[Userscript] Keisei 形声 Semantic-Phonetic Composition

Fixed, there were actually no kanji in WK that fit it as a kanji, it shows up in another tone mark 亲 -> 親.

I will check the DBs automatically for their consistency soon.

@acm2010 Do you auto-mark some of the kanji (making them yellow)? If not, I must have clicked a bunch without realizing.

They were auto-marked for me too.

I already pre-marked a handful that didn’t look like their tone mark …


… hmm more than expected :slight_smile:

魂 listed twice.

1 Like

One more.


@gzovak Thanks, there were 2 duplicates and 1 inconsistent non-compound in the DB, fixed now.

This script is a God send! Amazing work, and absolutely the best tool we could ask for!
I’ve missed this and wished they did something officially, it is a lacking feature that would be a great benefit to implement as a standard. For now we’ll have to rely on helpful souls to make it in script form instead, and you did! <3
Can’t thank you enough!!


@Toyger Thanks! The database required some work, judging from the speed of how “similar kanji” are added I wouldn’t hold my breath :wink:

Small new feature for 1.3.2:

Readings of phonetics are now sorted by their usefulness, and readings that do not appear in WK are grayed out, like this:


Can you do anything to indicate when the matching reading isn’t included on WaniKani? For example, the script says 噌 matches for the reading そう, but WaniKani only includes the reading そ. While it’s still definitely good to know that it has a phonetic component, it would also be good to make it clear that you can’t use that answer on WaniKani.

1 Like

I added a check and changed the text color a bit in version 1.3.5:


Note that I used the readings in the order as they appear in kanjipedia or jisho, so the reading in WK may just not be number one.

Not sure if けい is better than きょう for example. I could reorder the readings to match WK better.

Edit: version 1.3.6 because kanji that actually are not in WK were styled twice and it became unreadable.

1 Like

No pictures for the phonetic group of 昔.

Edit: Also for 九. Probably more.

@gzovak That was a bug of the “show missing readings in WK” feature, failed when something didn’t have an onyomi. Fixed in 1.4.0.

New feature:

  • I added the phonetic marks not in WK as kanji to see their original meaning (table in the example)

Just curious, did you mean to show the yellow text when the reading isn’t an accepted onyomi reading in WaniKani or when it’s not the primary onyomi reading in WaniKani? For example, shows the yellow text since せい isn’t the first reading in WaniKani, but it is an accepted reading.

@seanblue, there was a bug in the comparison (javascript split leaves the whitespaces, and “せい” != " せい"), fixed now in 1.4.1.

1 Like

Sorry if this has been said before but what on earth is 拠 doing in the 豦 tone mark group?

Some kanji have changed over time and have tone marks that are no longer visible in the kanji. If you look on jisho, you’ll see that it has a variant 據 that has the expected tone mark.

With that said, I think @acm2010 usually pre-marks those yellow and unless I accidentally unmarked mine, I think it should be added to the list.

P.S. I wanted to tweak my wordng a bit, but I saw @acm2010 typing and had to post my response before he beat me to it. :stuck_out_tongue:

1 Like

We actually did discussed this before :smile:

Sometimes the ways of the kanji masters are inexplicable, 足+責=跡?

1 Like

I was just watching before you posted :wink:

There is also a list of “obscure” kanji above:

1 Like

No information about compounds of 士「し」?

士 (samurai), 仕 (doing), 志 (intention), 誌 (magazine) all are read し