DictionaryforMIDs Esperanto-Japanese Dictionary

Request for Help

        This edition of the DictionaryforMIDs Japanese Dictionary uses romaji to input Japanese.  An input system based on hiragana would be a much better option. If you know the Java programming language and can add a hiragana input system, then please contact one of the then please contact one of the developers.

        The best place to start looking is at the "OwnTextField" Java code.  It is available from the DictionaryforMIDs forum.

1. IME (Input Method Editor)

        The "OwnTextField" Java code can be adapted to be used as an input method editor (IME).  An IME is a program that lets you enter Japanese on a computer.  The IME on a computer can convert hiragana to kanji.  The IME for a cell phone would be much simpler.  It would only convert cell phone number buttons to hiragana.  It will not convert the hiragana to kanji (a custom made IME to convert hiragana to kanji would be far too large to fit on current cell phones).  The IME will replicate the IME on a phone sold in Japan.  Here is the layout.

2. Database Setup

        Once an IME for Japanese is developed, then the search setup can be updated.  This edition of the dictionary uses a NormationClass to search with romaji.  The romaji is first converted to hiragana and a search is run.  Then the romaji is converted to katakana and a 2nd search is run.  Searching is slow since the search is done 2 times.

This search is done with 3 files:

• DictionaryForMIDs_EpoJpn_Lernu.jar\de\kugihan\dictionaryformids\translation\normation\NormationJpn.class
  [DictionaryForMIDs_DictionaryGeneration\source\DictionaryForMIDs\de\kugihan\dictionaryformids\translation\normation\NormationJpn.java]
• DictionaryForMIDs_EpoJpn_Lernu.jar\char_lists\romaji_hiragana_utf8.txt
• DictionaryForMIDs_EpoJpn_Lernu.jar\char_lists\romaji_katakana_utf8.txt

        A better option is to change the NormationClass to a DictionaryUpdateClass.  This will enable the search to be done 1 time instead of 2 times with the NormationClass.  This will lead to much faster searches.  See http://dictionarymid.sourceforge.net/newdict.html and http://dictionarymid.sourceforge.net/newdictNormation.html for information on DictionaryUpdateClass and NormationClass.

        This will be done with this file:

• DictionaryForMIDs_EpoJpn_Lernu.jar\char_lists\katakana_hiragana_utf8.txt

        Currently the katakana_hiragana_utf8.txt file is not used at all.

        All katakana dictionary entries will be hard-coded to hiragana in the index files:

• DictionaryForMIDs_EpoJpn_Lernu.jar\dictionary\indexJpnX.csv

        This conversion will happen during the creation process of the dictionary rather than at the search time.  Then at search time you would enter hiragana for all katakana words.  Only the indexes will contain the katakana words as hiragana.  All katakana words on the cell phone will still display normally as katakana.

        Say for example you want to look up the word 'bus' (バス). You would type in 「ばす」(hiragana). Then 「バス」(katakana) would show up as a result.  Another example is with kanji/hiragana words: say you want to look up 'sweet' (甘い). You would type in 「あまい」(hiragana). Then 「甘い」 (kanji/hiragana) would show up as a result.

        Long vowels aren't converted from katakana to hiragana. They are still "ー".  To enter long vowels on cell phones you type hiragana then add the katakana "ー" with the hiragana. This is to emulate how long vowels work on Japanese electronic dictionaries.

        So to search for "cart" （カート）, you type:
かーと [hiragana-katakana-hiragana]

        If you need any further assistance with the details then please contact dreamingclouds (at) gmail.com.