A kanji is a Japanese character. These characters are put together to make up words. PocketKanji is a program to help you look up kanji.
Kanji usually have two or more readings. The first is called the ON yomi (yomi means 'reading' in Japanese). The ON yomi is a reading that is somewhat close to the original Chinese reading. The other reading is the KUN yomi. This is the Japanese sound associated with a character. There is also a third reading that is used with people's names. This is called the NANORI reading.
Download the latest version.
This program has no warranty. I do my best to produce a good product but an oversight on my part could could crash your Palm Pilot. Please backup your data before using this program.
You need to sync the following files to your Palm Pilot:
kjpad.prc: The main programkjdef.pdb: The database for the kanji dictionaryjstroke.pdb: The database for kanji recognitionshinonome.pdb: A 16x16 pixels Japanese fontPockeKanji does not work with a hack called McPhling. It is a hack that makes it easy to switch between applications. You need to disable the hack in order to see the definition.
Write a kanji in the box. It is important that you use the correct stroke order for it to be recognized. Most kanji are written from the top left to the bottom right.
Tap on convert for the kanji to be recognized. The program will display the four kanji that best match your drawing. Tap on the correct kanji to find the definition. If the program does not match your character you will need to tap erase and try again.
This is the definition of the kanji which is displayed on the upper left. Next to the kanji is the EUC and Shift JIS code. Below the kanji are the ON reading, the KUN reading and the NANORI reading. Each reading is separated by a comma.
There is a dot separating the hiragana kanji reading and the hiragana that often comes after a kanji. For example in ito.shii (love), ito is the kanji and shii is the hiragana. The hyphen means that the reading is preceded or followed by another word.
To copy the character into the clipboard bring up the menu and select copy.
You need a Unix machine with GNU tools such a binutils and make. You
will also need the gcc compiler for pilot and the pilrc tools as well
as a program called par to create the kanji dictionary database for
PocketKanji.
Download par and compile it on your machine.
My development machine is a Linux machine, but you should be able to compile PocketKanji on any GNU environment. If you are running Windows you should be able to compile PocketKanji using the PRC-Tools with Cygwin.
All you need to do is type make prod in the top directory.
PocketKanji uses an algorithm invented by Todd David Rudick. It starts looking at the number of strokes entered. It then looks at the strokes direction. It adds points for stroke that are entered in the right direction. The highest score is hopefully the character the user has entered.
That is why the program is so picky with stroke order. The slightest mistake when entering the character will probably show the wrong character.
You can take a look at the database. You'll need to download the source file and expand it. It is split up in many files under a directory called stroke.
The files are named unistroke.n, where n is a number between 1 and 23. This number represents the number of stroke. For example unistrok.5 will contain characters having 5 strokes.
Let's take a look at an entry in unistrok.3:
53e3 3382 | 2 b 6
This is the entry for 'kuchi' which means mouth. The Japanese character looks like a square.
The first entry 53e3 is the unicode entry of the character. The next number is not used. The 3 values after the pipe characters represent the stroke direction.
The following diagrams shows the correspondence between the value and the stroke direction.
| 1 |  |
| 2 |  |
| 3 |  |
| 4 |  |
| 6 |  |
| 7 |  |
| 8 |  |
| 9 |  |
| b |  |
| c |  |
| x |  |
| y |  |
Looking at the table above you can see that the sequence 2 b 6 makes up kuchi.
Sometimes there is a pipe after the stroke sequence and they are more values. These are used to differentiate between characters that have identical stroke orders. I will document those values later.
With this information you should hopefully be able to start adding your own character. You can also use stroke order in the dictionary to make up more complicated characters.
The PocketKanji software and manual are licensed under the GNU General Public License version 2.0.
PocketKanji uses KANJIDIC which is licensed under the Monash University School Of Computer Science And Software Engineering Electronic Dictionary Research And Development Group General Dictionary License Statement.
This is the "brown paper bag" release. That is to cover my head in shame, I forgot to include a file in the zip file that is necessary for PocketKanji.
Many thanks to Michael Volkmer for reporting the bug.
This is the earthquake release. Seattle had a 6.2 earthquake and I am glad to still be alive.
Initial release.