New Forum | Previous | Next | (P-PDF) Developers
Topic: /FlateDecode chokes ZLib
Conf: (P-PDF) Developers, Msg: 51090
Date: 5/29/2002 04:43 PM
Coincidental to this discussion, I have just discovered a file that ZLib seems to not like.
I have a PDF file that is FlateCoded, and I use ZLib to decompress the streams.
I have an interesting file that is PDF version 1.2 and was produced with Distiller 4.0, according to the document info.
When I pass the data from the content stream of one of the pages to the inflate call from ZLib, I get back an "Invalid block type" error on one of the bytes. I tried passing the file in smaller chunks to ZLib, and it gets the same error on whatever chunk contains this particular byte.
Strangely, Acrobat displays the page without a problem. There is only one stream in the page Contents, so I know that this is the stream I am viewing. The bad byte is only a few dozen bytes into the stream, so I know that Acrobat is not partially decompressing the stream and displaying only what comes before the bad byte.
I have manually browsed the bytes in memory, and they are the same as the bytes from the file.
I am running out of ideas, and if anyone has a suggestion, I would be most appreciative.