Back

Using Koohii mnemonics in your apps & decks.

#1
Three "top voted stories" from Koohii community for 2200 kanji in RTK  Vol. 1

Included in the kanji-koohii-files repository, a sample database for the open source Kanji Koohii project.

For use in Anki decks or some other tools, see the usage requirements.

(Also includes 2 random non-voted stories per kanji, that is for the kanji-koohii development purposes so there are non-voted stories for different test cases)

Use the correct Study page links
Links to Study pages MUST USE THE UTF-8 KANJI. Frame numbers are relative to the user's selected kanji sequence on Koohii, which is different between RTK ecitions, and also could be completely different if we add KKLD, RTK Lite etc.

Code:
http://kanji.koohii.com/study/kanji/一     CORRECT

http://kanji.koohii.com/study/kanji/1      INCORRECT


In this thread I will answer questions related to the sample database only.

I won't provide any support for making your decks in whatever software you use besides Kanji Koohii. For Anki, ask questions in Anki Small Questions Thread.

Kanji Koohii isn't running on cloud. Please use the sample database and don't scrape the website and reduce performance for other Koohii users. Top voted stories are old stories for the most part, so it doesn't need to be updated much.
Edited: 2017-03-27, 11:05 am
Reply
#2
The sample database also includes dictionary data.

When building your spreadsheet / decks you can do things like that :

Code:
-- find priority entries using 大  ( UCS-2 code  22823)

SELECT jd.pri,jd.compound,jd.reading,jd.glossary,jd.misc,jd.verb
  FROM dictsplit ds
  LEFT JOIN  jdict jd using (dictid)
  WHERE ds.kanji=22823 AND (jd.pri & 0xCA)
  ORDER BY jd.pri DESC;

More info about the JMDICT related tables + split readings is found in rtkLabs.php.

If you make anything useful with that please share useful queries maybe they could be of use on Koohii.

About split readings
Note that split readings in dictsplit table do not match the actual reading in the compound, but the original On / Kun reading which the script found while "splitting" kanji compounds.

So eg. dictsplit entries for ざっし will return ざつ and し , but not ざっ and し (note the sokuon).
Reply