Issues with Double-Byte Languages

This document will describe the known issues associated with double-byte languages (Japanese, Korean, etc.) and ePublisher output. All of these issues have been tested and apply to projects on a Windows XP Japanese system unless specified otherwise.

Double-byte characters in the path of a source document will result in errors when adding or scanning the document.
If there are double-byte characters in the path leading to the output directory, then some files will not be copied into the output folder when you generate. It looks like JavaScript and XML files are not copied by Windows, so the output will not function correctly.
If you create output on an English system and then copy it to a Japanese system, the output will not load correctly if the path leading to those files has double-byte characters. This occurs only with FireFox; Internet Explorer will load the output correctly. This seems to be related to escaped characters in the URL field. The redirect JavaScript might be sending an escaped path to prevent errors with backslashes and special characters; IE apparently unescapes by default, so the double-byte characters show up correctly, while FireFox uses the escaped values, and the pages don't load. This applies to Linux Japanese systems as well.

Any insight into these issues is appreciated, as we continue to work toward a solution to the problems presented by double-byte file paths.