Localization support in the DITA Open Toolkit

The DITA Open Toolkit (DITA OT), a component which is used for generating output from DITA documents in XMetaL Author Enterprise Edition, includes various configurable features to support localization. This article summarizes where to find documentation on these features.

Identifying the Language of Your Content
For instructions on identifying the language of non-English documents so that the DITA Open Toolkit can process them correctly, see DITA: Indicating the language of content for DITA processing.

Generated Text
The DITA OT can implement appropriate generated text, such as “Note:” at the beginning of <note> elements for English and “Nota:” at the beginning of <note> elements for Spanish, for each locale. For details, see Editing "generated text" strings for DITA output for details.

Font and Character Support
No single font is optimal for all languages. By default, the DITA OT uses different fonts for different locales. For PDF output, you might need to change some settings in order for fonts to appear. See:

See also: DITA: Configuring XMetaL Author Enterprise to Generate Localized CHM Files.

Online Help Systems
If you are creating an online help system, you can make text on the help system’s buttons and tabs appear in the correct language. However, this does not happen automatically.

For WebHelp, strings used in buttons and tabs are stored in the DITA_OT\demo\webhelp\customization\common\index.html file. This file is preconfigured for English only. To produce WebHelp in another language, make a copy of this file, translate the English strings into the required language, and replace the original index.html file with the translated index.html file before you generate WebHelp output.

For Microsoft HTML Help (CHM) files, a different procedure is required. See DITA: Configuring XMetaL Author Enterprise to Generate Localized CHM Files.

Index Sorting
Indexes and glossaries are sorted according to locale-specific rules. For details, see Troubleshooting non-English indexes in DITA output.

Text Direction
Some languages, such as Arabic and Hebrew, must be displayed right-to-left instead of left-to-right. For non-PDF outputs, the DITA OT does this automatically based on the xml:lang attribute in the content. The preconfigured language files for PDF output do not include any right-to-left languages. It is possible to add support in PDF output for right-to-left languages, however we do not know of any publicly-available documentation on how to do it.