JGloss

JGloss is an application for adding reading and translation annotations to words in a Japanese text document. This can be done automatically and manually. When a text document is first opened, words will be looked up in dictionary files and the first reading and translation (if any) is used to annotate the word. The user can then edit the annotations: choose among the readings and translations found in the dictionaries, enter own readings and translations, remove annotations and add new annotations. The document can be exported as plain text with annotations, HTML (with support for the Ruby Annotation XHTML specification) or LaTeX.

The application is designed as a translation aid for people learning Japanese. With some new document, you can first skim the text and change the annotations to match the likeliest meaning of the word. Then you can print/export the text with annotations and start working on the details of understanding the text without having to resort to a paper dictionary all the time.

JGloss is written in Java and should work on any computer with support for the Java 2 Version 1.3/1.4 platform. It is distributed under the terms of the GNU General Public License.

The latest stable version is JGloss 1.0.10, released on 2004/06/20. You can download it here:

JGloss 2 alpha 3, released on 2004/06/20, is a snapshot of the development currently done on JGloss. It offers greater dictionary functionality and an improved user interface in the document editor. Many features from the JGloss 1.0.x series are still missing in this release, and there may be more bugs than in the stable release.

The dictionary lookup part of JGloss is now also available as a separate application: JDictionary. You can use this if for example you want to offer your language students a dictionary lookup program, but not the (too) easy text parsing of the full JGloss.

JGloss 2 is no longer compatible with Java2 1.3. You need a Java Runtime Environment compatible with at least Java2 1.4.

The archive of previous releases is here.

JGloss Information

• The JGloss User's Guide (PDF).
• More on Java and Japanese fonts.
• More LaTeX export templates.
• The JGloss project page at SourceForge.

JGloss mailing list/Newsgroup

• Mailing list (message archive)
• Newsgroup

The JGloss mailing list is the central discussion forum for all JGloss users and developers. It is open to everyone interested in JGloss. You can post questions about installation and usage, discuss how you currently use JGloss and suggest in what direction the application should be going. To encourage people to post, writing in any language is allowed.

The mailing list is also available as internet newsgroup, thanks to the gmane mail-news gateway. To access the list, point your favorite newsreader to news server news.gmane.org, group gmane.comp.java.jgloss or access the newsgroup with your web browser. All messages posted to the mailing list are also sent to the newsgroup, and vice versa. You do not need to be subscribed to the mailing list to read and post messages via the newsgroup.

Dictionary Files

Here are some links to dictionary files which can be used with JGloss. If you know of dictionaries not listed here, drop me a note.

• A list of EDICT and KANJIDIC dictionaries is available at the Monash Nihongo FTP archive.
• The WadokuJT is an extensive Japanese-German dictionary. Download WadokuJT.zip, unzip it and add the extracted file "WadokuJT.txt" in the JGloss dictionaries dialog. (The WadokuJT file format is only supported by JGloss if Java version 1.4 is installed, if you get a "unsupported dictionary format" error when you choose "WadokuJT.txt", install a current version of Java or download the EDICT formatted version of WadokuJT.)

Related Pages

• Jim Breen's Japanese Page has lots of interesting links. He is the author of several tools on which JGloss is based and one of the authors of EDICT and KANJIDIC.
• The ChaSen morphological analysis tool can be used by JGloss to parse Japanese text. Download it from its homepage.
• Two WWW interfaces which provide glosses for any Japanese web page without having to run an extra program:
  Todd Rudick's RIKAI server and Jim Breen's WWWJDIC.
• The Blue Sky Collection is an online archive of free Japanese texts.
• A Ruby extension for Netscape 7, Mozilla or Firefox enables the display of ruby in these browsers.

Here is an example for a text annotated using JGloss. It is the story "Kari no Douji" by Kenji Miyazawa. I downloaded the original text from the Blue Sky Collection. As you may have already noticed, my English is not perfect, and my Japanese is even worse. The translations are only chosen to allow basic understanding and will often be imprecise or just plain wrong. A website with information about Kenji Miyazawa and his stories is here. An English translation of "Kari no Douji" is available there. Here is the JGloss file from which the following documents where generated.

First an example of the HTML export function. The generated document uses a subset of the Ruby Annotation XHTML specification to display the readings in browsers which support it (currently only recent versions of the Microsoft Internet Explorer). In other browsers, if the neccessary JavaScript functions are available, the readings are shown in a tooltip popup and in the status bar. The translations are also displayed using JavaScript tooltips and in the status bar.

JGloss also supports exporting in LaTeX and plain text format:

• Here is a LaTeX example, TeXed from this source file and converted to PDF with the Ghostscript ps2pdf tool. This is my favorite export format. Here is an alternative version with the translation list at the end of the document (source).
• Plain text output can also be generated. Here is the text with reading and translation annotations and here the text with readings only.
• The list of annotations can be exported to a file.
• Finally, here is a printout of the document, generated via print to file and converted to PDF using Ghostscript's ps2pdf tool.

For a complete list of changes see the changelog.

Version 0.9 is the first public release of JGloss.

Version 0.9.1-0.9.4: refer to the changelog.

Version 1.0: Changes since the last version are mainly bug fixes and user interface cleanup. The formatting of translations in the LaTeX output is now nicer.

Version 1.0.1: The font used in the user interface can now be configured in the preferences, and a suitable font is autodetected at the first run. Editing the font.property files is thus no longer neccessary. A document title can be set which is used for HTML and LaTeX export. Import clipboard now works under X11.

Version 1.0.2: Fixes for compatibility problems with Java2 Version 1.4.

Version 1.0.3: Added support for the Japanese-German GDICT. The EDICT implementation is rewritten to use the J2SE 1.4 NIO API, which improves load time and decreases memory usage. Improved UI feedback when opening documents.

Version 1.0.4: LaTeX export is now configurable through templates. Both pLaTeX and LaTeX-CJK are supported. For long translations, a shorter translation entry is automatically added. Font auto-configuration is fixed. The font and font size of the word lookup result list is now configurable.

Version 1.0.5: Fixed a memory leak which occurred when a document was closed. Added an XML export option and improved paragraph detection on text import.

Version 1.0.6: Improved keyboard navigation and new options in word lookup dialog. Fixed the "access of A drive" bug.

Version 1.0.7: Improved dictionary recognition. Support for EDICT files with UTF-8 (Unicode) encoding.

Version 1.0.8: Fixed dictionary loading and preferences setting.

Version 1.0.9: Fixes for ChaSen parser and EDICT-UTF8 index creation and searching.

Version 1.0.10: Support for UTF-8 encoded KANJIDIC files. This means that the German kanjidic_de can now be used.

Version 2 alpha 0: Many changes, most notably an improved user interface.

Version 2 alpha 1: Fixed JGloss startup. Added LaTeX export.

Version 2 alpha 2: Improved dictionary recognition. Support for EDICT files with UTF-8 (Unicode) encoding.

Version 2 alpha 3: Support for UTF-8 encoded KANJIDIC files. This means that the German kanjidic_de can now be used. The build process now uses Apache ant. The documentation format is changed to Docbook XML and processed using Docbook XSL and FOP.

JGloss-WWW is a Java servlet which adds dictionary lookup results to Japanese HTML pages on the fly. It is currently experimental and only available in the source release of JGloss (and from the CVS repository). If you are running a web server with servlet engine installed you might want to try it out. The documentation is here.