what exactly are trying to test with the report? Are just concerned if it was created? Or that it is containing the same exact information each time? or that only specific areas of the pdf file contain information you are wanting to check?
You can't access the pdf file directly, but there are ways around testing it depending on the definition of what mean by "test"
Echoing what Tony said, testing is all about looking at your expected output for a given set of inputs. If your output is unknown, then so is success. There are a number of methods I have used to check for success with PDFs in the past.
<ul type="square">[*] For a given set of inputs, I should expect the report to be generated to be X. I can capture an expected result and then execute a file comparison for both size and content. If size passes, but content fails, I may still have a pass, as the report may contain a date stamp when the report is executed which may be beyond my control for setting initial conditions for the run. If size and content both fail, then clearly I have a failure condition.[*] Simple existence check, followed by a start of acrobat with the filename and see if an error occurs to check for a properly formatted file.[*] Existence check, plus a raw file open, checking for the beginning sequence of "%PDF" and the ending sequence of "%%EOF." Assuming that if both are present, then a valid PDF file has been written[*] Although it doesn't appear to be so, PDF files are text formatted files. You can open them in notepad and peruse them to youyr heart's extent. Since most of us don't speak PostScript natively, most of the time this is just giberish to us. But occasionally, in a non-compressed, non-encrypted PDF, you will happen upon text that does appear in your report, identifying text which can be used to assess success or failure. In such a case, you can combine a raw file open, a read traversal of the PDF along with checking for the existence of certain text tags in the PDF file for success. Minimally I have gone for the Beginning and end tags, plus two other elements that should be in a properly created and output PDF file of the type my application creates. Note, this is only an option for uncompressed and unencrypted PDF files. It will also help if you head to your local bookstore and pick up a copy of the PostScript Language Reference and the PDF document reference.[/list]