"incorrect header check" using ZLIB Compression

[Kevin Kalbfleisch]

I have written a simple compression program which uses ZLIB 1.1.4 to
compress a chunk of data in memory. I am using the ZLIB library as follows:

RWBoolean
ZLIBCompressionImpl::compress( const char* data_in,
unsigned int data_in_len,
char* data_out,
unsigned int* data_out_len )
{
int err = ::compress( (unsigned char *) data_out,
(unsigned long* ) data_out_len,
(unsigned char *) data_in,
(unsigned long) data_in_len );
if( err != Z_OK )
{
THROW_SEP( "ZLIB compression error" );
}

return TRUE;
}

Every thing works fine and a I can decompress the data using:

RWBoolean
ZLIBCompressionImpl::uncompress( const char* data_in,
unsigned int data_in_len,
char* data_out,
unsigned int* data_out_len )
{
int err = ::uncompress( (unsigned char *) data_out,
(unsigned long* ) data_out_len,
(unsigned char *) data_in,
(unsigned long) data_in_len );
if( err != Z_OK )
{
THROW_SEP( "ZLIB compression error" );
}

return TRUE;
}

As long as I am using the C++ code above I can compress and decompress at
will, and have never seen an error. The problem comes in when
I try to decompress the data in python.

If I write the compressed data to a file as follows:
ofstream oFile3;
oFile3.open( "rebundle.varz", ios:: out | ios::binary );
int compressedLen = compressed.length();
cout << "About to write: " << compressedLen
<< " bytes to rebundle.varz" << endl;
oFile3.write( (char*)&compressedLen , 4 );
oFile3.write( compressed.data(), compressedLen );
oFile3.flush();
oFile3.close();

then read it in with python and try to decompress it:

Python 2.1.1 (#1, Oct 9 2001, 13:27:21) [C] on aix4
Type "copyright", "credits" or "license" for more information.
>>>

fh = open( name, "rb" )
lenStr = fh.read( 4 )
length = struct.unpack( '!L', lenStr )[0]
blob = fh.read( length )
blob = zlib.decompress( blob )

I then get the following error:
zlib.error: Error -3 while decompressing data: incorrect header check.

Python is reading the correct number of bytes and the data looks the
same...

Is there some extra header that I should be writing? Does the python
version of ZLIB have known problems (I think that it is using ZLIB 1.1.3)

Any ideas?

Thanks,

Kevin Kalbfleisch
kev...@cosd.fedex.com

-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
Check out our new Unlimited Server. No Download or Time Limits!
-----== Over 80,000 Newsgroups - 19 Different Servers! ==-----

[Andreas Koltes]

If you are using 32 KB sliding window (default) the ZLib header should be:

0x78, 0xDA - max mode
0x78, 0x9C - normal mode
0x78, 0x5E - fast mode
0x78, 0x01 - no compression

If you use preset dictionaries the second header byte looks different. In
this case an extended header is required too.