PDFsharp & MigraDoc Foundation

Hi,

I've been using the provided example http://www.pdfsharp.net/wiki/ExportImages-sample.ashx to extract images from each page in a PDF.
The provided example shows how to determine if an image is /DCTDecode or /FlateDecode, I've provided a fragment of the code sample from your Wiki below (for reference):




But some of my PDFs seems to have images with two encodings and this causes the GetName() method to fail with a null exception. I've provided an example PDF fragment below:




My understanding is that /Filter [/FlateDecode/DCTDecode] means that the Image was encoding using Flate and then DCT and I should decode in reverse order. Is that correct?
But I don't know how to detect this condition in code since GetName("/Filter") returns null in this scenario.

So how do I detect this scenario in code?
And in the above /FlateDecode/DCTDecode example, what should I be doing to decode it and save to disk?

Thanks.

Simon Bond

If "/Filter" has an array then access the members of that array.

Use the "Decode" method of the "FlateDecode" filter class to obtain the embedded JPEG file.

_________________
Regards
Thomas Hoevel
PDFsharp Team

Thankyou for the reply.

Apologies for being dense, but would it be possible to provide a tiny snippet of C# which shows how to detect that "/Filter" is an array?
Maybe I've missed something in the documentation, but I can't see an example which shows how to do this.

I don't see how I can make use of IsArray() in C# when I can't even get a handle to /Filter

If the following code causes an exception then how do I get a handle to /Filter to determine that it is an array:



Thanks,

Simon

You can easily set a breakpoint and inspect the properties.
I first have to find a PDF that uses both filters before I can start trying anything...

Maybe you have to use "GetName(0)" and "GetName(1)" after checking the Count property.

A web search found this topic:
viewtopic.php?p=5587#p5587
Using "Contains" does not ensure the order of filters, but only "/FlateDecode /DCTDecode" makes sense.

_________________
Best regards
Thomas
(Freelance Software Developer with several years of MigraDoc/PDFsharp experience)

Excellent, thankyou for the advice and information.

Simon