Our application uses the .NET report viewer component to display reports that may contain any Unicode characters. When this report is exported (via the .NET report viewer) only western European and Arabic glyphs are displayed in the PDF file. Glyphs from other languages (Chinese, Japanese, Korean, Bengali, Malayalam, etc..) are rendered as empty boxes.
If this report is exported to any other format all of the glyphs can be displayed. The system where the report is viewed, exported, and where the exported report is viewed has the East Asian fonts and complex scripting installed. We can reproduce this on an XP SP3 machine and a Windows 2008 Server machine (these are the only systems we have tried).
If the report is exported to an excel document, and then the excel document is saved as a PDF then all glyphs can be rendered in the excel-generated PDF.
Looking at the PDF properties in Adobe Reader I see that the PDF generated by the Crystal Reports viewer contains only the font we specified in the report (Tahoma), while the excel-generated PDF includes several of the fonts to which the Tahoma OpenType font is linked.
This problem was observed using the Crystal Reports 2008 SP2 runtime.
How can we get the PDF export to include the linked fonts needed to render the non-western European glyphs?