IMERJ (IME/REXX for Japanese) Version 2.0.2 IMERJ is a simple input method editor for Japanese text, written in VX-REXX. IMERJ is intended primarily for use on non-Japanese OS/2 systems where no better method for typing Japanese is available. Users of native Japanese versions of OS/2 will probably not find it very useful, as the built-in IME is considerably more sophisticated. REQUIREMENTS IMERJ has only been tested on OS/2 version 4.5 and up. In theory, it should be usable on Warp 4 with FixPak 5 or later, but I cannot guarantee that. IMERJ requires the following to run: - VROBJ.DLL version 2.1d or later (available on Hobbes). - RXULS.DLL version 0.5.0 or later (available on Hobbes or my website). - VROBJEX.DLL version 1.3.2 or later (available on Hobbes or my website). - A Japanese font, or a Unicode font with Japanese support, if you want characters to display correctly. (Times New Roman MT 30 will suffice.) (The WarpIN package can install both RXULS and VROBJEX if necessary. VROBJ is already installed on almost all modern OS/2 systems.) The kanji conversion has the following additional requirements: - The EDICT dictionary (in classic format) or a compatible dictionary must be installed and configured under Options. - The KANJIDIC dictionary (or another dictionary in compatible format) must be installed and configured under Options. - The OS/2 codepage tables for Japanese (in particular, the file \LANGUAGE\CODEPAGE\IBM954) is also required, as this is used to translate to and from the EUC encoding used by the above dictionaries. If these files are not present, kanji conversion will be disabled. If the codepage table IBM954 is not already installed, you can install it manually by going into the directory \OS2IMAGE\FI\FONT on the OS/2 (or eComStation or ArcaOS) installation CD/DVD, and unzipping the file JPCODEPG.ZIP into the \LANGUAGE\CODEPAGE directory on your boot drive: unzip jpcodepg.zip -d C:\language\codepage (Replace C: with your boot drive.) The latest full version of EDICT is available from the public FTP site ftp://ftp.monash.edu.au/pub/nihongo (file edict.gz). However, the full version of EDICT is a very large file (and tends to grow larger over time as it is continually updated). Therefore, I recommend using this version instead: ftp://ftp.monash.edu.au/pub/nihongo/edict_sub.zip which contains 'only' the 22,000 most common words, and runs much faster. As for KANJIDIC, the latest version may be downloaded from: ftp://ftp.monash.edu.au/pub/nihongo/kanjidic.zip In both cases, only the main dictionary files (without extension) are needed; the supplementary files (indexes and documentation) are not used by IMERJ. Place them in any directory, and point IMERJ to their locations (via the Options dialog). IMERJ can also use an 'auxiliary' dictionary (configurable under Options). This is intended for users to be able to add their own words. A sample dictionary, AUXDICT, is provided with IMERJ; this contains around 500 common family and/or place names (since EDICT mostly doesn't include proper nouns). See the program help for details on editing this file. NOTE: The IMERJ WarpIN package includes both EDICT_SUB and KANJIDIC. USAGE Japanese text is generated in Shift_JIS format. Consequently, IMERJ must be run under codepage 932 (or 943) in order for it to be displayed correctly. To run under codepage 932, make sure that 932 is listed in the CODEPAGE setting in CONFIG.SYS (preferably as the second one, unless you're actually running Japanese OS/2). Then do a 'CHCP 932' before running IMERJ, or else (on later OS/2 versions) set the codepage to 932 on the 'Language' tab of the program object properties. Once IMERJ is running, type the desired text in the edit field. In the default input mode, any romaji text you enter (using the so-called "wapuro" style convention for romanizing Japanese) will be converted on the fly into Japanese kana (phonetic syllables). Lower case romaji text will be converted into hiragana, and upper case text into katakana. Prefixing a syllable with 'x' will cause the small version of the kana to be entered, if one exists. There are several alternate input modes available; see the help for details. The current mode is displayed in the status line above the edit window. Once you've typed the text to your satisfaction, you can copy and paste it to wherever it's needed. (Note that your target application should either support Unicode text and clipboard data, or else be running under codepage 932, if you want intelligible results.) LIMITATIONS - Kanji conversion only works for individual words in dictionary form, or single-kanji readings. Clause-based conversion is not possible. Similarly, conversion of conjugated kanji verbs or adjectives is not supported - you must either type the word in dictionary form, or use the "Single kanji" option to convert the kanji one by one. - Kanji conversion of proper nouns is limited, as EDICT does not cover most names, and the default AUXDICT only includes a few common surnames and place names. - Using mouse chord actions to paste text is not supported. (You can copy text this way, but the operation will be handled by the PM MLE control directly; as a result, text will not be copied in Unicode format.) This is a (probably permanent) limitation imposed by VX-REXX. - Some of the more esoteric CJK punctuation marks and graphics characters are not supported by the input conversion logic. You can, however, use my Extended Character Map program (DBCSMAP) to paste such characters in. (DBCSMAP is available from my website or from Hobbes.) - The edit field only operates in "insert" mode. Overstrike mode is not implemented at present. BUGS Almost certainly some bugs exist. Please report any you find, preferably with a description of how to reproduce them. (My contact information can be found on the web site listed at the end of this file.) HISTORY 2.0.2 (2022-12-21) - UI (not edit) font is now set to the PM default Window Text font, and the UI is rescaled as needed to fit. - Minor help file updates 2.0.1 (2019-03-01) - Kanji selection menu now reflects main window font changes. - Fixed several errors/typos in the help file. 2.0 (2018-04-14) - In light of the major new features, the version has been raised to 2.0. - Kanji conversion is now supported (with the limitations described), as long as EDICT/KANJIDIC are installed and configured. - Line breaks are now supported in the edit field. Multi-line clipboard text may now also be pasted in its entirety. - Added an Options dialog. - Added context menus and a toolbar. - Online help is now available. - Undo is now implemented (last action only). - There have been many fixes and improvements to romaji input conversion. - The commands for switching between conversion modes have been overhauled. - IMERJ is now licensed under version 3 of the GPL. 0.9 (2010-04-24) - Small 'tsu' is now entered correctly when in halfwidth mode. - The Ctrl+F keyboard shortcut should now work even with CapsLock on. - Added support for additional punctuation, in particular the ideographic quotation marks, and the middle-dot kana. Some more ASCII fullwidth equivalents (tilde, single-quote) are now also supported. 0.8 (2008-09-10) - Pressing 'x' before entering a kana will produce the small version of said kana, if one exists. 0.7 (2008-01-09) - Pressing Backspace or Del will cancel any partially-typed kana. - The "text/unicode" clipboard format is now supported (using RXULS.DLL). This format will be used in preference to plain text. 0.6 (2007-12-08) - You can now type an apostrophe (') after N to explicitly enter the standalone 'N' kana; the apostrophe in this case will not be entered. - Fixed a bug where N before S-, Z-, W- and Y-series kana wasn't working properly. - IMERJ now prompts for confirmation if you close the window while there is (uncopied) text in the edit window. 0.5 (not released) - Added support for clipboard cut and paste. - Window size/position, colour and font settings are now saved. 0.4 (not released) - Added support for halfwidth katakana. - Various user interface improvements. - Support use of 'M' for 'N' when typed immediately before a consonant. 0.3 (not released) - Added support for fullwidth ASCII. 0.2 (not released) - Added menubar and product information dialog. 0.1 (not released) - Initial build. LICENSE IMERJ (IME/REXX for Japanese) Copyright (C) 2007, 2019 Alex Taylor This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see . AUTHOR Alex Taylor http://www.altsan.org