Kanji Sieve 0.5 released

Now with added Kanji Notebook.


New features in this iteration

  • option to turn off tooltips
  • click to display kanji information in statistics view
  • shift-click for options on some buttons
  • October 2010 standard Jyouyou kanji
  • User defined keywords
  • preferred language preference
  • Spanish keywords
  • French keywords
  • Spanish, French and German automatic lookup on wwwjdic
  • added Hispadic and Wadoku to list of online dictionaries
  • Chuta.jp queried for English and preferred language
  • sieve for custom list of kanji
  • searched word in waeijiten is highlighted (verbs and adjectives decline)
  • minor interface improvements
  • improved character count for more accurate overall statistics
  • Kanji Notebook (2010 Jyouyou, KKLD kanji, Kanji Oddysey)
  • phonetic data (work in progress – 75 groups at present)
  • Import images
  • Images and Audio copied to media folder
  • Image Editor from pixlr.com


The kanji in the statistics panes can be clicked to get basic information about the kanji, its keyword and the KKLD index. there is also a checkbox to mark it as known.
Kanji Notebook contains more data on 2537 kanji, covering the 2010 Jyouyou, KKLD kanji, and Kanji Odyssey lists. My motivation to build Kanji Notebook was to get a sort of dictionary that showed the data I wanted, and allowed me to add my own notes and importantly change the keyword. The dataset from EDRDG originally started with keywords from Heisig. These are still there as the first word and were driving me crazy. Jim Breen said he was going to clear up these entries way back in 1995 but I still want to scream when I see “old boy” for 君. The keywords from KKLD are better, or sometimes my own keyword. I haven’t included KKLD keywords but there is the option to add them yourself. The KKLD index from EDRDG‘s data is included to help with a look-up. (If you managed to get the iPhone app before it disappeared putting a ~ before the index number allows you to search for it)
The example words are from vocabulary extracted from records in Kanji Sieve. Therefore all vocabulary should be relevant to your interests.

There is automated look-up of Internet dictionaries for a kanji. I find it interesting to search google images to see if I can guess a clear meaning or see how it is being used. You can explore kanji using the browse tab, to get lists by grade, radical, phonetic, kakitorikun screen, or Kanji Odyssey group.
Kakitorikun is a great piece of software for the Nintendo DS. It’s main drawback is it’s completely in Japanese. With Kanji Notebook I can at least easily put an English keyword to kanji on a given screen.
Kanji Odyssey from Coscom is an excellent e-book based on the frequency of use of kanji. It has many examples and recordings of native speech. I recommend it. An added bonus is their completely free update program. Recently I got a new CD from them when they updated for the 2010 Jyouyou. Absolutely free, not even postage costs. That’s a rare service these days.

There is the beginnings of language support for languages other than English. Spanish and French keywords are included. They will be displayed in preference to English when the preferred language is set in preferences. This also controls what language is used to query chuta.jp. However I found that the returns are very short so English is also submitted. wwwjdic will use the preferred language when searching for vocabulary.
Complete translations of the interface are unlikely at the moment however.


That’s it for now. Next potential release will be January. I really need to spend some more time on documentation! …


More about Kanji Sieve

09. November 2010 by ロバート
Categories: 01 news • 新聞, 02 reading • 読む事 | Tags: , , , , , , | 8 comments

Comments (8)

  1. Hello Robert.

    I am unable to insert a workable list of Kanji into the Custom List.

    Is there any special formatting needed?

    I`ve tried with Unicode, UTF8 and SJIS but to no avail. Kanji list appears pasted in window but when I press save everything canges to a single question mark.



  2. Hello Tom,
    Thanks for the bug report.
    I’ve managed to reproduce it.
    It’s caused by paragraph returns in the text.
    As a workaround you’d need to strip these out in a text editor first. It seems because of the routine I’ve used to strip duplicates from the list it will only accept a string of characters.
    This would also happen in the known kanji pane and filter non-jyouyou pane as well.

    時事時 is ok

    時 isn’t

    The character encoding is Unicode in the application.

    It should be easy enough to fix.

  3. Hi Robert,

    I tried again using JWPce to strip line returns and spaces between the characters but still have same problem.
    Characters copy OK to Custom List window but are reflected as single ? when saved.


  4. Strange behaviour Robert.

    I stripped line returns and spaces between characters with JWPce and exported UNICODE but still get same results.

    However, if I paste same exported text to Word 2010 and then re-export and paste to KanjiSieve everything works OK: Very strange.



  5. Hello again Robert.

    Any chance of having two (or more) Custom Lists?


  6. Hi Robert.

    On the main screen the maximum number of Kanji used in the record shown is 50.

    e.g. 208 Kanji Total
    using 123 different Kanji that you have yet to learn

    kanji kanji etc.
    Kanji etc.

    (50 kanji shown)

    and 73 more….

    How can I see the rest of the Kanji used (the other 73)? Are they shown on additional pages?


  7. Custom list preference behaviour.
    It’s not character encoding as such. If it were just character encoding you’d get mojibake not an error (“?”) I reckon it’s the different handling of EOL in JWPce and Word. I would think Word handles it better.
    I’ve added the stripping of EOL into the routine in Kanji Sieve so I’d hope it won’t be an issue in the future after 0.5.1

    More Custom Lists
    unlikely. It’s not that it can’t be done, but it’s going away from what I want to do, which is track the 1006 Kyouiku kanji, and my belief that it’s best to approach learning kanji within the grade 1 to 6 framework.
    I’ll have a think about an implementation though.

  8. On the stats screens you can only see the mos frequent 50 kanji per screen.
    It’s really only an issue with the unknown kanji screen, when there are many unknown kanji, and then the piece of writing is probably well beyond someones capability.
    You get full listings on a printout.

Leave a Reply

Required fields are marked *