What do you want now? (Request extensions here)

Thanks, that did the trick! :blush:

2 Likes

I want a countdown for my reviews like with memrise for example, I want it to be visual and not just show the seconds. And I want to be able to customize it, say 10-15 seconds would be plenty for me. Maybe if I get a wrong answer I get 15 seconds if not just go with 10

I think that if I create a bit of urgency I might be able to no get lost in thought as often while doing reviews

This is the closest existing script

2 Likes

Is there a script that tracks how long it took to answer each individual question and provides access to that data via WKOF-compatible filters? E.G., filter to return only reviews where answering meaning took longer than X seconds. It would probably make sense for the script to also track whether meaning/reading was answered first and make that data available too.

If that script doesn’t exist, would anyone care to write it?

I don’t believe there is any such script

There’s one tool i want and will probably write myself:
Multi Radical Kanji Search.
edit: website is online, see Multi-Radical Kanji Search thread

You enter multiple radicals, and get a list of kanji containing all of them.
It would look similar to this:
image
This already exists for RTK radicals, try it here:

The problem is, RTK radicals are very different from WK, not just the names, WK also has a few radicals RTK doesn’t.
(I know WK doesn’t use the “official” “radicals”, which are based on the 18th century chinese Kangxi dictionary, which is surely an unimprovable framework! They don’t even have “official” japanese names, only common nicknames - rant rant)

Why would this be handy?

Because it would be faster in many cases where a radical is contained in many kanji.
Let’s say we’re looking for . Try entering ‘turkey’. It would probably take a long time to find your kanji. But if you enter ‘turkey water’ (water in RTK is tsunami in WK), you’ll find it instantly.

If we search by turkey in WK, we get this wall to search through. Happy hunting!

A more ambitious improvement to WK search would be being able to search not just the WK set of 2055 Kanji, but at least the most common 2500 or more. Of course that would require “annotating” a lot of Kanji by hand. Ideally you’d be able to import and export the radical to kanji mappings, so it can always be extended, maybe collectively by the community, and you can swap between sets like WK, RTK or even your own. This may be be a future addition to the script.

There is actually one app which can basically do this already with WK radicals, Houhou. But:

  • It’s only on Windows.
  • Have to do a lot of clicks to enter multiple radicals
  • Doesn’t filter interactively (live), and you can’t delete filters (radicals) individually.

There is also already a visual version of this:



It’s nice, but still, searching for the radicals is a bit slow and painful, even if you know the stroke count.
(i suppose you could memorize the position of each radical in the grid with enough patience)

My questions to you all:
Would you find this useful?
Any feature requests or comments? What would you like improved compared to the RTK version?

Features i’m planning so far:

  • Copy to clipboard button for each kanji (maybe just click on the kanji)
  • Search radicals by name or (maybe/later) graphic
    • order radicals by stroke count, but not with each stroke number neatly on the same row or column
    • display the name for each radical (maybe just by hover)
  • Radical to Kanji associations: Can modify, export and import (later)
    • Switch between WK, RTK and custom names (later)

One note about the RTK version

The original github repository for it was taken down by a DMCA claim from RTK. I’ll have to be careful not to infringe on RTK either, maybe think about what exactly what the problem was and how to avoid it. Curiously, i found a fork of it that wasn’t taken down (that i probably shouldn’t link). In the worst case, just leave out RTK support and leave it for users to import the radical names.

I’ll probably try to get a prototype started this weekend, unless the replies make it obvious that it’s a completely stupid idea or something ^^ i’ll also give this its own thread then.
edit: Multi-Radical Kanji Search thread is up

3 Likes

Make your search bar into a wkof filter and Item Inspector will be able to use it. The related kanji will be automatically displayed in a table with popups to display associated information. You can already configure Item Inspector to display the radicals within in a kanji in a popup so people will be able to look back at which radical this kanji respond to. Item Inspector also has an export function built-in. Maybe you will be able to use it.

The idea is you make your filter select the kanji and Item Inspector will do the rest of the job for you.

There are several filter scripts out there you may use as a reference code on how to write filters. They are not hard to do.

Not sure if requested, but something that reads aloud the Kanji readings instead of only vocab readings. I am a slave to my job so I only have time for “Anki Mode” and audio confirmation is a big help to not messing with my progress.

1 Like

Cool, i’ll look into using that, thanks!
But is wkof+Item Inspector restricted to the 2055 WK-Kanji or can it also search an extended Kanji set?

1 Like

It is only the WK kanjis.

  • The WK API has the information of which radials are in which kanji. This information is not available for other kanji or radicals in WK.
  • If you plan to use the WK radicals I don’t see how to apply that to another kanji set unless you can find some source that has done that categorization.

If you want an extended kanji set you will need something entirely different.

The limitation is that WK doesn’t supply audio files for kanji readings. I don’t know anyone who does. There is also the problem of what to do with kanji that have more than one reading. Which audio shall we play?

I would do the radical categorization myself for an extended kanji set.
I think the visual multi radical search has a larger set already.

1 Like

I think it may boil down to how much information you want to display about the kanji. With Item Inspector the WK information is built-in. You will get that without much effort. And you get export too. If you do a separate project you will have to integrate everything.

One possible approach may be to do the wkof+Item Inspector first because you can get a lot of functionality up and running quickly this way. Then you can move to the extended kanji as a separate project.

Yes, that’s a good idea, the wkof + item inspector version may be my first prototype (and useful for WK learners).

However, my ultimate goal is for this to be my default kanji lookup tool, and i regularly see non-WK kanji these days.
For me (high typing speed), it would be much faster and less annoying than drawing the kanji or any of the kanji search methods i tried.

1 Like

Just a heads up. You will run into a user interface issue. You may look at the Item List filter for an example. By default you have to type the radicals into a small one-line text input. The Visually Similar Kanji filter has solved this issue by inserting a dialog in the settings with a better UI. You may look at this code for a reference.

1 Like

Hello,
Is there a script that would show how many items have moved to apprentice/guru level based on your latest review? I’m thinking something like Guru1 100(+10)

1 Like

The Heatmap script shows this information for the whole day. If you click on any day in the Heatmap you get this pop-up. The SRS column displays the SRS counts of the items before and after the review. There is also the score in the top right (+27), which indicates the net number of SRS levels advanced

image

1 Like

I completely missed this one O_o Thank you so much :blush:

1 Like

I miss the stroke order script that recently stopped working. Complementary to that, I would like to have an option to display a handwriting font by default, because I write out my reviews to practice writing.

1 Like

I like the transitive/intransitive pair item id for all those pesky verbs that have those parts of speech that so infuriate me.