like #{こうじ}耕司# in the Saya no Uta HD edition.

We can filter this with the following Regex #(\u007B.*?\u007D)? and get a much nicer output.

The final text hooking result.

Useful links: