[Userscript] Self-Study Quiz


Finally got around to installing the new quiz version from @rfindley and the additional filters from @seanblue . Wow, this is great! I’ve mainly been drilling leeches and new kanji, and this really does the job well. For leeches, I use “All Questions” which is a new experience for me, having to recall the kana after hearing the reading or seeing the English. And for new kanji, I’ve selected all the Apprentices and set the “Time Until Review” to 50%, to not interfere too much with the SRS. I really appreciate the hard work you guys have put in.


One nitpick: on vocab items, I expect to hear audio (when I have it enabled) after a correct reading. I hear it after typing in an incorrect reading. That’s not normal is it?

I’m running chrome on Win10.

Edit: I have lightning mode enabled. If I turn lightning mode off, then I hear audio properly for both correct and incorrect readings.


I think there’s an issue with Katakana input, I tested this word a few times and only the Hiragana variant seems to work.


:point_right: [ v3.0.16 ] - Fix audio in lightning mode. Fix mixed katakana/hiragana answers.

@nerdymartian @blerian


Thanks, it works!

Not a bug, but wondering if it’s possible for me to pre-study the next five (or ten) lessons in my WK queue? For example, my lesson queue typically remain stable on WK’s server until I quiz them, unless new items become unlocked, at which point the server will randomize the lessons again. But until then, I normally have many hours available for me while lessons remain unchanged until I decide to quiz.

I notice your self-quiz app (or maybe it’s the filters) randomizes the in-queue lessons. I’d like the option to choose the same ones WK gives me, so I can pre-study them. And if that’s possible, it would be nice to associate the number of lessons separately for each saved setting. For example, one saved setting for pre-study could choose 10 in-queue items (5 readings + 5 meanings), without messing up the number of items for other saved settings.


This would be better suited as companion script. The Quiz script has an interface that other scripts can use to pop up a custom quiz. The interface is somewhat primitive, but I figured I would leave it that way until people find uses for it.

I’d be happy to give input to another developer, but I want to stay focused on getting wkstats.com updated.


@rfindley When hovering over “Pairing” it says the shortcut key is Ctrl-P, but that isn’t reflected in the code. I have added it for myself and share it below for the benefit of others.
P.S. Your code is always so intuitive.
line 1628-1631

    var keycode_xlat = {
        '8':'Backspace', '13':'Enter', '27':'Escape', '37':'ArrowLeft', '39':'ArrowRight', '65':'KeyA',
        '69':'KeyE', '72':'KeyH', '76':'KeyL', '79':'KeyO', '80':'KeyP', '82':'KeyR', '83':'KeyS', '112':'F1',

line 1673-1678

                case 'KeyL': toggle_lightning(); break; // Lightning
                case 'KeyP': // Pairing
                    e.preventDefault(); // Ctrl-P usually prints the page
                case 'KeyR': // Re-quiz


Great, thanks!

In the Framework, I’m adding support for a Scripts dropdown menu on the Reviews page, which affected some of the CSS customizations in this script’s Settings dialog. Once I have that in place, I’ll push an update with your changes.


Just wanted to say I really like the audio quiz. it wasn’t I started trying to figure out what the person is saying in the quizzes that I realized how different the pronunciation is (in my head) for some words. LIke ぶ (bu) is not really like a shortened “boo”, but it has a different vowel sound. Also, I couldn’t figure out what the speaker was saying for なげつける (nagetsukeru). It sounded like nyanyetsukeru to me, which apparently is the Tokyo accent. Since that’s where I plan to live eventually, I better start pronouncing like that. Anyway, good stuff!


As always, a follow up complaint: I got marked wrong for answering the English for 聞く(kiku) as “to work well” (利く). They’re pronounced exactly the same. I suppose there’s no fix for homonyms.


That’s something I had wanted to handle, but decided I had too large of a stack of stuff to do. Maybe I’ll address it eventually.

The solution is to index all vocab by reading, and if an audio quiz answer is wrong, check the index to see if any other words with the same reading have a meaning that matches the answer. The down side is that it would be easy to always answer with one word, and end up ignoring all of the homonyms.


You could always do what Bunpro does and say “can you answer this another way?”.


True. Maybe a setting to require answering all of the learned homonyms.

It would still require adding the detection, though. Maybe someday…


@rfindley Hey Thank you for making this script. :slight_smile:

I was wondering if you could please clarify something for me? This script has a built in srs? like the same as wanikani? meaning if I get an answer correct, it gets pushed back weeks months days etc?
I know Wanikani has capped their srs after burning items. But I was wonder what are the srs settings for this script?



There is no SRS for the script. Its purpose is to let you study as much as you want, whenever you want.


Seanblue is correct.


Did my eyes deceive me, or did you just edit your post to change who you were replying to? I’ve never figured out how to do that, if it’s even possible.


If it’s not too much trouble, would it be possible to please implement an uncapped srs function for the burned items?


I may have initially clicked reply to you, then canceled and replied to RAIJIN… but I don’t think that’s what you mean.


I’m not planning to make any significant changes to this script since I’m focusing on rewriting the stats site, but this script does have enough of a programmable interface that someone could write an SRS add-on script.