I have managed to generate a pdf output from xml data that contain non-Roman characters, e.g., Chinese. This is done by selecting a Truetype font, generating the font matrics in xml, etc.
What is puzzling to me is that by not doing any of the above, or introducing a font type that can handle Chinese characters, thus leaving Apache fop to work on its default mode, I get an output where
- In the Table of Contents, the Chinese texts come out perfectly
- In the body, all the Chinese characters are replaced by "#", i.e., it needs the font.
A screen shot is given below.
Can anybody explain why the TOC behaves differently please? Many thanks!
Read this for your answer: https://forums.adobe.com/thread/1271916?tstart=0
Basically all fonts information for "bookmarks" which is what you have are chosed by the Reader application, not the file. Therefore, Adobe "finds" you characters within the application and not the PDF file.