Attached is a CS3 .zip archive containing the INDD file before running the datamerge and after running the datamerge - including the Unicode UTF16 .csv database file.
Kyle...thanks for posting about this. I did try to use GREP before I posted but I didn't know what the hell I was doing. You are correct about my file setup. I will report back tomorrow after I get back to work and try this new information. The client said I could just delete the Chinese characters for this run, so thats what I did to get the job done friday. I'll try this and see how it goes on following orders. Thanks again for all the help to both of you.
I had the same problem as you, when producing variable data jobs with mixed Hebrew and English characters.
To phrase it simply: I wanted InDesign to use a specific font for the Hebrew text, and another for the English text. Neither InDesign nor my VDP plugin (InData) had a way to set this preference.
As you suspected, I finally opened the Hebrew font in FontLab, pasted the English font's glyphs at the appropriate character range, saved it as a new OpenType font, and voilà! Using this single new font, InDesign automatically takes the correct glyphs for every language in the text. Maybe the open-source FontForge editor will work too.
By the way, I have no idea whether copying/pasting the glyphs violates the license agreement of both fonts, but I couldn't find a better way to solve this, and at the bottom line, I am using the fonts as they were, it's just a small hack behind the scenes to make life easier...