A translation table is a module in DBT that provides the rules to convert (translate) a document from print-to-braille or from braille-to-print. Normally, it is selected by the DBT template that controls production of the current document. All documents have a template. In fact, for many languages there are multiple templates, with differences in translation rules or formatting, but each references at least one translation table. (For more on templates, see DBT Templates, the Basics.)
Regardless of your template, you can choose a different translation table to translate your current document using the Translation Table selection from the DBT Document Menu.
You can also select different translation tables to use for particular passages in your document. See the section below on Language Table Switching.
The Japanese tables support print-to-braille translation of Japanese-language literary text into uncontracted Japanese braille.
Translation from braille-to-print is supported for this language.
Table Designator: jpn identifies this translation table for Language Table Switching.
Braille Contractions: This language is usually produced in uncontracted braille. This means that the letters of each word are rendered into braille on a one-for-one basis.
Capital Sign: The Japanese translator does not use any capitalization mark in braille because the native script does not employ capitalization. Text written with the Roman alphabet (such as English) uses dot 6 for capitalization.
Emphasis: The Japanese translator ignores all bold, italics, and underlining from the inkprint text. These are not represented in the braille.
Mathematical Braille: Duxbury Systems is not yet supporting the national mathematics braille system with this translator.
Script Systems Used: This translator handles Hiragana, Katakana, and Roman characters (generally introduced to handle words of foreign origin). Any Kanji (Chinese or Han) characters are converted to Hiragana characters during file import. (See the important note about Kanji at the bottom of this page.)
A number of DBT codes affect the mode of the translation or create special translation effects on specific letters or symbols. Some translation modes are specific to particular translator tables.
There are no special translator modes for this table, only those which are present for all tables, such as the [lnb~] code (language-switch) and a few codes for internal testing.
For more about DBT codes that affect the mode of translation, search on the two words, "Translation Code", in the topic, DBT Codes Quick Reference.
DBT has translation tables for over 200 world languages. Modern versions of DBT allow using multiple language translation tables within a single document.
Suppose that you are working on a document using this base translation table, but it has passages in a foreign language, or that need a technical braille code. At the beginning of each such passage, insert the DBT code lnb, followed by ~ (tilde) and the table designator for the desired language table. (The table designator for each language table is listed in the Key Characteristics.) Note that using the lnb code you can change from the base table to virtually any other translation table and back again.
For some language tables, the table designator is short, like ise for Icelandic. Thus, to switch to Icelandic braille translation, insert [lnb~ise]. The table designators are more elaborate for mathematics code tables and for languages that have multiple translation tables. As an example, the designator for Unified French Braille is fra-xuf. To start a passage in the French Unified Braille code, insert [lnb~fra-xuf]. At the end of a foreign language passage, use the plain [lnb] code to switch back to the original, base, language translation table.
Some translation tables, and hence their table designators, are for braille codes but not for natural languages. Some examples are the International Phonetic Alphabet (designator: qip) and Nemeth Code (designator: qmt-xnem72m) for mathematics. Using lnb with those table designators allows you to switch to the IPA braille code or the Nemeth braille math code.
While a plain [lnb] code returns translation to the base language, it does not restore any other translation properties that might have been in effect before the switch. For example, if you had been using a [g1L] code (for "grade 1 lock") to prevent contractions, you need to repeat that code after the [lnb] code to restore that effect. Fortunately, you can build lnb codes into DBT styles, to customize what modes to enter and exit at the switch in and out of a translation table.
Note that DBT templates whose names contain the word "basic" all have a number of styles defined for switching between different translation tables. (For the list, see Basic Templates.)
Notice that if you include Roman alphabet characters in a document but do not use a language style or code to switch to a different translation table, you will get uncontracted English with the same marks of punctuation, emphasis, and capitalization used for the base language.
To create the Japanese tables, Duxbury consulted several resources, including the article on Japanese braille at en.wikipedia.org/wiki/Japanese_Braille and the "Japanese Braille Tutorial", by Mitsuji Kadota.
Written Japanese is a combination of three script systems: Hiragana, Katakana, and Kanji. Braille translations for Hiragana and Katakana are reasonably straight forward. For Kanji, however, the conversion to Hiragana depends on the correct pronunciation of the Kanji in the linguistic context of the text being translated. This implementation of the Japanese braille tables performs a fixed replacement of Kanji characters with Hiragana or Katakana during the importation of the text into DBT. This transformation is based on The KANJIDIC/KANJD212 Project. Duxbury thanks the language specialists that offered this material on the Web for our use.
In the current state of development, generating quality Japanese braille depends on the intervention of a skilled user. Relying on the automated output of DBT without making a careful review of the results of the import process is not expected to produce braille of sufficient quality, in most cases, where Kanji conversion is involved.
Duxbury Systems welcomes suggestions on ways to improve on this work.
If you have information about changes in the braille rules for your language, please email Duxbury so that we may update our translation tables. Likewise, if you find errors in your translated document, in either the print-to-braille or braille-to-print direction, please contact us. To describe your problem, it is best to send both the *.dxp and *.dxb files along with a full explanation, because the correct braille is often a matter of very specific circumstances. Contact Duxbury at: languages@duxsys.com.