[Userscript] Multiple Answer Input revamped

@ganbareniichan Hi again! Sorry for the late reply, got distracted. I released an update, please tell me what you think!
(Keep in mind that it does require WaniKani Open Framework from now on, so make sure you have it installed!)

You can find the setting to turn off the requirement of all answers being correct in the top left during a review (together with all other WKOF script settings)

2 Likes

Awesome! Thank you so much. I donā€™t have any more reviews until tomorrow but Iā€™ll be sure to check back with how I like it!!!

1 Like

Itā€™s broken for me (Firefox + Tampermonkey).

As for my idea of improvement (although I havenā€™t used it yet): let users decide what readings / meanings are important (and without any decision, only one answer is required).

1 Like

Thanks for the report, Iā€™ll have a look at it later today!

Concerning your suggestion, would I be right to assume you mean a way to only mark specific meanings/readings as relevant enough to fail the user for?

Making such a system on a blacklist-basis would mean a lot of manual configuration for the user.

What would you think about making it possible to ignore a few select readings and meanings manually instead? e.g. being failed for something you consider unimportant and then manually whitelisting it, a la ā€œI donā€™t care about this reading, donā€™t fail me again if I donā€™t mention itā€?

This would mean all answers are required except for the ones you specify arenā€™t. I think that would mean a lot less configuration work for the user.

1 Like

This is good for readings, but not for meanings. Also, WaniKani can easily miss some important ā€œsensesā€.

Because there tend to be a lot more meanings, you mean? I can understand that, but selecting a few meanings you donā€™t want when it comes up seems a lot less tedious than having to go through each vocab manually and essentially defining user synonyms for all of them. I fear that would significantly reduce ease-of-use out of the box, with only the benefit of avoiding a single fail one time (which could even be mitigated via answer override or a similar script).
Are there any more downsides to this that I am missing? Or is the one I described worse than I realize?

What do you mean by this?

Some meanings are quite similar to others (with some Englishā€™s, rather than Japaneseā€™s, differences); some are just different entities. Therefore, meanings can be categorized into meaning groups, as you can see on Jisho.

1 Like

I see, that makes sense.

Then, how about checking to see if at least one meaning from each meaning group + all user synonyms have been entered?

This would ensure you donā€™t have to type essentially synonyms multiple times because multiple ways of saying the same thing are listed, while at the same time ensuring you list all actual, topical meanings.

Though, meaning groups arenā€™t really defined in WaniKani. Everything are really just comma separated list, or the such.

It might be possible normalize (partially) with regex or something.

1 Like

Yeah, Iā€™ll have a look if there are any good databases I can yoink from. Weā€™ll see.

First Iā€™ll get it to work again :slight_smile:

Thankies for the feedback btw

@polv

Alrighty, so, a few things:

First off, the script doesnā€™t match for extra studies yet, which is a simple enough fix. It is also, however, not broken on vanilla WaniKani.
That said, I did manage to break it eventually, by also using other scripts.

Are you perchance using Wanikani Double-Check as well? My script works fine, but Double-Check appears to override window.answerChecker.evaluate with its own function, return_new_answer, which returns undefined.

The strange thing is that this override has been part of Double-Check since before I even started writing this script, so it doesnā€™t make any sense for the issue to surface now. Maybe execution order in Tampermonkey?

@rfindley sorry to ping you on this as well, but do you have any idea what could have caused this to happen? Additionally, ideas on how to make my script compatible with yours would be appreciated (it should be possible as I only need the answerChecker for its results, I donā€™t need to change it - Iā€™m not sure how to enforce execution order in tampermonkey or similar, though.)

1 Like

Yes, although I donā€™t actively do WaniKani. Actually, I find WaniKani not good enough, and decide to drop WaniKani once again, for a while now. (while I still do Anki)

1 Like

Just sort by script number:
image

Then drag the scripts to the order you want:
image

If yours runs before Double-Check, it will probably work. The key is for yours to hijack the answer function before Double-Check, so Double-Check will get whatever answer results your script returns.

I think you nailed the culprit, but

Simply reordering alone isnā€™t enough. The Enter key will be dead, if both scripts are active.

Is there a possibility this could be done to work the same way where it checks all answers but as long as your first answer is correct, even if following ones are incorrect, it still marks the submission as correct?

I like this idea of being able to test your knowledge on all possible answers instead of one, but WaniKaniā€™s system is as long as you know ONE of the possible answers itā€™s looking for it marks it as correct. With the current system, I would never want to attempt listing the alternate answers because if I could for sure get a correct answer and progress, why risk actually losing progress when I could confidently move forward? Listing the for sure correct answer first would then take the stress out and let me work on learning the other answers.

Hi, there is a setting from version 1.2 onwards that will not instafail if at least one answer is correct; although it doesnt specifically check for it to be the first one.

Does that work for you?

Oh I didnā€™t know there were options to tweak and change. If so then yes thatā€™s exactly what Iā€™m looking for. If it ends up that was there all along, Iā€™m sorry for Necro-bumping lol

1 Like

@Buschwichtel Your script is great, thank you for making it. Occasionally though, it doesnā€™t accept multiple answers that are correct. See example attached. Is this something that you are aware of, or is it a conflict with other scripts?

Would it help you if I reported them as they occur (or not)?

Thanks! It depends, what other scripts are you using in tandem with it? Iā€™d have to check if they could interfere with my hacks. Does it always fail for 実?

No, it just passed with the same input (see attached). I had another fail (kurushii) which Iā€™m pretty sure had previously passed.

I have quite a few scripts installed (list attached) - not all are enabled.

Let me know if you want me to alert you to any more of these issues.