Is it possibile to retrieve .nav from pdf portfolio?

Possible in the sense you can extract the navigator SWF and supporting assets from the PDF COS streams byte-by-byte? Yes.

Is that the same as the NAV file? No.

Can you extract any of that with Acrobat? No. Neither do I know of any third-party software that'll do it, as the likely uses for such a feature are overwhelmingly illegal.

The contents of a .NAV file (which is - more or less - a .ZIP file) get broken up and spread across several COS dictionaries in the PDF file when you apply the layout to your portfolio. Assuming you know what you are looking for, any decent PDF library that can access arbitrary COS level objects can pull out the components and you can then reconstruct the .NAV.

Dave: Assuming the file has not been encrypted, this wouldn't be illegal, just very, very hard.

Assuming we're talking about a NAV created by someone other than Adobe, if it wasn't your NAV in the first place and you don't have the author's permission to decompile the PDF, pulling content out to re-use would be a no-no under copyright law, as anything you made with your phoenix navigator would be an unauthorized derivative work.

If you have permission to extract it, then it's easier to ask the guy who built the PDF to send you his original file. If you have to pull it out of a PDF with a tow truck, the balance of probability is that you're doing something you shouldn't be.

The only legitimate case I can see is if you've written your own navigator, used it, then somehow deleted both the NAV file [u]and[/u] the Flex project you used to build it.

Thanks for your answers!
I do not know if it's illegal, my friend send me a pdf portfolio with some information that I think is contained in an XML file contained in .Nav.
I can copy this information by hand or try to retrieve the xml and save lot of my time.

what is the "PDF COS streams byte-by-byte" and "decent PDF library that can access arbitrary COS level objects can pull out the components"?

PDF files are made up of a series of objects - blocks of text or binary data - and since the PDF specification is public knowledge, it's possible (though very difficult in all but the simplest cases) to open a PDF file in a text editor and copy out the relevant sections, once you've worked out which they are. The problem is that almost all the objects in a PDF are compressed, so your XML file won't appear in human-readable form unless you decompress them first, and that is also far from a trivial task. As portfolios are typically very large, searching through the file by hand won't be much fun either.

There are code libraries for many programming languages, as well as a C++ API for Acrobat, that programmers can use to directly manipulate the content of a PDF - but working with binary objects and converting them back into their original form is not a novice-level project.

Finally... I done it!!!

I'm a php programmer so i search for some library that work with pdf and the object inside, so i can retrieve the xml object coded by "Flatedecode" filter.
decode it and get the xml data!

Thank you again!!