Fritz: All Fritz/All Fritz.zip/All Fritz/FILES/EDUCUAGE/MOKE2.LZH/MK11DOC.ZOO/moke.doc

Text File | 1990-04-24 | 48KB | 1,316 lines

             MOKE  -  Mark's Own Kanji Editor

                 User Reference Guide  -  1.1 (24 April 1990)

                       Mark Edwards

                   KiCompWare


1.0  Overview

   MOKE is Japanese language editor for the PC with a command set
   similar to the UNIX editor 'vi'. The name MOKE is partially derived
   from JOVE (Jonathan's Own Version of Emacs) and shows some lack of
   creativity. However the only reward I may get in writing this editor
   may be that it bares my first name. This User's Reference Guide was
   entered using MOKE and MicroEmacs 3.10.

   This program is the result of many long nights in front of my PC and
   is based on several other programs (view, rtk, fk). This could not
   really be a Japanese editor without the Simple Kana to Kanji conversion
   program by Masahiko Sato ([e-mail]) and the
   Wnn dictionaries.

   MOKE allows input of english, hiragana, katakana, and most importantly
   kanji. As far as I know there is "no good" way of entering kanji on
   a computer. All methods require a number of key strokes first to enter
   kana, then more strokes to select a kanji character. In that respect
   this editor is no different. It should be noted that this is not a
   word processor, but a text editor.

   MOKE can also aid in the entry of Kanji. When writing japanese I frequently
   must consult a dictionary (or my wife) to find out where the kanji stops
   and the kana starts. Many times I know the onyomi of the kanji. If the
   kanji is displayed in on the screen and the cursor is below it by using
   the F7 key all the on/kun yomi's of the kanji will be displayed in the
   ENTERLINE line. By using the ALT-m combination to mark a place and the
   F7 key the japanese to english translation of the word (if found in the
   dictionary) will be displayed as well as the yomikata of a kanji word.
   It also has the ability to search for a word given one of the kanjis
   in that word.

   As the UNIX "vi" editor is a mode editor so is MOKE. Perhaps the major
   reason for this is the need to enter Japanese kana and kanji. Though
   it doesn't necessitate it. MOKE has two modes: CURSOR mode, and INSERT
   mode. In CURSOR mode all movement in the file is done, as well as editing
   of existing text. In INSERT mode all entering or inputing of new text is
   done.

   MOKE may be freely distributed. If you pass the program along, please
   include all documentation, help files, dictionary files and executable
   files. The only connection that this program has with the University of
   Wisconsin is that I am a full time employee there.

   Section 2.0 states the System Requirements and Section 3. details the
   installation of MOKE. While the remainder of this document describes the
   operation of the editor. If you have any questions, problems, bug reports,
   or other please report them to me.


     mark edwards  [ $B%^!<%/!#%(%I%o%:(J ]
     29 Sandy ct.
     Madison, Wi. 53717 (USA)
     Phone: [phone]

     Email:
         Internet: [e-mail]
         Bitnet:   [e-mail]

Note:  MOKE and its utility programs have more than 5000 lines of code. The
       code is still undergoing development. MOKE the editor was started
       around the middle of January 1990. Being that this code is only
       around 3 months old it is not surprising that MOKE still has some
       bugs. Most of the bugs while being bothersome are harmless. Some
       bugs, however, may lock up your PC causing loss of text. So please
       insure that you have adequate backups and that you save the current
       buffer every once and a while. (If you keep me informed of bugs
       you find I'll keep everyone else informed, if not fix them as
       soon as the bugs surface.)

2.0   System Requirements

    -    IBM PC (XT, AT, 386, PS/2, or clone)
    -    MSDOS 3.1 or higher
    -    512k memory or more (MOKE uses 230k on startup)
    -    Graphics Adapter (Hercules, CGA, EGA, VGA)
    -    Harddrive or one high density floppy disk drive
         (depending on which Kana to Kanji dictionary used)

  Note: Because MOKE uses bitmapped fonts the faster your PC the better
        the response time will be. The reason for this is that for each
    Japanese Character a 16x16 matrix must be loaded into the screen
    buffer bit by bit. Thats 256 bits for each character. In contrast
      the characters for the normal text mode of your machine are 8x8
    and are displayed by special hardware in your machine.


    Useable lines for Japanese

        CGA  - 7
        HGC  - 14
    EGA  - 20
        VGA  - 20


3.0   Installing MOKE

   MOKE is fairly easy to install, but since it uses many files
  the installation is not trivial. There are several data files
  including the Kana/Kanji bitmaps, the SKK or Wnn jishou, the
  indexes for the SKK and Wnn jishou, the reverse lookup file for
  kanji on/kun yomi, and the japanese to english dictionary, and
  also has kana yomikata for some japanese words. For ease put all
  the data files  in a directory called "\kanji". See section 3.1
  for a list of data files for more information.

  Make sure you go through all the 3.x sections for proper installation
  of MOKE.


3.1   Data Files

    The following is the list of data file for MOKE, the distribution
    you receive may not have all these files. The files that are marked
    with a plus are necessary for MOKE to work as a Japanese Editor. The
    ".ind" files can be built from the Utility programs.

       FILE              Description

  +  \kanji\k16jis1.fnt    - 16x16 JIS Level 1 bitmap fonts (from KD dist)
  +  \kanji\k16jis2.fnt    - 16x16 JIS Level 2 bitmap fonts (from KD dist)
     \kanji\k14bdf       - 14x14 JIS Level 1 & 2 bitmap fonts (from X-window)
     \kanji\k16bdf       - 16x16 JIS Level 1 & 2 bitmap fonts (from X-window)
     \kanji\k24bdf       - 24x24 JIS Level 1 & 2 bitmap fonts (from X-window)
        (k14, k16, k24 are still under development and are not being
         release yet)

  +  \kanji\skkjis      - Simple Kana to Kanji dictionary
  +  \kanji\skkjis.ind       - index into skkjis
     \kanji\wnndict      - Wnn and SKK combined dictionaries
     \kanji\wnndict.ind   - Index into wnndict dictionary

     \kanji\mydict      - Alternate Dictionary to search for Kana to kanji
                lookup

     \kanji\revtab      - reverse lookup index to skkjis
                (to display on/kun)
     \kanji\wsktok      - combined Wnn & SKK Kanji to kana dictionary
     \kanji\wsktok.ind      - index into above file
     \kanji\skkktok.dat      - SKK Kanji to kana dictionary
     \kanji\skkktok.ind      - index to above file

     \kanji\edict      - japanese to english dictionary

     \kanji\moke.rc      - startup file for MOKE


Note:    The hashtab file from MOKE 1.0 is no longer used. Please
        remove that file from your disk if it is there.

3.1.1  Data Files and Configuration

 A.    When disk space is at a premium (ex. floppy based systems)

         skkjis, skkjis.ind

       On the command line specify the -s option or use the "dictionary"
       parameter in "moke.rc".


 B.    When you have plenty of disk space

       wnndict, wnndict.ind, wsktok, wsktok.ind, revtab, edict

       This is the default and my normal configuration.


3.1.2  Making the index files

   The most important data file for MOKE is the kana to Kanji dictionaries,
   "skkjis", or "Wnndict" . I would like to distribute the Wnn dictionaries
   and skkjis dictionaries as is but I haven't found a suitable  sorting
   program. I may have to write my own. In the mean time I am distributing
   a sorted version of each dictionary. If the index file to these files is
   not included with the distribution you receive you can make it by running
   the "wnn_ind" program.

     wnn_ind \kanji\skkjis  (or \kanji\wnndict)

   yields the file "wnn_ind" then copy and rename the file appropriately.

    copy wnn_ind \kanji\skkjis.ind (or \kanji\wnndict.ind)


  Making "revtab" for skkjis

   revtab  \kanji\skkjis
   copy revtab \kanji

  for wnndict

   revtab \kanji\wnndict
   copy revtab \kanji

3.2  Graphics Driver

    The Graphics Driver file (BGI) is located by default in current the
   directory or by the environment variable "bgi". For example if the bgi
   graphics interface files are located in the \kanji directory

3.3  Moke.rc

    Like the Unix VI MOKE also has a startup file where additional options
   can be specified. Some of the options are similar to the switches on
   the command line, but provide an easier way to start MOKE up rather
   than specifying each on the command line. Each of the option should be
   on a single line. "moke.rc" should go in the \kanji directory.

   Option        Possible values (The default value are listed first)

   filetype        Unix/MSDOS
   codetype        EUC/JIS
   graphicstype        HGC/CGA/EGA/VGA
   drivepath        <drive:path> or <path> (the place where all the data
                    files are located. Default is \kanji)
   dictionary        SKKJIS/WNNDICT     (wnndict is combination of skkjis & wnn)
   edict        Yes/no    (If using the Jdict functions and have a edict)
   mydict        Yes/no  (default is no)
   fonthalfsize        9 for readibility (default is 8)
              The font size is 16x16 for kanji, 8x8 for ascii

3.4  Autoexec and config.sys

     Since MOKE uses many data files you'll need to insure that DOS
   can open several files and for optimum preformance make sure there
   are enough buffers. Put the following in your config.sys file


   [config.sys]

   Files=20
   buffers=20

   MOKE uses the bgi graphics drivers from Borland. For moke to load the proper
   driver it must know where to look for it. Add the following line in your
   autoexec.bat file so MOKE can find the bgi files.

   [autoexec.bat]

   set bgi=c:\kanji     (or whatever the path is)



4.0   Using MOKE

    See Section 3.0 for information on how to install moke
   before trying to start MOKE. MOKE can be started up at the MSDOS
   prompt by simply typing "moke".    You can also ask it to load a
   file into the buffer upon startup as shown below.

    moke myfile

   MOKE can be given some arguments (or parameters) on the command line.
   The arguments include "-j" to load in a JIS compatible file or "-e"
   for a Extended Unix Code (EUC) compatible file.


4.1  MOKE's Screen

   MOKE's screen is currently divided into three parts, the editing window,
   the STATUS line, and the ENTERLINE. The editing window or screen is
   where all editing, and displaying of the buffer takes place. The STATUS
   line displays information useful in using the program. It tells you the
   current mode (INSERT or CURSOR), the current INSERT mode state (hiragana,
   katakana, or ascii), and a one or two word help for some of the command
   keys or function keys.

   The ENTERLINE is a general purpose line where the entering of romaji or
   ascii takes place, displaying and selecting of the Kanji Choices for a
   particular romaji string, the displaying of on/kun yomi of a single
   Kanji, and the display of the english translation of words found in the
   japanese to english dictionary as well as the yomikata of that word.


4.2 Getting Help in MOKE

   Help is available in both CURSOR and INSERT mode. To get a brief display
   of the available commands for each mode press F1. For more detailed help
   press "h". To continue on in the current mode press any other key.

   The following is a list of the current help topics.

   1.  Entering Hiragana
   2.  Entering Katakana
   3.  Entering ascii
   4.  Entering JIS ascii
   5.  Entering Kanji
   6.  Entering Punctuation etc.
   7.  Kanji Guess Lookup
   8.  English to katakana lookup
   9.  Entering Tutorial #1
  10.  Verb Conjugation


 You may enter additional help topics up to Twenty. Twenty is the number
 of topics that will fit in a single column. This interface is still
 primitive. There will be enhancements to this in the future.

4.2.1  The help.hlp file

  A help file is potentially any EUC file. Most of the current help files
  were derived from this document. The format of "help.hlp" is as follows.

  <help file name>.hlp  <40 character or less description of the help>

  Example:

      tut.hlp         Entering Tutorial #1


5.0 Quick Tutorial

   To run program type:

     moke <filename>

   where <filename> is an optional argument. If <filename> is supplied
   it will read the file into the buffer.

    It will open various files and load the kanji fonts. If everything goes
   right the text in the file will be displayed on the screen with the
   the cursor ('^' for ascii characters, '^^' for japanese characters).
   Initially you are in the CURSOR mode. In the CURSOR mode you can move
   about in the file with the arrow keys. To enter something type i (for
   INSERT MODE). In order to exit moke you must be in CURSOR mode. To
   change INSERT mode to CURSOR mode press the ESCAPE key. To exit press
   F9. If the buffer has been modified it will ask you if you really want
   to leave. If you want to save the buffer type 'n' and use the F10 key
   to save the file (it prompts for a file name). Then press F9 and MOKE
   will exit.

   The rest of this section will take you through entering two lines of
   japanese text.


5.1 Entering a simple Japanese phrase

   Let's see how to enter the following japanese sentence using moke

     $B$"$7$?!"$o$?$7$OEl5~$K9T$-$^$9!#(J

      (Ashita, watashi wa toukyou ni ikimasu)

   Start moke with no filename and we will then have a fresh buffer to
   play around with. The first word to enter is $B$"$7$?(J. To enter
   we must first go to the INSERT mode. Typing an 'i' changes the mode to
   INSERT mode. When you pressed the 'i' key the "Enter:" prompt was
   displayed on the ENTERLINE or the last line of your screen. Now to enter
   $B$"$7$?(J simply type "ashita" using romaji as shown below.

      Enter: ashita

   Press return and the hiragana for ashita appears on the first line
   of the screen. The comma can be entered as is. Enter it and press
   return. Next enter "watashi", press return. The "wa" must be entered
   like it appears in hiragana that is $B$O(J (ha), press return.
   Now we get to the first kanji word in the sentence. Enter toukyou
   just as you would any other word but instead of using the return key
   to send the line to the screen use the F2 function key. The romaji
   string toukyou has only one mapping in the SKK jishou so it is displayed
   immediatedly. Had there been more than one mapping the choices would have
   been displayed in the ENTERLINE and selection is done with the arrow keys
   and the return key.

   Now let's finish the sentence. Next is "ni", enter it like "ashita" and
   press the return key. Now we have the last word "ikimasu". Since this
   word uses both kanji and hiragana to represent the word entering it is
   slight different from kanji only words. To lookup the kanji $B9T(J in the
   dictionary enter the romaji for the kanji part of the word then enter
   the first romaji character of the hiragana portion of the word which
   is a "k". But capitalize it. Capitalizing it tells the program that this
   character should be treated differently and should not be translated into
   hiragana. Press the F2 key. Two choice will appear on the ENTERLINE. The
   first choice is the question mark (?) if you select this by pressing the
   SPACE BAR or the return key nothing will be selected. So use the right
   arrow and move the cursor under $B9T(J. Press the SPACE BAR or the
   return key. Lastly enter the hiragana porition of "ikimasu" with a period
   and press the return key.

   If you think you've got it skip the next section otherwise for another
   look at entering continue on.


5.2  A second example


    $B:#!";d$O%+%k%U%)%k%K%"$K=;$s$G$$$^$9!#(J

     (Ima, watashi ha California ni sunde imasu.)

   If you have just finished the first sentence press F5 for a new line.
   Having to press a key other than the enter key for a newline is rather
   kludgy, but that is how it currently works. The first word to enter is
   $B:#(J. On the enterline type "ima" and then press the F2 key. The
   kanji for "ima" should appear. Enter "watashi" the same as "ima" and
   press the F2 key. We want the katakana version of California so we do
   this by surrounding "California" with tildes.

     Enter: ~california~

   The katakana version of California is in the SKKjis, but if we didn't
   know we could check anyways. Press the F2 key. The katakana for California
   should have appeared on the upper screen. Next is "ni". Enter it and press
   return. Now we get to "sunde" which is a verb. The dictionary form of the
   verb is "sumu", so we type in "suM" and press the F2 key.  This time there
   are three possible choices. The correct choice is the 2nd one. Position
   the cursor underneath it and press return. You can finish the rest of
   the sentence by typing in "ndeimasu.".


6.0  INSERT mode in detail

     This section describes the input method of MOKE. Entering text
   in MOKE is done in the INSERT mode. The normal mode of MOKE is
   CURSOR mode or command mode. Several command keys invoke the
   INSERT mode including "i" for insert, "a" for append, and
   "o" for open. To quit INSERT and return to CURSOR mode press the
   ESCAPE key.


Note:  This section primarily is valid only for the SKKJIS dictionary.
       It is also valid for the WNNDICT, however the kanji in WNNDICT
       are listed in other ways also. I haven't done extensive experimentation
       with it though.

6.1  Hiragana

     There are three states for entering text. The normal state is
   hiragana. This state can be changed via the F3 key or via special
   non alpha flags which will be described below. Hiragana is enter
   via romaji. That is you use a ordinary ascii keyboard and enter
   romaji then you press the return key to change the romaji into
   hiragana. The hiragana state accepts only romaji and a few punctuation
   characters. You can enter as few as one romajinized hiragana or
   a whole line of romajinized hiragana. DO NOT USE SPACES to separate
   words.

   Example:

     Enter: gohan              $B$4$O$s(J


      For words like $B$7$C$Q$$(J (shippai) or $B$$$C$?$i(J
   (ittara) use "t-" ($B$C(J) to denote that the next consonant
   is long or that it is doubled.


   Example:

      Enter: shit-pai     $B$7$C$Q$$(J

      Enter: it-tara    $B$$$C$?$i(J


6.2  Katakana

     Katakana is the second state for entering text. Katakana words
   are entered similar to hiragana words. That is with romajinized
   katakana. To enter katakana either press the F3 key until katakana
   is listed under the INSERT mode on the status line or toggle the
   KATAKANA flag by first typing the "|". To end the katakana state
   type another "|".

   Example:

     Enter: |programming|         $B%W%m%0%i%_%s%0(J

   The long vowel sound is denoted by the "-" character.

   Example:

     Enter: |konpyu-ta|        $B%3%s%T%e!<%?(J


6.3  Ascii

    Ascii is the last state for entering text. In the Ascii state
   any printable Ascii character can be entered. The Ascii state, like
   the katakan state, is invoked either with the F3 key or with a Ascii
   flag, the Ascii flag. The Ascii flag is the tilde character ("~") and
   is used in the same manner as the katakana flag.

   Example:

     Enter: ~Example~            Example


   Since the "~" and "|" characters are special control characters they
  cannot be entered normally. To enter these two characters precede them
  with the backslash ("\") character. The backslash will escape the
  normal meaning of these characters. Using the backslash with any other
  character will enter the backslash and that character.


6.3.1 JIS Ascii

   While in INSERT mode press the F3 key (the k/h/a key) until "jascii"
   is displayed under INSERT mode. The English Alphabet in lower and upper
   case can be entered. The characters will be the same size as hiragana,
   katakana, and kanji.

   Example:

  $B#A#B#C#D#E#F#G#H#I#J#K#L#M#N#O#P#Q#R#S#T#U#V#W#X#Y#Z(J
  $B#a#b#c#d#e#f#g#h#i#j#k#l#m#n#o#p#q#r#s#t#u#v#w#x#y#z(J


6.4  Entering Kanji

    To enter Kanji start by entering the yomikata of the kanji in
   romaji. Enter the yomikata for a single kanji or the romaji string
   that represents a combination of two or more kanji. This time instead
   of pressing the return key in order to change the romaji into kana,
   press the F2 key. If the SKK Jishou only maps to one possible Kanji(s)
   then it will be displayed directly in the window. If the string maps
   into more than one possible Kanji then a list of possible kanji will
   be displayed in the ENTERLINE.

    Use the left and right arrow keys to move the cursor to the correct
   Kanji. If the Kanji you desire is not listed in the ENTERLINE use the
   up and down arrows to see if any more possible kanji are available.
   When you have moved the cursor underneath the correct kanji press the
   SPACE BAR or the return key to select that KANJI. The selected kanji
   will then appear at the end of the current line in the window. If the
   Kanji you want is not listed in any of the lines move the cursor
   underneath the question mark ("?") and press the SPACE BAR or the
   return key.


6.4.1  Entering Kanji only words

    In this section I will detail how to enter the yomikata of a kanji
   and then select the desired kanji from the possibilities list. The
   kanji I wish to look up is $BJs(J . The yomikata for this kanji
   is ho- ($B$[$&(J). So I type:

    Enter: ho-

   and then press the F2 function key. The ENTERLINE is erased and the
   following list of kanji is displayed.


    $B$[$&(J          ?  $BJ}(J $BK!(J $BJs(J $BJ|(J $BK.(J $BK"(J $BK%(J $BK$(J $BJq(J $BJz(J $BK-(J $BJo(J $BJp(J $BJr(J $BJ~(J $BJw(J

   The kanji I wish to enter is the third kanji from the left. I use the
   right arrow key and move the cursor underneath it. Then I press the
   return key. If I had pressed the down arrow the following list would
   be displayed.


    $B$[$&(J          ?  $BK,(J $BJx(J $BK2(J $BJu(J $BK&(J $BK)(J $BK'(J $BJ{(J $BJy(J $BJv(J $BJt(J $BK#(J $BK((J $BK*(J $BK+(J $BK/(J

   If I had pressed the down arrow key again the remainder would have been
   displayed.

    $B$[$&(J          ?  $BK0(J $BK1(J


6.4.2   Entering Verbs and Adjectives with a Kanji part

    Entering verbs and adjectives is some what different from entering words
    that are all kanji. How verbs are entered in MOKE relys much on how the
    the entry appears in the skk dictionary. Verbs that have a kanji part and
    a hiragana part are looked up in the skkjis with hiragana for the kanji and
    an ascii character representing the start of the hiragana part of the verb.
    For the verb $B8+$k(J (miru) it is listed as $B$_(Jr,  $B=q$/(J is listed
    as $B$+(Jk, and $B;W$&(J is listed as $B$*$b(Ju. For MOKE just
    capitalize the character representing the start of the hiragana part of the
    verb or the 'r', 'k', and 'u' of the above verbs. Below is a couple of
    examples and how they are to be entered.


      kanji           romaji       enter      displayed

       $B=q$/(J       kaku         kaK        $B=q(J

       $B8+$k(J        miru           miR        $B8+(J

       $B;W$&(J        omou         omoU       $B;W(J

       $B?)$Y$k(J    taberu          taB        $B?)(J

       $BBT$D(J      matsu        maT        $BBT(J

       $B=P$9(J      dasu         daS        $B=P(J

       $B?7$7$$(J    atarashii    ataraS     $B?7(J

       $BH~$7$$(J    utsukushi    utsukuS    $BH~(J

       $BBg$-$$(J    ookii        ooK        $BBg(J

       $B9b$$(J      takai        takaI      $B9b(J

       $BDc$$(J      hikui        hikuI      $BDc(J


6.4.3    Entering Katakana versions of English words.


    The skkjis dictionary also has some english words. Try computer, path,
    program, programming, etc (computer related words). For more words
    do a strings on skkjis. Enter english words as in rtk.

    Example:

    Enter: ~computer~


6.4.4  Guess Kanji command

    Many times you will know only one kanji of the word. In this case
    MOKE can assist you in finding the right word. To use this function
    type in the known kanji and use ALT-g (for guess). It will act similar
    to the F2 command and ask you to choose a kanji if the kana maps to
    more than one kanji. Select the desired kanji as you normally do. Next
    you'll get the following message.

        Which kanji in the word (1)st, 2)nd, ...):

    Enter the position number of the known kanji (Example: 1). The disk
    will spin and if the kanji is found (in the wsktok.dat dictionary)
    it will display one entry. To go to the next entry hit any key
    except ESCAPE or RETURN. To select the entry hit RETURN. To Exit without
    selecting anything press ESCAPE.

    Of course this function is only as good as the dictionary itself. If the
    word is not in the dictionary you won't find it. You should note that
    searches on the first kanji of the word is much faster then searching on
    subsequent kanji. This is because the dictionary is sorted on the first
    kanji and therefore the program can jump to the general area of that
    kanji. An other kanji in the word must necessarily search the entire
    dictionary and the dictionary is over 400k.


6.4.5  English to kanji

    Kanji characters and compounds can be entered via english also. This
    presupposes that the english already exists in the edict dictionary
    however. The english word(s) is entered either in the ascii state of
    the INSERT mode or by surrounding the word with "~" (tildes) and then
    pressing ALT-e (for english lookup) to lookup the word.

    Example:

        ~idea~

    To go to the next entry hit any key except ESCAPE or RETURN. To select
    the entry hit RETURN. To Exit without selecting anything press ESCAPE.

    To find any match the entire dictionary is searched.

    Note: The pattern matching is currently for any string in a particular
      english entry in edict. This means that "plan" will also return
      explanation. (This will be fixed in upcoming releases).


6.5  Punctuation

    Some japanese punctuation character are also available. Many of the
    punctuations are available by name. (In latter versions of documentation
    I'll try to include a list)


       Punctuation     Enter      Description

    $B!V(J      `        kagikakko
    $B!W(J      '         kagikakko
        $B!t(J      #
        $B!)(J      ?
        $B!#(J      .       period
        $B!"(J      ,          comma

6.6  Other

   Greek and Russian and other misc characters are available,  and various
   other things also by japanese name.  Try "shikaku", "maru".....

   You can use moke with the "fonttest.jis" file to find out how to enter
   special characters also. The fonttest.jis file is part of the kd1.0
   distribution.


7.0  Kanji to kana/english lookup

   The readings (yomikata) of Kanji characters as well as Kanji words can be
   looked up by MOKE. The possible readings of a single Kanji character may
   be looked up via the F7 key in Cursor Mode. The readings if found will be
   displayed on the ENTERLINE. To lookup a Kanji word the start of the word
   must be marked with ALT-m. Then move the cursor to the end of the word.
   The word will become underlined as you move the cursor. Use the F7 key
   here to display the reading (if any).

   If edict is "yes" in moke.rc first the edict file will be searched and
   the reading along with any english translations for the word will be
   displayed. Press ESCAPE or RETURN to resume CURSOR MODE or any other key
   to continue searching the file for additional readings.


   Note: To get the english translation of a single kanji character you must
         always mark it with ALT-m. Using F7 without ALT-m will always search
         only the "revtab" dictionary.


7.1  Entering English translations while in the editor

    Brute Force

    English can be entered via the ALT-d (for dictionary entry) key. Entering
    is done as usual but to lookup a word use ALT-d instead of F2. Select the
    kanji word if there are multiple mappings and if not it will be displayed
    in the following format


    <kanji word> [<hiragana yomikata>] /

    The ENTERLINE is now ready to accept the english (ascii mode). Enter
    individual translation separated by "/". To end close the last translation
    with "/" and "~". This changes the mode back to hiragana. Press F5 to
    go to the next line.

    While Editing a Text

    The english translation for a kanji word can also be entered while
    editing a text. To do this lookup a kanji word marking it with ALT-m
    and instead of using the F7 key use F8. If there is no english it
    will ask you for the english. Even if there is english for this word
    it is possbile that there are multiple readings. Hit any character
    (except for ESCAPE or RETURN which will bring you back to CURSOR MODE)
    to get the next reading (I usually press the Space Bar).

    When entering english you'll get the prompt

     English (1):

     then

     English (2):

    Enter however many english translations there are pressing return to
    go to the next one. When no more press ESCAPE on right after presing
    RETURN.


8.0  MOKE Command Keys


   This is a visual line editor. I looked at patterning commands after
   ed, ex or vi, but choose it to be vi like. If you don't like the letters
   I choose you can change them in the moke.h file and recompile the
   program.

   Like vi when a line is deleted it is copied to a buffer. This buffer
   can be put (p) anywhere you like. Another delete line will destroy
   the current buffer and again make a copy of the line just deleted.

   Copying a line, like vi, can be done with the yank line command (y).


8.1  CURSOR MODE keys

   a - append  (insert to the left of cursor)

   d - delete current line

   i - insert  (insert to the right of cursor)

   o - open line for editing (insert line after currentline)

   p - put last line deleted or yanked below current line

   x - delete cursor character

   y - yank line (copies line into buffer)

   j - join lines (currentline with next line)

   D - clear from cursor position to the end of line

   b - goto beginning of line

   e - goto end of line

   g - goto line number (prompts for the line number)

   / - search forward from currentline

   ? - search reverse from currentline

   n - search forward next

   N - search forward reverse

   ALT-g     report current line of how many (not implemented yet)

   ALT-r    insert file after current line

   ALT-w    write current file (does not ask)

   ALT-z    write current file and quit.

   F1 - Help for CURSOR mode

   F6 - toggle between EUC and JIS output

   F7 - on/kun yomi lookup of single/(string of) kanji

   F8 - on/kun yomi lookup of single/(string of) kanji and asks you if you'd
    like to enter the english of the string if it is not found in the
        edict dictionary.

   F9 -  Quit (If buffer has been modified it asks if you wish to save file)

   F10 - Save file (always asks you the file name to save to)

   ALT-M -  Used with the F7 key to look up a japanese string.


8.1.1  Searching


   MOKE can primatively search forward and backwards. To search forward
   use the "/" key and to search backwards use the "?" key. When you
   press the forward search in the ENTERLINE "Fsearch: " will appear
   type in the Japanese (or english) string to search for. You can use
   the F2 key to get kanji. The search string as you build it will appear
   on the enterline also. You can combined kanji, kana, english.

   To find the next occurrence of the string press "n", to search reverse
   use "N".


8.2  INSERT mode keys

   F1 -  Help for INSERT mode

   F2 -  lookup word on enter line

   F3 -  toggle between hiragana/katakana/ascii input modes
         (default is hiragana)

   F4 -  copy last string entered in screen window to enterline
     (erases anything currently displayed in the enterline)

   F5 -  Start Newline (very kludgy)

   bs  - (back space) erase one character on enter line

   q   - quit and save

   e   - quit (doesn't ask to save)


   ALT-d - like F2. For input of dictionary words for the Japanese->english
           dictionary project.

   ALT-e   Looks kanji by way of english using the EDICT dictionary

   ALT-g - guess kanji (used like F2 or ALT-o, if you know one of the kanji
       in a word use this and MOKE will display possible kanji using
       the kanji you give it.)

   ALT-o - search other dictionary (\kanj\mydict) for word


9. Future Enhancements

   Some future enhancements may include getting rid of the entering only
   on the enterline and do the kana/kanji editing inline, and possibly
   multi-buffer capability.


10. Utility Programs

   jc    - Converts Japanese files between the various methods of coding
          EUC, JIS, SJIS

      Source of the code is Japan but I don't have anything more then that.
      The program guesses what form the input file is and changes it to
      the format that you request.


        Syntax:       jc <-e, -j, -s> <input file> > <output file>

             -e     some format to EUC
              -j    some format to JIS
             -s     some format to SJIS


    Note: Use the output redirection switch (>) to direct output to the
          output file name.


   jprint       prints Japanese/ascii text on a 24 pin Epson compatible Printer.
        Jprint uses the 16x16 fonts that are distributed with MOKE.

        Syntax:    jprint  <-s, -L, -9, -in> <filename>
                     -s   suppress FF,
                     -L   Letter Quality (24x24)
                     -9   Epson 9 pin compatible mode (16x16)
                      -in  indent 'n' amount

        The file can be either JIS/EUC coded.


   revtab    Creates the Reverse lookup table. This is for the lookup of
            of the on/kun yomi of an individual file. There are 7305
        kanji entries. (Creates a file named "revtab" size 224000
        bytes. Copy the file to \kanji when done.

        Syntax:  revtab skkjis

          or

             revtab wnndict

    wnn_ind    Make index file from wnndict or skkjis

    ktok_ind    Make index file to wsktok.dat or skkktok.dat


11. Special thanks and Acknowledgements

     Special thanks to my wife Ritsuko [ $BN';R(J ] who learned
     well a new english phrase and the meaning, computer widow.
     You see since receiving the Junet Reader program middle December
     I've bought a faster PC for home, bought Minix 1.3 and am trying
     to update it to 1.5.6 with cdifs posted to the net, play with our
     first baby (almost eight months old now, Akane [ $B0+(J ]), and
     write all these programs.

     A big thank you to Bill Weaver ([e-mail]) for many valuable
     suggestions in debugging and enhancing MOKE.

11.1 Acknowledgements

   This program would not be useful without the SKK dictionary provided in
   the SKK interface to Nemacs by masahiko sato or the Wnn dictionaries
   provided with from Kyoto University.

11.1.1 SKK

    Below is the first few lines from the SKK manual.

    $@$+$J4A;zJQ49%7%9%F%`(J SKK (Simple Kana to Kanji conversion system)

            $@ElKLBg3XEE5$DL?.8&5f=j(J $@:4F#(J $@2mI'(J
            ([e-mail])

SKK $@$O(J NEmacs $@>e$GF0$/4JC1$J$+$J4A;zJQ49%7%9%F%`$G$"$k!#0J2<$G$O(J SKK $@$NFC(J
$@D'$*$h$S;HMQK!$K$D$$$F=R$Y$k!#(J

11.1.2  Wnn

   Below is the Copyright notice with Wnn.


    Wnn Version 4.0

This distribution contains Version 4.0 of Wnn Japanese Input
System.  Wnn is a network-extensible Kana-to-Kanji conversion system
and was jointly developed and released by the Software Research Group
of Kyoto University Research Institute for Mathematical Science, OMRON
TATEISI ELECTRONICS CO., and Astec, Inc.  If you have Wnn up to
Version 3.3, refer to "manual/intro" for revision information.

This distribution has been tested on the following systems.
    Sun-3 Sun-4,  SunOS3.4, SunOS4.0
    OMRON SX9100, UniOS-B (4.3BSD) UniOS-U (SystemV R2.1) (DT DS M90)
--------------------------------------------------------------------------
/*
    Copyright
        Kyoto University Research Institute for Mathematical Sciences
        1987, 1988, 1989
    Copyright OMRON TATEISI Electronics, CO. 1987, 1988, 1989
    Copyright ASTEC,Inc. 1987, 1988, 1989


   Permission to use, copy, modify, and distribute this software
   and its documentation for any purpose and without any fee is
   hereby granted, subject to the following restrictions:

   The above copyright notice and this premisson notice must appear
   in all version of this software;

   The name of "Wnn" may not be changed;

   All documentation of software based on "Wnn" must contain the wording
   "This software is based on the original version of Wnn developed by
   Kyoto University Research Institute for Mathematical Sciences (KURIMS),
   OMRON TATEISI Electronics, CO. and
   ASTEC,Inc.", followed by the above copyright notice;

   The name of KURIMS, OMRON and ASTEC may not be used
   for any purposes related to the marketing or advertising
   of any product based on this software.

   KURIMS, OMRON and ASTEC make no representations about
   the suitability of this software for any purpose.
   It is provided "as is" without express or implied warranty.

   KURIMS, OMRON and ASTEC DISCLAIM ALL WARRANTIES WITH REGARD TO THIS
   SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS,
   IN NO EVENT SHALL OMRON, ASTEC and K.U.R.I.M.S BE LIABLE FOR ANY SPECIAL,
   INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM
   LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE
   OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
   PERFORMANCE OF THIS SOFTWARE.

   Author: Hideki Tsuiki    Kyoto University
                [e-mail]

        Hiroshi Kuribayashi    Omron Tateisi Electronics, Co.
                [e-mail]
                uunet!nff.ncl.omron.co.jp!kuri

        Naouki Nide        Kyoto University
                [e-mail]

        Shozo Takeoka    ASTEC, Inc
                [e-mail]

        Takasi Suzuki    Advanced Software Technology & Mechatronics
                Research Institute of KYOTO
                [e-mail]

*/


Appendix A     Alternate Dictionary (mydict)


 MOKE will search an alternate dictionary if the mydict variable is set
 in moke.rc. The dictionary form is similar to the SKKJIS or WNNDICT
 format.

 FORMAT

  <hiragana/english> /<kanji/katakana>/<repeat as many times as needed>/

 Examples

   mark /$B%^!<%/(J/
   edwards /$B%(%I%o!<%9(J/
   $B$j$D$3(J /$BN';R(J/

 The dictionary should be in \kanji and should be named "mydict".
 Be sure to turn the Alternate dictionary on in the startup file
 "moke.rc". The following line is all that is needed.

    mydict     yes


Appendix B     Dictionary of Technical Words


Term        Meaning

baseline    The horizontal line where characters without descenders
        rest. An example of a character with a descender is p.

bitmap        a bitmap in its simplistic form is essentially a two
        dimensional array of pixels or bits where an image is formed
        depending on the binary on or off state of each pixel in the
        array.

bitmap font    uses pattern of pixels for each character. With bitmap
        fonts there exists a different set of pixels for each
        character in every fonts size you want.
        Sometimes called raster

buffer        Area in the computers memory used to store what the computer
        is currently working on.

EUC        Extended UNIX CODE

font metric     The physical characteristics of each character in a
        font including height, width, .

glyph        a symbol that conveys information nonverbally. (Websters)

graphic font    See vector font.

hiragana    One the two kanas. Hiragana is used to write japanese words
        and to decline japanese verbs, put endings on adjectives,
        etc. Characterized by more rounded and artistic strokes.

JIS        Japanese Industrial Standard

kana        The alphabet like symbols used with Kanji to represent the
        Japanese langugage. There are two parallel systems hiragana
        and katakana.

katakana    One of the two kanas. katakana is usually used to represent
        foreign words, but not always. Characterized by straighter
        strokes.

kanji        The Chinese characters used with the kanas to represent the
        Japanese characters. There are around 2000 basic kanji with
        an additional 5000+ other kanji. Some of the kanji were
        greatly simplified after WWII.

kerning        adjusting the amount of space between letter.

pixel        the individual elements or dots of the digitized image.
        Derived from "picture elements". Some times called dots as
        in dots per inch.

raster        The area where a picture is displayed on a crt.

raster font     See bitmap font.

raster imaging    The creation of a complete dot-for-dot image of a page
        or graphic. In most cases one dot or pixel represents
        one unit in memory.

resolution    the process or capability of making distinguishable the
        individual parts of an object or closely adjacent optical
        images. (Websters)

        Resolution is usually measured in dots per inch or pixels
        per inch.

        computer displays are typically 50 to 100 dpi
        dot matrix printers are typically 100 to 200 dpi
        laser printers are typically 300 to 600 dpi
        Photographic technology, above 1000 dpi

romaji        Method of writting Japanese with roman characters. It is
        highly ambiguous due to the number of homonyms in the
        japanese language.

SJIS        Shift JIS.

stroke        a mark or dash made by a single movement. One of the
        lines of a letter of the alphabet. (websters)

vector font    define lines that make up characters. Only need one
        set of characters for each font. They can be drawn
        at other sizes by varying the length of the lines.
        Vector fonts can be easily rotated and scaled.

vi        The UNIX visual editor developed by UC Berkeley. Is now
        widely available on other Operating Systems. The version
        of vi called Stevie also has also been patched for the
        Japanese language.

UNIX        Operating System originally developed by AT&T. There are now
        several different flavors (VERSIONS) of UNIX the two most
        popular are V R3 by AT&T and the Berkeley Software
        Development (BSD) now at level 4.4. UNIX is responsible for
        many of the current programming tools like AWK, LEX, YACC.
        But is probably most noted for the C Programming Language.


Appendix B. Related Programs

B.1  Programs to display Japanese on a PC


C.1.1  KD 1.0     - Kanji Driver by Izumi Ohzawa

    KD is probably the best and most powerful program for displaying
    Japanese on a PC. It really is different from the rest of the
    display programs here in that it allows one to display japanese
    from program on a remote computer without having to download the
    file to the PC. In such it is kind of a terminal emulator.

       Izumi Ohzawa     [ $@BgBt8^=;(J ]
    University of California, Berkeley
    [e-mail] or ...!ucbvax!violet!izumi

C.1.2.  Reader - Junet Reader Program by Seiichi Nomura and Seke Wei

    This is the program where Izumi Ohzawa and I got our inspiration.
    In that respect this is probably the most important program of the
    group. It is mainly aimed at reading News from Junet. It does have
    some capabilities of printing Japanese on a Dot Matrix printer.

        Seke Wei
        Institute of Computer Science
        TsingHua University
        HsinChu, Taiwan 30043
        [e-mail]


        Seiichi Nomura
        Associate Professor of
        Mechanical Engineering
        University of Texas at Arlington
        [e-mail]
        [e-mail]


C.1.3.  View   - JIS viewer by me [Mark Edwards]

   I first translated the Junet Reader program into TC 2.0 and then
   further enhanced it to my liking. I improved the speed at which the
   bitmapped fonts were displayed probably over 50% by converting the
   kjfont16.ver file into the form that it is displayed in. This
   program serves as the basis to all my other programs. I also have
   added support for the k14.bdf file that is distributed by X11R4.


C.2.  Programs to print to a printer

   jprint - See above

   As stated in C.1.2. the Junet Reader program has some facilities for
   this. The best program is  Kanjips - JIS to Postscript translater
   by Toshiyuki Masui. His development platform isa Unix machine capable
   of running PERL. PERL isn't available on the PC so I did a quick hack
   to make it work on the PC. It is still buggy, and I plan to work on
   it in the future to make it more usable.


     Toshiyuki Masui
     [e-mail].

C.3  Miscellaneous Programs and Guides


C.3.1  jis.pas  by Ken Lunde

    Ken has written a pascal program for the VAX running VMS to
    convert between the various JIS coding techniques. The program
    was converted to run on the PC by Seiichi Nomura.

    [e-mail]

C.3.1 Electronic Transfer of Japanese by Ken Lunde

    This is a good introduction for sending and receiving mail from
    Japan and how to do it in Japanese. (See above for mail address)

D. Japanese to English Dictionary Project

    The on/kun yomi enhancement to view (and moke) led me to think
    of another enhancement, namely a lookup based on a japanese word
    that would yield english. What would be needed here is a kanji/kana
    to english dictionary much like the kana to kanji dictionary that
    the SKK dictionary provides.

    I am curious to what kind of interest the readers of this news
    group might have for this facility. A workable online dictionary
    would require much effort and is not a one-man job. If there is
    any interest at all I would be glad to collect contributions from
    anyone and then build the interface and distribute the dictionary
    to all.

    Tentatively I think a good submital format would be best patterned
    after the SKK dictionary format.


     Example:

      <kanji> [on/kun yomi] /english one/ english two/ english n/

      miru in kanji [miru in hiragana]  /to look/to see/


      What do you think?

      moke or nemacs or whatever program for editing
      japanese could be used to construct the entries.


E.  Using Japanese on other Platforms

E.1 Unix Machines

    Nemacs
    kterm     - uses kinput and wnn server
    wterm
    kinput
    SKK
    Wnn

E.2 Macintosh

    Japanese System Software available from ADPA (Apple Computer)

E.3 Other

F. Problems

   I have tested MOKE on Hercules and VGA graphics cards. The program
   will not work on a monochrome only machine, and it doesn't work on
  at least some Toshiba portables.


   Bug reports, comments, or to get on the mailing list to receive bug reports
   and bug fixes.

   Send to:

      mark edwards  [ $B%^!<%/!#%(%I%o%:(J ]
     29 Sandy ct.
     Madison, Wi. 53717 (USA)
     Phone: [phone]

     Email:
         [e-mail]
         [e-mail]