Is the issue document specific? Or occurs with any document?
Do you get a crash, or just the "cannot open file" exception?
Can you reproduce using our TextSearch sample?
Please post code, or modified TextSearchTest file, and if needed the input PDF file.