The check for bIsSigContents was done on data in the dictionary that may not be available at the moment of the check because /Contents may be written before /Type in the pdf. The solution is to defer the check later when all keys has been added to the dictionary. To do that we just store the char buffer for "Contents" when we find it and decode the string with or without the PdfEncrypt instance after decision. At that point we finally add it to the dictionary. Also check for signature dictionaries of type /Type/DocTimeStamp, as done in PDFBox[1].
[1] https://issues.apache.org/jira/browse/PDFBOX-3173
fix-reading-contents-string-signature-dictionaries.diff
Description: Binary data
_______________________________________________ Podofo-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/podofo-users
