I'm using PDFsharp (with some other tools) to cut up a PDF file into pieces. It works very well for the actual PDF, but I find that the metadata comes out all mangled, with Chinese characters and such. Looking more closely, it seems the character %xy (x and y hexadecimal numbers) is being 'translated' into %Ex%(80+4*y)%80. Checking shows that internally the right string is being used, and if I load the resulting PDF file into PDFsharp again, I see the 'wrong' text there as well. Anyone knowing what is going on? Do I have to change some encoding setting somewhere? Or is this a bug (and if so, why hasn't anyone noticed it before)?
|