Handling raw PDF source in ActionScript

I am using the JODConverter web service to convert an ODT-document to a PDF-fi开发者_C百科le. I have some working Ruby code, that will load up the ODT-file and convert it using the web service. The resulting PDF-file is then returned to me and I can easily save it.

When I try to do the same thing in ActionScript, I seem to be facing some issues with FlateDecode blocks in the PDF source. They are somehow altered (possibly because ActionScript strings are UTF-8). The result is that the resulting PDF-file is incomplete. Meta-information is correct, but the file appears to be blank.

I would appreciate any kind of feedback relating to this issue.