xref issue

Hi,

We are using PDFSharp for PDF file editing at server side. We have an issue with PDF file which was edited by RadaeePDF Cordova plugin. When we are trying to open it by using PDFSharp we receive the following exception:

Unexpected token 'xref' in PDF stream. The file may be corrupted. If you think this is a bug in PDFsharp, please send us your PDF file.

It looks like RadaeePDF Cordova plugin corrupts file. The XREF table should contain a list to the objects of the PDF. In this case the XREF table includes a link to the XREF table.

Original file: www.dropbox.com/s/wyjtoqokthrrryb/Original%20File.pdf
Edited file: www.dropbox.com/s/3076hz16l19wcys/File.pdf

There is something I'm not understanding.

The input file is a normal PDF while the output one is an pdf/a-1.

May you clarify the full flow?

Original -> compiled by radaee -> transformed to pdf/a-1 by pdfsharp?

There are some errors I can't recognize as generated by our side (for example the missing Zapf Dingbats font is a requirement from pdf/a-1 transformation).
May you send us the missing pdf file I guess you've in your full process?

File.pdf is not processed by PDFSharp. This file was created by RadaeePDF Cordova plugin for Android.

Ok, I'll double check with dev team.

Thank you.

It seems the original file has some corruption in the AcroForm structure.
I'm uploading the report generated by AcrobatPro XI.
Sorry if it is in italian.

It seems the original PDF generator (OpenOffice?) has some issue while generating forms.