DeflateStream throws exception when inflating PDF streams

DeflateStream throws exception when inflating PDF streams

• Teymur Hajiyev

  Hi!

  Did you ever get a solution to this problem

  Thomas

  
  

• Pirgher

  Still no replies
  I will wait one more day, then try posting a bug report.
  John
  
  

• Looooooka

  DeflateStream and GZipStream have some bugs in them. I tried to compress a block of 64k worth of Es and the decompressed buffer is missing 2 E's. Here is the code.

  using System;

  using System.Collections.Generic;

  using System.IO;

  using System.Linq;

  using System.Security;

  using System.Security.Cryptography;

  using System.Text;

  using NUnit.Framework;

  [TestFixture]

  public class PositiveZip

  {

  /// <summary>

  /// Test the block compression and decompression APIs.

  /// </summary>

  [Test]

  public void ZipTestCase01001()

  {

  int plainbuffsize = 1024 * 64;

  byte[] plainblock64K = new byte[plainbuffsize];

  for (int pbindex = 0; pbindex < plainbuffsize; pbindex++)

  {

  plainblock64K[pbindex] = 0x45;

  }

  MemoryStream uncompstream = new MemoryStream(plainblock64K);

  MemoryStream compstream = new MemoryStream();

  Zip.CompressStreamToStream(uncompstream, compstream);

  byte[] compbuff = compstream.ToArray();

  Assert.Less(compbuff.Length, plainblock64K.Length);

  compstream.Position = 0;

  MemoryStream recoverstream = new MemoryStream();

  Zip.DeCompressStreamToStream(compstream, recoverstream);

  byte[] recoverbuffer = recoverstream.ToArray();

  Assert.AreEqual(plainblock64K.Length, recoverbuffer.Length);

  Assert.AreEqual(plainblock64K, recoverbuffer);

  }

  }

  static public class Zip

  {

  static public void CompressStreamToStream(Stream instream, Stream outstream)

  {

  GZipStream deflate = new GZipStream(outstream, CompressionMode.Compress, true);

  byte[] buffer = new byte[4096];

  do

  {

  int bytesread = instream.Read(buffer, 0, 4096);

  if (bytesread <= 0)

  break;

  deflate.Write(buffer, 0, bytesread);

  } while (true);

  return;

  }

  static public void DeCompressStreamToStream(Stream instream, Stream outstream)

  {

  GZipStream deflate = new GZipStream(instream, CompressionMode.Decompress, true);

  byte[] buffer = new byte[4096];

  do

  {

  int bytesread = deflate.Read(buffer, 0, 4096);

  if (bytesread <= 0)

  break;

  outstream.Write(buffer, 0, bytesread);

  } while (true);

  return;

  }

  }

  
  

• Raymond Chen

  Turns out it's bloody PDF at fault. I went hunting through the iTextSharp source, and found out (not that the comments indicate this) that PDF slaps a 2 byte header onto the front of the DEFLATE stream.

  Simply ReadByte twice on your input stream, and THEN pass it to DeflateStream and it works a treat.

  
  
  

• lucasjordan

  And it looks like there's been no movement on this issue in 2 months...
  
  

• Karsinogeeni

  Looks like he already has:

  http://lab.msdn.microsoft.com/productfeedback/viewfeedback.aspx feedbackid=3f6e3d2a-23d1-46dd-aaa0-53df18df4190.

  
  
  

• Shula

  I got the same problem here ... any solution
  
  

• paragp

  This issue with this is that I wasn't closeing the deflate stream. When you don't close a deflate stream it doesn't flush immediately to the underlying stream.

  
  

• Fidencio Monroy

  John,

  Can you post a bug report on the Microsoft Product Feedback Center with a project that reproduces the problem

  Regards

  David

  
  
  
• DeflateStream throws exception when inflating PDF streams