summary refs log tree commit diff
path: root/Crypto/src/util
diff options
context:
space:
mode:
Diffstat (limited to 'Crypto/src/util')
-rw-r--r--Crypto/src/util/Arrays.cs226
-rw-r--r--Crypto/src/util/BigIntegers.cs72
-rw-r--r--Crypto/src/util/Enums.cs72
-rw-r--r--Crypto/src/util/Platform.cs171
-rw-r--r--Crypto/src/util/Strings.cs107
-rw-r--r--Crypto/src/util/collections/CollectionUtilities.cs71
-rw-r--r--Crypto/src/util/collections/EmptyEnumerable.cs44
-rw-r--r--Crypto/src/util/collections/EnumerableProxy.cs25
-rw-r--r--Crypto/src/util/collections/HashSet.cs99
-rw-r--r--Crypto/src/util/collections/ISet.cs19
-rw-r--r--Crypto/src/util/collections/LinkedDictionary.cs178
-rw-r--r--Crypto/src/util/collections/UnmodifiableDictionary.cs64
-rw-r--r--Crypto/src/util/collections/UnmodifiableDictionaryProxy.cs66
-rw-r--r--Crypto/src/util/collections/UnmodifiableList.cs67
-rw-r--r--Crypto/src/util/collections/UnmodifiableListProxy.cs61
-rw-r--r--Crypto/src/util/collections/UnmodifiableSet.cs59
-rw-r--r--Crypto/src/util/collections/UnmodifiableSetProxy.cs56
-rw-r--r--Crypto/src/util/date/DateTimeObject.cs25
-rw-r--r--Crypto/src/util/date/DateTimeUtilities.cs47
-rw-r--r--Crypto/src/util/encoders/Base64.cs95
-rw-r--r--Crypto/src/util/encoders/Base64Encoder.cs307
-rw-r--r--Crypto/src/util/encoders/BufferedDecoder.cs117
-rw-r--r--Crypto/src/util/encoders/BufferedEncoder.cs117
-rw-r--r--Crypto/src/util/encoders/Hex.cs131
-rw-r--r--Crypto/src/util/encoders/HexEncoder.cs164
-rw-r--r--Crypto/src/util/encoders/HexTranslator.cs108
-rw-r--r--Crypto/src/util/encoders/IEncoder.cs18
-rw-r--r--Crypto/src/util/encoders/Translator.cs19
-rw-r--r--Crypto/src/util/encoders/UrlBase64.cs127
-rw-r--r--Crypto/src/util/encoders/UrlBase64Encoder.cs31
-rw-r--r--Crypto/src/util/io/BaseInputStream.cs54
-rw-r--r--Crypto/src/util/io/BaseOutputStream.cs54
-rw-r--r--Crypto/src/util/io/PushbackStream.cs52
-rw-r--r--Crypto/src/util/io/StreamOverflowException.cs27
-rw-r--r--Crypto/src/util/io/Streams.cs94
-rw-r--r--Crypto/src/util/io/TeeInputStream.cs54
-rw-r--r--Crypto/src/util/io/TeeOutputStream.cs42
-rw-r--r--Crypto/src/util/io/pem/PemGenerationException.cs26
-rw-r--r--Crypto/src/util/io/pem/PemHeader.cs55
-rw-r--r--Crypto/src/util/io/pem/PemObject.cs47
-rw-r--r--Crypto/src/util/io/pem/PemObjectGenerator.cs13
-rw-r--r--Crypto/src/util/io/pem/PemObjectParser.cs17
-rw-r--r--Crypto/src/util/io/pem/PemReader.cs94
-rw-r--r--Crypto/src/util/io/pem/PemWriter.cs120
-rw-r--r--Crypto/src/util/net/IPAddress.cs197
-rw-r--r--Crypto/src/util/zlib/Adler32.cs88
-rw-r--r--Crypto/src/util/zlib/Deflate.cs1640
-rw-r--r--Crypto/src/util/zlib/InfBlocks.cs618
-rw-r--r--Crypto/src/util/zlib/InfCodes.cs611
-rw-r--r--Crypto/src/util/zlib/InfTree.cs523
-rw-r--r--Crypto/src/util/zlib/Inflate.cs387
-rw-r--r--Crypto/src/util/zlib/JZlib.cs73
-rw-r--r--Crypto/src/util/zlib/StaticTree.cs152
-rw-r--r--Crypto/src/util/zlib/Tree.cs367
-rw-r--r--Crypto/src/util/zlib/ZDeflaterOutputStream.cs155
-rw-r--r--Crypto/src/util/zlib/ZInflaterInputStream.cs131
-rw-r--r--Crypto/src/util/zlib/ZInputStream.cs196
-rw-r--r--Crypto/src/util/zlib/ZOutputStream.cs220
-rw-r--r--Crypto/src/util/zlib/ZStream.cs214
59 files changed, 9084 insertions, 0 deletions
diff --git a/Crypto/src/util/Arrays.cs b/Crypto/src/util/Arrays.cs
new file mode 100644
index 000000000..7fa05ebee
--- /dev/null
+++ b/Crypto/src/util/Arrays.cs
@@ -0,0 +1,226 @@
+using System;
+using System.Text;
+
+namespace Org.BouncyCastle.Utilities
+{
+
+    /// <summary> General array utilities.</summary>
+    public sealed class Arrays
+    {
+        private Arrays()
+        {
+        }
+
+		public static bool AreEqual(
+			bool[]  a,
+			bool[]  b)
+		{
+			if (a == b)
+				return true;
+
+			if (a == null || b == null)
+				return false;
+
+            return HaveSameContents(a, b);
+		}
+
+        public static bool AreEqual(
+            char[] a,
+            char[] b)
+        {
+            if (a == b)
+                return true;
+
+            if (a == null || b == null)
+                return false;
+
+            return HaveSameContents(a, b);
+        }
+
+        /// <summary>
+        /// Are two arrays equal.
+        /// </summary>
+        /// <param name="a">Left side.</param>
+        /// <param name="b">Right side.</param>
+        /// <returns>True if equal.</returns>
+        public static bool AreEqual(
+			byte[]	a,
+			byte[]	b)
+        {
+			if (a == b)
+				return true;
+
+			if (a == null || b == null)
+				return false;
+
+			return HaveSameContents(a, b);
+		}
+
+		[Obsolete("Use 'AreEqual' method instead")]
+		public static bool AreSame(
+			byte[]	a,
+			byte[]	b)
+		{
+			return AreEqual(a, b);
+		}
+
+		/// <summary>
+		/// A constant time equals comparison - does not terminate early if
+		/// test will fail.
+		/// </summary>
+		/// <param name="a">first array</param>
+		/// <param name="b">second array</param>
+		/// <returns>true if arrays equal, false otherwise.</returns>
+		public static bool ConstantTimeAreEqual(
+			byte[]	a,
+			byte[]	b)
+		{
+			int i = a.Length;
+			if (i != b.Length)
+				return false;
+			int cmp = 0;
+			while (i != 0)
+			{
+				--i;
+				cmp |= (a[i] ^ b[i]);
+			}
+			return cmp == 0;
+		}
+
+		public static bool AreEqual(
+			int[]	a,
+			int[]	b)
+		{
+			if (a == b)
+				return true;
+
+			if (a == null || b == null)
+				return false;
+
+			return HaveSameContents(a, b);
+		}
+
+        private static bool HaveSameContents(
+            bool[] a,
+            bool[] b)
+        {
+            int i = a.Length;
+            if (i != b.Length)
+                return false;
+            while (i != 0)
+            {
+                --i;
+                if (a[i] != b[i])
+                    return false;
+            }
+            return true;
+        }
+
+        private static bool HaveSameContents(
+            char[] a,
+            char[] b)
+        {
+            int i = a.Length;
+            if (i != b.Length)
+                return false;
+            while (i != 0)
+            {
+                --i;
+                if (a[i] != b[i])
+                    return false;
+            }
+            return true;
+        }
+
+        private static bool HaveSameContents(
+			byte[]	a,
+			byte[]	b)
+		{
+			int i = a.Length;
+			if (i != b.Length)
+				return false;
+			while (i != 0)
+			{
+				--i;
+				if (a[i] != b[i])
+					return false;
+			}
+			return true;
+		}
+
+		private static bool HaveSameContents(
+			int[]	a,
+			int[]	b)
+		{
+			int i = a.Length;
+			if (i != b.Length)
+				return false;
+			while (i != 0)
+			{
+				--i;
+				if (a[i] != b[i])
+					return false;
+			}
+			return true;
+		}
+
+        public static string ToString(
+			object[] a)
+		{
+			StringBuilder sb = new StringBuilder('[');
+			if (a.Length > 0)
+			{
+				sb.Append(a[0]);
+				for (int index = 1; index < a.Length; ++index)
+				{
+					sb.Append(", ").Append(a[index]);
+				}
+			}
+			sb.Append(']');
+			return sb.ToString();
+		}
+
+		public static int GetHashCode(
+			byte[] data)
+		{
+			if (data == null)
+			{
+				return 0;
+			}
+
+			int i = data.Length;
+			int hc = i + 1;
+
+			while (--i >= 0)
+			{
+				hc *= 257;
+				hc ^= data[i];
+			}
+
+			return hc;
+		}
+
+		public static byte[] Clone(
+			byte[] data)
+		{
+			return data == null ? null : (byte[]) data.Clone();
+		}
+
+		public static int[] Clone(
+			int[] data)
+		{
+			return data == null ? null : (int[]) data.Clone();
+		}
+
+		public static void Fill(
+			byte[]	buf,
+			byte	b)
+		{
+			int i = buf.Length;
+			while (i > 0)
+			{
+				buf[--i] = b;
+			}
+		}
+	}
+}
diff --git a/Crypto/src/util/BigIntegers.cs b/Crypto/src/util/BigIntegers.cs
new file mode 100644
index 000000000..df82e1f22
--- /dev/null
+++ b/Crypto/src/util/BigIntegers.cs
@@ -0,0 +1,72 @@
+using System;
+
+using Org.BouncyCastle.Math;
+using Org.BouncyCastle.Security;
+
+namespace Org.BouncyCastle.Utilities
+{
+	/**
+	 * BigInteger utilities.
+	 */
+	public sealed class BigIntegers
+	{
+		private const int MaxIterations = 1000;
+
+		private BigIntegers()
+		{
+		}
+
+		/**
+		* Return the passed in value as an unsigned byte array.
+		*
+		* @param value value to be converted.
+		* @return a byte array without a leading zero byte if present in the signed encoding.
+		*/
+		public static byte[] AsUnsignedByteArray(
+			BigInteger n)
+		{
+			return n.ToByteArrayUnsigned();
+		}
+
+		/**
+		* Return a random BigInteger not less than 'min' and not greater than 'max'
+		* 
+		* @param min the least value that may be generated
+		* @param max the greatest value that may be generated
+		* @param random the source of randomness
+		* @return a random BigInteger value in the range [min,max]
+		*/
+		public static BigInteger CreateRandomInRange(
+			BigInteger		min,
+			BigInteger		max,
+			// TODO Should have been just Random class
+			SecureRandom	random)
+		{
+			int cmp = min.CompareTo(max);
+			if (cmp >= 0)
+			{
+				if (cmp > 0)
+					throw new ArgumentException("'min' may not be greater than 'max'");
+
+				return min;
+			}
+
+			if (min.BitLength > max.BitLength / 2)
+			{
+				return CreateRandomInRange(BigInteger.Zero, max.Subtract(min), random).Add(min);
+			}
+
+			for (int i = 0; i < MaxIterations; ++i)
+			{
+				BigInteger x = new BigInteger(max.BitLength, random);
+				if (x.CompareTo(min) >= 0 && x.CompareTo(max) <= 0)
+				{
+					return x;
+				}
+			}
+
+			// fall back to a faster (restricted) method
+			return new BigInteger(max.Subtract(min).BitLength - 1, random).Add(min);
+		}
+	}
+}
diff --git a/Crypto/src/util/Enums.cs b/Crypto/src/util/Enums.cs
new file mode 100644
index 000000000..abab5266b
--- /dev/null
+++ b/Crypto/src/util/Enums.cs
@@ -0,0 +1,72 @@
+using System;
+using System.Text;
+
+#if NETCF_1_0 || NETCF_2_0 || SILVERLIGHT || PORTABLE
+using System.Collections;
+using System.Reflection;
+#endif
+
+using Org.BouncyCastle.Utilities.Date;
+
+namespace Org.BouncyCastle.Utilities
+{
+    internal sealed class Enums
+    {
+        private Enums()
+        {
+        }
+
+		internal static Enum GetEnumValue(System.Type enumType, string s)
+		{
+			if (!enumType.IsEnum)
+				throw new ArgumentException("Not an enumeration type", "enumType");
+
+			// We only want to parse single named constants
+			if (s.Length > 0 && char.IsLetter(s[0]) && s.IndexOf(',') < 0)
+			{
+				s = s.Replace('-', '_');
+
+#if NETCF_1_0
+				FieldInfo field = enumType.GetField(s, BindingFlags.Static | BindingFlags.Public);
+				if (field != null)
+				{
+					return (Enum)field.GetValue(null);
+				}
+#else
+				return (Enum)Enum.Parse(enumType, s, false);
+#endif		
+			}
+
+			throw new ArgumentException();
+		}
+
+		internal static Array GetEnumValues(System.Type enumType)
+		{
+			if (!enumType.IsEnum)
+				throw new ArgumentException("Not an enumeration type", "enumType");
+
+#if NETCF_1_0 || NETCF_2_0 || SILVERLIGHT || PORTABLE
+            IList result = Platform.CreateArrayList();
+			FieldInfo[] fields = enumType.GetFields(BindingFlags.Static | BindingFlags.Public);
+			foreach (FieldInfo field in fields)
+			{
+                // Note: Argument to GetValue() ignored since the fields are static,
+                //     but Silverlight for Windows Phone throws exception if we pass null
+				result.Add(field.GetValue(enumType));
+			}
+            object[] arr = new object[result.Count];
+            result.CopyTo(arr, 0);
+            return arr;
+#else
+			return Enum.GetValues(enumType);
+#endif
+		}
+
+		internal static Enum GetArbitraryValue(System.Type enumType)
+		{
+			Array values = GetEnumValues(enumType);
+			int pos = (int)(DateTimeUtilities.CurrentUnixMs() & int.MaxValue) % values.Length;
+			return (Enum)values.GetValue(pos);
+		}
+	}
+}
diff --git a/Crypto/src/util/Platform.cs b/Crypto/src/util/Platform.cs
new file mode 100644
index 000000000..52ace89e5
--- /dev/null
+++ b/Crypto/src/util/Platform.cs
@@ -0,0 +1,171 @@
+using System;
+using System.IO;
+using System.Text;
+
+#if SILVERLIGHT || PORTABLE
+using System.Collections.Generic;
+#else
+using System.Collections;
+#endif
+
+namespace Org.BouncyCastle.Utilities
+{
+	internal sealed class Platform
+	{
+		private Platform()
+		{
+		}
+
+#if NETCF_1_0 || NETCF_2_0
+		private static string GetNewLine()
+		{
+			MemoryStream buf = new MemoryStream();
+			StreamWriter w = new StreamWriter(buf, Encoding.UTF8);
+			w.WriteLine();
+			w.Close();
+			byte[] bs = buf.ToArray();
+            return Encoding.UTF8.GetString(bs, 0, bs.Length);
+		}
+#else
+        private static string GetNewLine()
+        {
+            return Environment.NewLine;
+        }
+#endif
+
+        internal static int CompareIgnoreCase(string a, string b)
+        {
+#if (SILVERLIGHT || PORTABLE)
+            return String.Compare(a, b, StringComparison.OrdinalIgnoreCase);
+#else
+            return String.Compare(a, b, true);
+#endif
+        }
+
+#if NETCF_1_0 || NETCF_2_0 || SILVERLIGHT || PORTABLE
+		internal static string GetEnvironmentVariable(
+			string variable)
+		{
+			return null;
+		}
+#else
+		internal static string GetEnvironmentVariable(
+			string variable)
+		{
+			try
+			{
+				return Environment.GetEnvironmentVariable(variable);
+			}
+			catch (System.Security.SecurityException)
+			{
+				// We don't have the required permission to read this environment variable,
+				// which is fine, just act as if it's not set
+				return null;
+			}
+		}
+#endif
+
+#if NETCF_1_0
+		internal static Exception CreateNotImplementedException(
+			string message)
+		{
+			return new Exception("Not implemented: " + message);
+		}
+
+		internal static bool Equals(
+			object	a,
+			object	b)
+		{
+			return a == b || (a != null && b != null && a.Equals(b));
+		}
+#else
+		internal static Exception CreateNotImplementedException(
+			string message)
+		{
+			return new NotImplementedException(message);
+		}
+#endif
+
+#if SILVERLIGHT || PORTABLE
+        internal static System.Collections.IList CreateArrayList()
+        {
+            return new List<object>();
+        }
+        internal static System.Collections.IList CreateArrayList(int capacity)
+        {
+            return new List<object>(capacity);
+        }
+        internal static System.Collections.IList CreateArrayList(System.Collections.ICollection collection)
+        {
+            System.Collections.IList result = new List<object>(collection.Count);
+            foreach (object o in collection)
+            {
+                result.Add(o);
+            }
+            return result;
+        }
+        internal static System.Collections.IList CreateArrayList(System.Collections.IEnumerable collection)
+        {
+            System.Collections.IList result = new List<object>();
+            foreach (object o in collection)
+            {
+                result.Add(o);
+            }
+            return result;
+        }
+        internal static System.Collections.IDictionary CreateHashtable()
+        {
+            return new Dictionary<object, object>();
+        }
+        internal static System.Collections.IDictionary CreateHashtable(int capacity)
+        {
+            return new Dictionary<object, object>(capacity);
+        }
+        internal static System.Collections.IDictionary CreateHashtable(System.Collections.IDictionary dictionary)
+        {
+            System.Collections.IDictionary result = new Dictionary<object, object>(dictionary.Count);
+            foreach (System.Collections.DictionaryEntry entry in dictionary)
+            {
+                result.Add(entry.Key, entry.Value);
+            }
+            return result;
+        }
+#else
+        internal static System.Collections.IList CreateArrayList()
+        {
+            return new ArrayList();
+        }
+        internal static System.Collections.IList CreateArrayList(int capacity)
+        {
+            return new ArrayList(capacity);
+        }
+        internal static System.Collections.IList CreateArrayList(System.Collections.ICollection collection)
+        {
+            return new ArrayList(collection);
+        }
+        internal static System.Collections.IList CreateArrayList(System.Collections.IEnumerable collection)
+        {
+            ArrayList result = new ArrayList();
+            foreach (object o in collection)
+            {
+                result.Add(o);
+            }
+            return result;
+        }
+        internal static System.Collections.IDictionary CreateHashtable()
+        {
+            return new Hashtable();
+        }
+        internal static System.Collections.IDictionary CreateHashtable(int capacity)
+        {
+            return new Hashtable(capacity);
+        }
+        internal static System.Collections.IDictionary CreateHashtable(System.Collections.IDictionary dictionary)
+        {
+            return new Hashtable(dictionary);
+        }
+#endif
+
+        internal static readonly string NewLine = GetNewLine();
+	}
+}
diff --git a/Crypto/src/util/Strings.cs b/Crypto/src/util/Strings.cs
new file mode 100644
index 000000000..1d35920c9
--- /dev/null
+++ b/Crypto/src/util/Strings.cs
@@ -0,0 +1,107 @@
+using System;
+using System.Text;
+
+namespace Org.BouncyCastle.Utilities
+{
+	/// <summary> General string utilities.</summary>
+	public sealed class Strings
+	{
+		private Strings()
+		{
+		}
+
+		internal static bool IsOneOf(string s, params string[] candidates)
+		{
+			foreach (string candidate in candidates)
+			{
+				if (s == candidate)
+					return true;
+			}
+			return false;
+		}
+
+		public static string FromByteArray(
+			byte[] bs)
+		{
+			char[] cs = new char[bs.Length];
+			for (int i = 0; i < cs.Length; ++i)
+			{
+				cs[i] = Convert.ToChar(bs[i]);
+			}
+			return new string(cs);
+		}
+
+        public static byte[] ToByteArray(
+            char[] cs)
+        {
+            byte[] bs = new byte[cs.Length];
+            for (int i = 0; i < bs.Length; ++i)
+            {
+                bs[i] = Convert.ToByte(cs[i]);
+            }
+            return bs;
+        }
+
+        public static byte[] ToByteArray(
+			string s)
+		{
+			byte[] bs = new byte[s.Length];
+			for (int i = 0; i < bs.Length; ++i)
+			{
+				bs[i] = Convert.ToByte(s[i]);
+			}
+			return bs;
+		}
+
+        public static string FromAsciiByteArray(
+            byte[] bytes)
+        {
+#if (SILVERLIGHT || PORTABLE)
+            // TODO Check for non-ASCII bytes in input?
+            return Encoding.UTF8.GetString(bytes, 0, bytes.Length);
+#else
+            return Encoding.ASCII.GetString(bytes, 0, bytes.Length);
+#endif
+        }
+
+        public static byte[] ToAsciiByteArray(
+            char[] cs)
+        {
+#if SILVERLIGHT || PORTABLE
+            // TODO Check for non-ASCII characters in input?
+            return Encoding.UTF8.GetBytes(cs);
+#else
+            return Encoding.ASCII.GetBytes(cs);
+#endif
+        }
+
+        public static byte[] ToAsciiByteArray(
+            string s)
+        {
+#if SILVERLIGHT || PORTABLE
+            // TODO Check for non-ASCII characters in input?
+            return Encoding.UTF8.GetBytes(s);
+#else
+            return Encoding.ASCII.GetBytes(s);
+#endif
+        }
+
+        public static string FromUtf8ByteArray(
+			byte[] bytes)
+		{
+			return Encoding.UTF8.GetString(bytes, 0, bytes.Length);
+		}
+
+        public static byte[] ToUtf8ByteArray(
+            char[] cs)
+        {
+            return Encoding.UTF8.GetBytes(cs);
+        }
+
+		public static byte[] ToUtf8ByteArray(
+			string s)
+		{
+			return Encoding.UTF8.GetBytes(s);
+		}
+	}
+}
diff --git a/Crypto/src/util/collections/CollectionUtilities.cs b/Crypto/src/util/collections/CollectionUtilities.cs
new file mode 100644
index 000000000..fd0bdcc7a
--- /dev/null
+++ b/Crypto/src/util/collections/CollectionUtilities.cs
@@ -0,0 +1,71 @@
+using System;
+using System.Collections;
+using System.Text;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public sealed class CollectionUtilities
+	{
+		private CollectionUtilities()
+		{
+		}
+
+        public static void AddRange(IList to, ICollection range)
+        {
+            foreach (object o in range)
+            {
+                to.Add(o);
+            }
+        }
+
+        public static bool CheckElementsAreOfType(
+			IEnumerable e,
+			Type		t)
+		{
+			foreach (object o in e)
+			{
+				if (!t.IsInstanceOfType(o))
+					return false;
+			}
+			return true;
+		}
+
+        public static IDictionary ReadOnly(IDictionary d)
+        {
+            return new UnmodifiableDictionaryProxy(d);
+        }
+
+        public static IList ReadOnly(IList l)
+        {
+            return new UnmodifiableListProxy(l);
+        }
+
+        public static ISet ReadOnly(ISet s)
+        {
+            return new UnmodifiableSetProxy(s);
+        }
+
+        public static string ToString(
+			IEnumerable c)
+		{
+			StringBuilder sb = new StringBuilder("[");
+
+			IEnumerator e = c.GetEnumerator();
+
+			if (e.MoveNext())
+			{
+				sb.Append(e.Current.ToString());
+
+				while (e.MoveNext())
+				{
+					sb.Append(", ");
+					sb.Append(e.Current.ToString());
+				}
+			}
+
+			sb.Append(']');
+
+			return sb.ToString();
+		}
+	}
+}
diff --git a/Crypto/src/util/collections/EmptyEnumerable.cs b/Crypto/src/util/collections/EmptyEnumerable.cs
new file mode 100644
index 000000000..a61a0789a
--- /dev/null
+++ b/Crypto/src/util/collections/EmptyEnumerable.cs
@@ -0,0 +1,44 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public sealed class EmptyEnumerable
+		: IEnumerable
+	{
+		public static readonly IEnumerable Instance = new EmptyEnumerable();
+
+		private EmptyEnumerable()
+		{
+		}
+
+		public IEnumerator GetEnumerator()
+		{
+			return EmptyEnumerator.Instance;
+		}
+	}
+
+	public sealed class EmptyEnumerator
+		: IEnumerator
+	{
+		public static readonly IEnumerator Instance = new EmptyEnumerator();
+
+		private EmptyEnumerator()
+		{
+		}
+
+		public bool MoveNext()
+		{
+			return false;
+		}
+
+		public void Reset()
+		{
+		}
+
+		public object Current
+		{
+			get { throw new InvalidOperationException("No elements"); }
+		}
+	}
+}
diff --git a/Crypto/src/util/collections/EnumerableProxy.cs b/Crypto/src/util/collections/EnumerableProxy.cs
new file mode 100644
index 000000000..9eec4af21
--- /dev/null
+++ b/Crypto/src/util/collections/EnumerableProxy.cs
@@ -0,0 +1,25 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public sealed class EnumerableProxy
+		: IEnumerable
+	{
+		private readonly IEnumerable inner;
+
+		public EnumerableProxy(
+			IEnumerable inner)
+		{
+			if (inner == null)
+				throw new ArgumentNullException("inner");
+
+			this.inner = inner;
+		}
+
+		public IEnumerator GetEnumerator()
+		{
+			return inner.GetEnumerator();
+		}
+	}
+}
diff --git a/Crypto/src/util/collections/HashSet.cs b/Crypto/src/util/collections/HashSet.cs
new file mode 100644
index 000000000..1facb58e3
--- /dev/null
+++ b/Crypto/src/util/collections/HashSet.cs
@@ -0,0 +1,99 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public class HashSet
+		: ISet
+	{
+		private readonly IDictionary impl = Platform.CreateHashtable();
+
+		public HashSet()
+		{
+		}
+
+		public HashSet(IEnumerable s)
+		{
+			foreach (object o in s)
+			{
+				Add(o);
+			}
+		}
+
+		public virtual void Add(object o)
+		{
+			impl[o] = null;
+		}
+
+		public virtual void AddAll(IEnumerable e)
+		{
+			foreach (object o in e)
+			{
+				Add(o);
+			}
+		}
+
+		public virtual void Clear()
+		{
+			impl.Clear();
+		}
+
+		public virtual bool Contains(object o)
+		{
+			return impl.Contains(o);
+		}
+
+		public virtual void CopyTo(Array array, int index)
+		{
+			impl.Keys.CopyTo(array, index);
+		}
+
+		public virtual int Count
+		{
+			get { return impl.Count; }
+		}
+
+		public virtual IEnumerator GetEnumerator()
+		{
+			return impl.Keys.GetEnumerator();
+		}
+
+		public virtual bool IsEmpty
+		{
+			get { return impl.Count == 0; }
+		}
+
+		public virtual bool IsFixedSize
+		{
+			get { return impl.IsFixedSize; }
+		}
+
+		public virtual bool IsReadOnly
+		{
+			get { return impl.IsReadOnly; }
+		}
+
+		public virtual bool IsSynchronized
+		{
+			get { return impl.IsSynchronized; }
+		}
+
+		public virtual void Remove(object o)
+		{
+			impl.Remove(o);
+		}
+
+		public virtual void RemoveAll(IEnumerable e)
+		{
+			foreach (object o in e)
+			{
+				Remove(o);
+			}
+		}
+
+		public virtual object SyncRoot
+		{
+			get { return impl.SyncRoot; }
+		}
+	}
+}
diff --git a/Crypto/src/util/collections/ISet.cs b/Crypto/src/util/collections/ISet.cs
new file mode 100644
index 000000000..1f8edba40
--- /dev/null
+++ b/Crypto/src/util/collections/ISet.cs
@@ -0,0 +1,19 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public interface ISet
+		: ICollection
+	{
+		void Add(object o);
+		void AddAll(IEnumerable e);
+		void Clear();
+		bool Contains(object o);
+		bool IsEmpty { get; }
+		bool IsFixedSize { get; }
+		bool IsReadOnly { get; }
+		void Remove(object o);
+		void RemoveAll(IEnumerable e);
+	}
+}
diff --git a/Crypto/src/util/collections/LinkedDictionary.cs b/Crypto/src/util/collections/LinkedDictionary.cs
new file mode 100644
index 000000000..933d38ded
--- /dev/null
+++ b/Crypto/src/util/collections/LinkedDictionary.cs
@@ -0,0 +1,178 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public class LinkedDictionary
+		: IDictionary
+	{
+		internal readonly IDictionary hash = Platform.CreateHashtable();
+		internal readonly IList keys = Platform.CreateArrayList();
+
+		public LinkedDictionary()
+		{
+		}
+
+		public virtual void Add(object k, object v)
+		{
+			hash.Add(k, v);
+			keys.Add(k);
+		}
+
+		public virtual void Clear()
+		{
+			hash.Clear();
+			keys.Clear();
+		}
+
+		public virtual bool Contains(object k)
+		{
+			return hash.Contains(k);
+		}
+
+		public virtual void CopyTo(Array array, int index)
+		{
+			foreach (object k in keys)
+			{
+				array.SetValue(hash[k], index++);
+			}
+		}
+
+		public virtual int Count
+		{
+			get { return hash.Count; }
+		}
+
+		IEnumerator IEnumerable.GetEnumerator()
+		{
+			return GetEnumerator();
+		}
+
+		public virtual IDictionaryEnumerator GetEnumerator()
+		{
+			return new LinkedDictionaryEnumerator(this);
+		}
+
+		public virtual void Remove(object k)
+		{
+			hash.Remove(k);
+			keys.Remove(k);
+		}
+
+		public virtual bool IsFixedSize
+		{
+			get { return false; }
+		}
+
+		public virtual bool IsReadOnly
+		{
+			get { return false; }
+		}
+
+		public virtual bool IsSynchronized
+		{
+			get { return false; }
+		}
+
+		public virtual object SyncRoot
+		{
+			get { return false; }
+		}
+
+		public virtual ICollection Keys
+		{
+            get { return Platform.CreateArrayList(keys); }
+		}
+
+		public virtual ICollection Values
+		{
+			// NB: Order has to be the same as for Keys property
+			get
+			{
+                IList values = Platform.CreateArrayList(keys.Count);
+				foreach (object k in keys)
+				{
+					values.Add(hash[k]);
+				}
+				return values;
+			}
+		}
+
+		public virtual object this[object k]
+		{
+			get
+			{
+				return hash[k];
+			}
+			set
+			{
+				if (!hash.Contains(k))
+					keys.Add(k);
+				hash[k] = value;
+			}
+		}
+	}
+
+	internal class LinkedDictionaryEnumerator : IDictionaryEnumerator
+	{
+		private readonly LinkedDictionary parent;
+		private int pos = -1;
+
+		internal LinkedDictionaryEnumerator(LinkedDictionary parent)
+		{
+			this.parent = parent;
+		}
+
+		public virtual object Current
+		{
+			get { return Entry; }
+		}
+
+		public virtual DictionaryEntry Entry
+		{
+			get
+			{
+				object k = CurrentKey;
+				return new DictionaryEntry(k, parent.hash[k]);
+			}
+		}
+
+		public virtual object Key
+		{
+			get
+			{
+				return CurrentKey;
+			}
+		}
+
+		public virtual bool MoveNext()
+		{
+			if (pos >= parent.keys.Count)
+				return false;
+			return ++pos < parent.keys.Count;
+		}
+
+		public virtual void Reset()
+		{
+			this.pos = -1;
+		}
+
+		public virtual object Value
+		{
+			get
+			{
+				return parent.hash[CurrentKey];
+			}
+		}
+
+		private object CurrentKey
+		{
+			get
+			{
+				if (pos < 0 || pos >= parent.keys.Count)
+					throw new InvalidOperationException();
+				return parent.keys[pos];
+			}
+		}
+	}
+}
diff --git a/Crypto/src/util/collections/UnmodifiableDictionary.cs b/Crypto/src/util/collections/UnmodifiableDictionary.cs
new file mode 100644
index 000000000..0bdf70ad7
--- /dev/null
+++ b/Crypto/src/util/collections/UnmodifiableDictionary.cs
@@ -0,0 +1,64 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public abstract class UnmodifiableDictionary
+		: IDictionary
+	{
+		protected UnmodifiableDictionary()
+		{
+		}
+
+		public virtual void Add(object k, object v)
+		{
+			throw new NotSupportedException();
+		}
+
+		public virtual void Clear()
+		{
+			throw new NotSupportedException();
+		}
+
+		public abstract bool Contains(object k);
+
+		public abstract void CopyTo(Array array, int index);
+
+		public abstract int Count { get; }
+
+		IEnumerator IEnumerable.GetEnumerator()
+		{
+			return GetEnumerator();
+		}
+
+		public abstract IDictionaryEnumerator GetEnumerator();
+
+		public virtual void Remove(object k)
+		{
+			throw new NotSupportedException();
+		}
+
+		public abstract bool IsFixedSize { get; }
+
+		public virtual bool IsReadOnly
+		{
+			get { return true; }
+		}
+
+		public abstract bool IsSynchronized { get; }
+
+		public abstract object SyncRoot { get; }
+
+		public abstract ICollection Keys { get; }
+
+		public abstract ICollection Values { get; }
+
+		public virtual object this[object k]
+		{
+			get { return GetValue(k); }
+			set { throw new NotSupportedException(); }
+		}
+
+		protected abstract object GetValue(object k);
+	}
+}
diff --git a/Crypto/src/util/collections/UnmodifiableDictionaryProxy.cs b/Crypto/src/util/collections/UnmodifiableDictionaryProxy.cs
new file mode 100644
index 000000000..0fca909a3
--- /dev/null
+++ b/Crypto/src/util/collections/UnmodifiableDictionaryProxy.cs
@@ -0,0 +1,66 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public class UnmodifiableDictionaryProxy
+		: UnmodifiableDictionary
+	{
+		private readonly IDictionary d;
+
+		public UnmodifiableDictionaryProxy(IDictionary d)
+		{
+			this.d = d;
+		}
+
+		public override bool Contains(object k)
+		{
+			return d.Contains(k);
+		}
+
+		public override void CopyTo(Array array, int index)
+		{
+			d.CopyTo(array, index);
+		}
+
+		public override int Count
+		{
+			get { return d.Count; }
+		}
+
+		public override IDictionaryEnumerator GetEnumerator()
+		{
+			return d.GetEnumerator();
+		}
+
+		public override bool IsFixedSize
+		{
+			get { return d.IsFixedSize; }
+		}
+
+		public override bool IsSynchronized
+		{
+			get { return d.IsSynchronized; }
+		}
+
+		public override object SyncRoot
+		{
+			get { return d.SyncRoot; }
+		}
+
+		public override ICollection Keys
+		{
+			get { return d.Keys; }
+		}
+
+		public override ICollection Values
+		{
+			get { return d.Values; }
+		}
+
+		protected override object GetValue(object k)
+		{
+			return d[k];
+		}
+	}
+}
diff --git a/Crypto/src/util/collections/UnmodifiableList.cs b/Crypto/src/util/collections/UnmodifiableList.cs
new file mode 100644
index 000000000..28e49eac3
--- /dev/null
+++ b/Crypto/src/util/collections/UnmodifiableList.cs
@@ -0,0 +1,67 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public abstract class UnmodifiableList
+		: IList
+	{
+		protected UnmodifiableList()
+		{
+		}
+
+		public virtual int Add(object o)
+		{
+			throw new NotSupportedException();
+		}
+
+		public virtual void Clear()
+		{
+			throw new NotSupportedException();
+		}
+
+		public abstract bool Contains(object o);
+
+		public abstract void CopyTo(Array array, int index);
+
+		public abstract int Count { get; }
+
+		public abstract IEnumerator GetEnumerator();
+
+		public abstract int IndexOf(object o);
+
+		public virtual void Insert(int i, object o)
+		{
+			throw new NotSupportedException();
+		}
+
+		public abstract bool IsFixedSize { get; }
+
+		public virtual bool IsReadOnly
+		{
+			get { return true; }
+		}
+
+		public abstract bool IsSynchronized { get; }
+
+		public virtual void Remove(object o)
+		{
+			throw new NotSupportedException();
+		}
+
+		public virtual void RemoveAt(int i)
+		{
+			throw new NotSupportedException();
+		}
+
+		public abstract object SyncRoot { get; }
+		
+		public virtual object this[int i]
+		{
+			get { return GetValue(i); }
+			set { throw new NotSupportedException(); }
+		}
+
+		protected abstract object GetValue(int i);
+	}
+}
diff --git a/Crypto/src/util/collections/UnmodifiableListProxy.cs b/Crypto/src/util/collections/UnmodifiableListProxy.cs
new file mode 100644
index 000000000..9d00737ef
--- /dev/null
+++ b/Crypto/src/util/collections/UnmodifiableListProxy.cs
@@ -0,0 +1,61 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public class UnmodifiableListProxy
+		: UnmodifiableList
+	{
+		private readonly IList l;
+
+		public UnmodifiableListProxy(IList l)
+		{
+			this.l = l;
+		}
+
+		public override bool Contains(object o)
+		{
+			return l.Contains(o);
+		}
+
+		public override void CopyTo(Array array, int index)
+		{
+			l.CopyTo(array, index);
+		}
+
+		public override int Count
+		{
+			get { return l.Count; }
+		}
+
+		public override IEnumerator GetEnumerator()
+		{
+			return l.GetEnumerator();
+		}
+
+		public override int IndexOf(object o)
+		{
+			return l.IndexOf(o);
+		}
+
+		public override bool IsFixedSize
+		{
+			get { return l.IsFixedSize; }
+		}
+
+		public override bool IsSynchronized
+		{
+			get { return l.IsSynchronized; }
+		}
+
+		public override object SyncRoot
+		{
+			get { return l.SyncRoot; }
+		}
+
+		protected override object GetValue(int i)
+		{
+			return l[i];
+		}
+	}
+}
diff --git a/Crypto/src/util/collections/UnmodifiableSet.cs b/Crypto/src/util/collections/UnmodifiableSet.cs
new file mode 100644
index 000000000..8792815ac
--- /dev/null
+++ b/Crypto/src/util/collections/UnmodifiableSet.cs
@@ -0,0 +1,59 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public abstract class UnmodifiableSet
+		: ISet
+	{
+		protected UnmodifiableSet()
+		{
+		}
+
+		public virtual void Add(object o)
+		{
+			throw new NotSupportedException();
+		}
+
+		public virtual void AddAll(IEnumerable e)
+		{
+			throw new NotSupportedException();
+		}
+
+		public virtual void Clear()
+		{
+			throw new NotSupportedException();
+		}
+		
+		public abstract bool Contains(object o);
+
+		public abstract void CopyTo(Array array, int index);
+
+		public abstract int Count { get; }
+
+		public abstract IEnumerator GetEnumerator();
+
+		public abstract bool IsEmpty { get; }
+
+		public abstract bool IsFixedSize { get; }
+
+		public virtual bool IsReadOnly
+		{
+			get { return true; }
+		}
+
+		public abstract bool IsSynchronized { get; }
+
+		public abstract object SyncRoot { get; }
+
+		public virtual void Remove(object o)
+		{
+			throw new NotSupportedException();
+		}
+
+		public virtual void RemoveAll(IEnumerable e)
+		{
+			throw new NotSupportedException();
+		}
+	}
+}
diff --git a/Crypto/src/util/collections/UnmodifiableSetProxy.cs b/Crypto/src/util/collections/UnmodifiableSetProxy.cs
new file mode 100644
index 000000000..e119e2957
--- /dev/null
+++ b/Crypto/src/util/collections/UnmodifiableSetProxy.cs
@@ -0,0 +1,56 @@
+using System;
+using System.Collections;
+
+namespace Org.BouncyCastle.Utilities.Collections
+{
+	public class UnmodifiableSetProxy
+		: UnmodifiableSet
+	{
+		private readonly ISet s;
+
+		public UnmodifiableSetProxy (ISet s)
+		{
+			this.s = s;
+		}
+
+		public override bool Contains(object o)
+		{
+			return s.Contains(o);
+		}
+
+		public override void CopyTo(Array array, int index)
+		{
+			s.CopyTo(array, index);
+		}
+
+		public override int Count
+		{
+			get { return s.Count; }
+		}
+
+		public override IEnumerator GetEnumerator()
+		{
+			return s.GetEnumerator();
+		}
+
+		public override bool IsEmpty
+		{
+			get { return s.IsEmpty; }
+		}
+
+		public override bool IsFixedSize
+		{
+			get { return s.IsFixedSize; }
+		}
+
+		public override bool IsSynchronized
+		{
+			get { return s.IsSynchronized; }
+		}
+
+		public override object SyncRoot
+		{
+			get { return s.SyncRoot; }
+		}
+	}
+}
diff --git a/Crypto/src/util/date/DateTimeObject.cs b/Crypto/src/util/date/DateTimeObject.cs
new file mode 100644
index 000000000..793376b6d
--- /dev/null
+++ b/Crypto/src/util/date/DateTimeObject.cs
@@ -0,0 +1,25 @@
+using System;
+
+namespace Org.BouncyCastle.Utilities.Date
+{
+	public sealed class DateTimeObject
+	{
+		private readonly DateTime dt;
+
+		public DateTimeObject(
+			DateTime dt)
+		{
+			this.dt = dt;
+		}
+
+		public DateTime Value
+		{
+			get { return dt; }
+		}
+
+		public override string ToString()
+		{
+			return dt.ToString();
+		}
+	}
+}
diff --git a/Crypto/src/util/date/DateTimeUtilities.cs b/Crypto/src/util/date/DateTimeUtilities.cs
new file mode 100644
index 000000000..311ad5d37
--- /dev/null
+++ b/Crypto/src/util/date/DateTimeUtilities.cs
@@ -0,0 +1,47 @@
+using System;
+
+namespace Org.BouncyCastle.Utilities.Date
+{
+	public class DateTimeUtilities
+	{
+		public static readonly DateTime UnixEpoch = new DateTime(1970, 1, 1);
+
+		private DateTimeUtilities()
+		{
+		}
+
+		/// <summary>
+		/// Return the number of milliseconds since the Unix epoch (1 Jan., 1970 UTC) for a given DateTime value.
+		/// </summary>
+		/// <param name="dateTime">A UTC DateTime value not before epoch.</param>
+		/// <returns>Number of whole milliseconds after epoch.</returns>
+		/// <exception cref="ArgumentException">'dateTime' is before epoch.</exception>
+		public static long DateTimeToUnixMs(
+			DateTime dateTime)
+		{
+			if (dateTime.CompareTo(UnixEpoch) < 0)
+				throw new ArgumentException("DateTime value may not be before the epoch", "dateTime");
+
+			return (dateTime.Ticks - UnixEpoch.Ticks) / TimeSpan.TicksPerMillisecond;
+		}
+
+		/// <summary>
+		/// Create a DateTime value from the number of milliseconds since the Unix epoch (1 Jan., 1970 UTC).
+		/// </summary>
+		/// <param name="unixMs">Number of milliseconds since the epoch.</param>
+		/// <returns>A UTC DateTime value</returns>
+		public static DateTime UnixMsToDateTime(
+			long unixMs)
+		{
+			return new DateTime(unixMs * TimeSpan.TicksPerMillisecond + UnixEpoch.Ticks);
+		}
+
+		/// <summary>
+		/// Return the current number of milliseconds since the Unix epoch (1 Jan., 1970 UTC).
+		/// </summary>
+		public static long CurrentUnixMs()
+		{
+			return DateTimeToUnixMs(DateTime.UtcNow);
+		}
+	}
+}
diff --git a/Crypto/src/util/encoders/Base64.cs b/Crypto/src/util/encoders/Base64.cs
new file mode 100644
index 000000000..5dc9fae63
--- /dev/null
+++ b/Crypto/src/util/encoders/Base64.cs
@@ -0,0 +1,95 @@
+using System;
+using System.IO;
+using System.Text;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+	public sealed class Base64
+	{
+		private Base64()
+		{
+		}
+
+		/**
+		 * encode the input data producing a base 64 encoded byte array.
+		 *
+		 * @return a byte array containing the base 64 encoded data.
+		 */
+		public static byte[] Encode(
+			byte[] data)
+		{
+			string s = Convert.ToBase64String(data, 0, data.Length);
+            return Strings.ToAsciiByteArray(s);
+		}
+
+		/**
+		 * Encode the byte data to base 64 writing it to the given output stream.
+		 *
+		 * @return the number of bytes produced.
+		 */
+		public static int Encode(
+			byte[]	data,
+			Stream	outStream)
+		{
+			string s = Convert.ToBase64String(data, 0, data.Length);
+            byte[] encoded = Strings.ToAsciiByteArray(s);
+			outStream.Write(encoded, 0, encoded.Length);
+			return encoded.Length;
+		}
+
+		/**
+		 * Encode the byte data to base 64 writing it to the given output stream.
+		 *
+		 * @return the number of bytes produced.
+		 */
+		public static int Encode(
+			byte[]	data,
+			int		off,
+			int		length,
+			Stream	outStream)
+		{
+			string s = Convert.ToBase64String(data, off, length);
+            byte[] encoded = Strings.ToAsciiByteArray(s);
+			outStream.Write(encoded, 0, encoded.Length);
+			return encoded.Length;
+		}
+
+		/**
+		 * decode the base 64 encoded input data. It is assumed the input data is valid.
+		 *
+		 * @return a byte array representing the decoded data.
+		 */
+		public static byte[] Decode(
+			byte[] data)
+		{
+            string s = Strings.FromAsciiByteArray(data);
+			return Convert.FromBase64String(s);
+		}
+
+		/**
+		 * decode the base 64 encoded string data - whitespace will be ignored.
+		 *
+		 * @return a byte array representing the decoded data.
+		 */
+		public static byte[] Decode(
+			string data)
+		{
+			return Convert.FromBase64String(data);
+		}
+
+		/**
+		 * decode the base 64 encoded string data writing it to the given output stream,
+		 * whitespace characters will be ignored.
+		 *
+		 * @return the number of bytes produced.
+		 */
+		public static int Decode(
+			string	data,
+			Stream	outStream)
+		{
+			byte[] decoded = Decode(data);
+			outStream.Write(decoded, 0, decoded.Length);
+			return decoded.Length;
+		}
+	}
+}
diff --git a/Crypto/src/util/encoders/Base64Encoder.cs b/Crypto/src/util/encoders/Base64Encoder.cs
new file mode 100644
index 000000000..c94ce9d3c
--- /dev/null
+++ b/Crypto/src/util/encoders/Base64Encoder.cs
@@ -0,0 +1,307 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+	public class Base64Encoder
+		: IEncoder
+	{
+		protected readonly byte[] encodingTable =
+		{
+			(byte)'A', (byte)'B', (byte)'C', (byte)'D', (byte)'E', (byte)'F', (byte)'G',
+			(byte)'H', (byte)'I', (byte)'J', (byte)'K', (byte)'L', (byte)'M', (byte)'N',
+			(byte)'O', (byte)'P', (byte)'Q', (byte)'R', (byte)'S', (byte)'T', (byte)'U',
+			(byte)'V', (byte)'W', (byte)'X', (byte)'Y', (byte)'Z',
+			(byte)'a', (byte)'b', (byte)'c', (byte)'d', (byte)'e', (byte)'f', (byte)'g',
+			(byte)'h', (byte)'i', (byte)'j', (byte)'k', (byte)'l', (byte)'m', (byte)'n',
+			(byte)'o', (byte)'p', (byte)'q', (byte)'r', (byte)'s', (byte)'t', (byte)'u',
+			(byte)'v',
+			(byte)'w', (byte)'x', (byte)'y', (byte)'z',
+			(byte)'0', (byte)'1', (byte)'2', (byte)'3', (byte)'4', (byte)'5', (byte)'6',
+			(byte)'7', (byte)'8', (byte)'9',
+			(byte)'+', (byte)'/'
+		};
+
+		protected byte padding = (byte)'=';
+
+		/*
+		* set up the decoding table.
+		*/
+		protected readonly byte[] decodingTable = new byte[128];
+
+		protected void InitialiseDecodingTable()
+		{
+			for (int i = 0; i < encodingTable.Length; i++)
+			{
+				decodingTable[encodingTable[i]] = (byte)i;
+			}
+		}
+
+		public Base64Encoder()
+		{
+			InitialiseDecodingTable();
+		}
+
+		/**
+		* encode the input data producing a base 64 output stream.
+		*
+		* @return the number of bytes produced.
+		*/
+		public int Encode(
+			byte[]	data,
+			int		off,
+			int		length,
+			Stream	outStream)
+		{
+			int modulus = length % 3;
+			int dataLength = (length - modulus);
+			int a1, a2, a3;
+
+			for (int i = off; i < off + dataLength; i += 3)
+			{
+				a1 = data[i] & 0xff;
+				a2 = data[i + 1] & 0xff;
+				a3 = data[i + 2] & 0xff;
+
+				outStream.WriteByte(encodingTable[(int) ((uint) a1 >> 2) & 0x3f]);
+				outStream.WriteByte(encodingTable[((a1 << 4) | (int) ((uint) a2 >> 4)) & 0x3f]);
+				outStream.WriteByte(encodingTable[((a2 << 2) | (int) ((uint) a3 >> 6)) & 0x3f]);
+				outStream.WriteByte(encodingTable[a3 & 0x3f]);
+			}
+
+			/*
+			* process the tail end.
+			*/
+			int b1, b2, b3;
+			int d1, d2;
+
+			switch (modulus)
+			{
+				case 0:        /* nothing left to do */
+					break;
+				case 1:
+					d1 = data[off + dataLength] & 0xff;
+					b1 = (d1 >> 2) & 0x3f;
+					b2 = (d1 << 4) & 0x3f;
+
+					outStream.WriteByte(encodingTable[b1]);
+					outStream.WriteByte(encodingTable[b2]);
+					outStream.WriteByte(padding);
+					outStream.WriteByte(padding);
+					break;
+				case 2:
+					d1 = data[off + dataLength] & 0xff;
+					d2 = data[off + dataLength + 1] & 0xff;
+
+					b1 = (d1 >> 2) & 0x3f;
+					b2 = ((d1 << 4) | (d2 >> 4)) & 0x3f;
+					b3 = (d2 << 2) & 0x3f;
+
+					outStream.WriteByte(encodingTable[b1]);
+					outStream.WriteByte(encodingTable[b2]);
+					outStream.WriteByte(encodingTable[b3]);
+					outStream.WriteByte(padding);
+					break;
+			}
+
+			return (dataLength / 3) * 4 + ((modulus == 0) ? 0 : 4);
+		}
+
+		private bool ignore(
+			char c)
+		{
+			return (c == '\n' || c =='\r' || c == '\t' || c == ' ');
+		}
+
+		/**
+		* decode the base 64 encoded byte data writing it to the given output stream,
+		* whitespace characters will be ignored.
+		*
+		* @return the number of bytes produced.
+		*/
+		public int Decode(
+			byte[]	data,
+			int		off,
+			int		length,
+			Stream	outStream)
+		{
+			byte b1, b2, b3, b4;
+			int outLen = 0;
+
+			int end = off + length;
+
+			while (end > off)
+			{
+				if (!ignore((char)data[end - 1]))
+				{
+					break;
+				}
+
+				end--;
+			}
+
+			int  i = off;
+			int  finish = end - 4;
+
+			i = nextI(data, i, finish);
+
+			while (i < finish)
+			{
+				b1 = decodingTable[data[i++]];
+
+				i = nextI(data, i, finish);
+
+				b2 = decodingTable[data[i++]];
+
+				i = nextI(data, i, finish);
+
+				b3 = decodingTable[data[i++]];
+
+				i = nextI(data, i, finish);
+
+				b4 = decodingTable[data[i++]];
+
+				outStream.WriteByte((byte)((b1 << 2) | (b2 >> 4)));
+				outStream.WriteByte((byte)((b2 << 4) | (b3 >> 2)));
+				outStream.WriteByte((byte)((b3 << 6) | b4));
+
+				outLen += 3;
+
+				i = nextI(data, i, finish);
+			}
+
+			outLen += decodeLastBlock(outStream, (char)data[end - 4], (char)data[end - 3], (char)data[end - 2], (char)data[end - 1]);
+
+			return outLen;
+		}
+
+		private int nextI(
+			byte[]	data,
+			int		i,
+			int		finish)
+		{
+			while ((i < finish) && ignore((char)data[i]))
+			{
+				i++;
+			}
+			return i;
+		}
+
+		/**
+		* decode the base 64 encoded string data writing it to the given output stream,
+		* whitespace characters will be ignored.
+		*
+		* @return the number of bytes produced.
+		*/
+		public int DecodeString(
+			string	data,
+			Stream	outStream)
+		{
+			// Platform Implementation
+//			byte[] bytes = Convert.FromBase64String(data);
+//			outStream.Write(bytes, 0, bytes.Length);
+//			return bytes.Length;
+
+			byte b1, b2, b3, b4;
+			int length = 0;
+
+			int end = data.Length;
+
+			while (end > 0)
+			{
+				if (!ignore(data[end - 1]))
+				{
+					break;
+				}
+
+				end--;
+			}
+
+			int  i = 0;
+			int  finish = end - 4;
+
+			i = nextI(data, i, finish);
+
+			while (i < finish)
+			{
+				b1 = decodingTable[data[i++]];
+
+				i = nextI(data, i, finish);
+
+				b2 = decodingTable[data[i++]];
+
+				i = nextI(data, i, finish);
+
+				b3 = decodingTable[data[i++]];
+
+				i = nextI(data, i, finish);
+
+				b4 = decodingTable[data[i++]];
+
+				outStream.WriteByte((byte)((b1 << 2) | (b2 >> 4)));
+				outStream.WriteByte((byte)((b2 << 4) | (b3 >> 2)));
+				outStream.WriteByte((byte)((b3 << 6) | b4));
+
+				length += 3;
+
+				i = nextI(data, i, finish);
+			}
+
+			length += decodeLastBlock(outStream, data[end - 4], data[end - 3], data[end - 2], data[end - 1]);
+
+			return length;
+		}
+
+		private int decodeLastBlock(
+			Stream	outStream,
+			char	c1,
+			char	c2,
+			char	c3,
+			char	c4)
+		{
+			if (c3 == padding)
+			{
+				byte b1 = decodingTable[c1];
+				byte b2 = decodingTable[c2];
+
+				outStream.WriteByte((byte)((b1 << 2) | (b2 >> 4)));
+
+				return 1;
+			}
+
+			if (c4 == padding)
+			{
+				byte b1 = decodingTable[c1];
+				byte b2 = decodingTable[c2];
+				byte b3 = decodingTable[c3];
+
+				outStream.WriteByte((byte)((b1 << 2) | (b2 >> 4)));
+				outStream.WriteByte((byte)((b2 << 4) | (b3 >> 2)));
+
+				return 2;
+			}
+
+			{
+				byte b1 = decodingTable[c1];
+				byte b2 = decodingTable[c2];
+				byte b3 = decodingTable[c3];
+				byte b4 = decodingTable[c4];
+
+				outStream.WriteByte((byte)((b1 << 2) | (b2 >> 4)));
+				outStream.WriteByte((byte)((b2 << 4) | (b3 >> 2)));
+				outStream.WriteByte((byte)((b3 << 6) | b4));
+
+				return 3;
+			}
+		}
+
+		private int nextI(string data, int i, int finish)
+		{
+			while ((i < finish) && ignore(data[i]))
+			{
+				i++;
+			}
+			return i;
+		}
+	}
+}
diff --git a/Crypto/src/util/encoders/BufferedDecoder.cs b/Crypto/src/util/encoders/BufferedDecoder.cs
new file mode 100644
index 000000000..633cf1e97
--- /dev/null
+++ b/Crypto/src/util/encoders/BufferedDecoder.cs
@@ -0,0 +1,117 @@
+using System;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+    /// <summary>
+    ///  A buffering class to allow translation from one format to another to
+    ///     be done in discrete chunks.
+    /// </summary>
+    public class BufferedDecoder
+    {
+        internal byte[]        buffer;
+        internal int           bufOff;
+
+        internal ITranslator   translator;
+
+        /// <summary>
+        /// Create a buffered Decoder.
+        /// </summary>
+        /// <param name="translator">The translater to use.</param>
+        /// <param name="bufferSize">The size of the buffer.</param>
+        public BufferedDecoder(
+            ITranslator translator,
+            int         bufferSize)
+        {
+            this.translator = translator;
+
+            if ((bufferSize % translator.GetEncodedBlockSize()) != 0)
+            {
+                throw new ArgumentException("buffer size not multiple of input block size");
+            }
+
+            buffer = new byte[bufferSize];
+//            bufOff = 0;
+        }
+
+        /// <summary>
+        /// Process one byte of data.
+        /// </summary>
+        /// <param name="input">Data in.</param>
+        /// <param name="output">Byte array for the output.</param>
+        /// <param name="outOff">The offset in the output byte array to start writing from.</param>
+        /// <returns>The amount of output bytes.</returns>
+        public int ProcessByte(
+            byte        input,
+            byte[]      output,
+            int         outOff)
+        {
+            int         resultLen = 0;
+
+            buffer[bufOff++] = input;
+
+            if (bufOff == buffer.Length)
+            {
+                resultLen = translator.Decode(buffer, 0, buffer.Length, output, outOff);
+                bufOff = 0;
+            }
+
+            return resultLen;
+        }
+
+
+        /// <summary>
+        /// Process data from a byte array.
+        /// </summary>
+        /// <param name="input">The input data.</param>
+        /// <param name="inOff">Start position within input data array.</param>
+        /// <param name="len">Amount of data to process from input data array.</param>
+        /// <param name="outBytes">Array to store output.</param>
+        /// <param name="outOff">Position in output array to start writing from.</param>
+        /// <returns>The amount of output bytes.</returns>
+        public int ProcessBytes(
+            byte[]      input,
+            int         inOff,
+            int         len,
+            byte[]      outBytes,
+            int         outOff)
+        {
+            if (len < 0)
+            {
+            throw new ArgumentException("Can't have a negative input length!");
+            }
+
+            int resultLen = 0;
+            int gapLen = buffer.Length - bufOff;
+
+            if (len > gapLen)
+            {
+                Array.Copy(input, inOff, buffer, bufOff, gapLen);
+
+                resultLen += translator.Decode(buffer, 0, buffer.Length, outBytes, outOff);
+
+                bufOff = 0;
+
+                len -= gapLen;
+                inOff += gapLen;
+                outOff += resultLen;
+
+                int chunkSize = len - (len % buffer.Length);
+
+                resultLen += translator.Decode(input, inOff, chunkSize, outBytes, outOff);
+
+                len -= chunkSize;
+                inOff += chunkSize;
+            }
+
+            if (len != 0)
+            {
+                Array.Copy(input, inOff, buffer, bufOff, len);
+
+                bufOff += len;
+            }
+
+            return resultLen;
+        }
+    }
+
+}
diff --git a/Crypto/src/util/encoders/BufferedEncoder.cs b/Crypto/src/util/encoders/BufferedEncoder.cs
new file mode 100644
index 000000000..5c3b1ab46
--- /dev/null
+++ b/Crypto/src/util/encoders/BufferedEncoder.cs
@@ -0,0 +1,117 @@
+using System;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+    /// <summary>
+    /// A class that allows encoding of data using a specific encoder to be processed in chunks.
+    /// </summary>
+    public class BufferedEncoder
+    {
+        internal byte[]        Buffer;
+        internal int           bufOff;
+
+        internal ITranslator   translator;
+
+
+        /// <summary>
+        /// Create.
+        /// </summary>
+        /// <param name="translator">The translator to use.</param>
+        /// <param name="bufferSize">Size of the chunks.</param>
+        public BufferedEncoder(
+            ITranslator translator,
+            int         bufferSize)
+        {
+            this.translator = translator;
+
+            if ((bufferSize % translator.GetEncodedBlockSize()) != 0)
+            {
+                throw new ArgumentException("buffer size not multiple of input block size");
+            }
+
+            Buffer = new byte[bufferSize];
+//            bufOff = 0;
+        }
+
+
+        /// <summary>
+        /// Process one byte of data.
+        /// </summary>
+        /// <param name="input">The byte.</param>
+        /// <param name="outBytes">An array to store output in.</param>
+        /// <param name="outOff">Offset within output array to start writing from.</param>
+        /// <returns></returns>
+        public int ProcessByte(
+            byte        input,
+            byte[]      outBytes,
+            int         outOff)
+        {
+            int         resultLen = 0;
+
+            Buffer[bufOff++] = input;
+
+            if (bufOff == Buffer.Length)
+            {
+                resultLen = translator.Encode(Buffer, 0, Buffer.Length, outBytes, outOff);
+                bufOff = 0;
+            }
+
+            return resultLen;
+        }
+
+        /// <summary>
+        /// Process data from a byte array.
+        /// </summary>
+        /// <param name="input">Input data Byte array containing data to be processed.</param>
+        /// <param name="inOff">Start position within input data array.</param>
+        /// <param name="len">Amount of input data to be processed.</param>
+        /// <param name="outBytes">Output data array.</param>
+        /// <param name="outOff">Offset within output data array to start writing to.</param>
+        /// <returns>The amount of data written.</returns>
+        public int ProcessBytes(
+            byte[]      input,
+            int         inOff,
+            int         len,
+            byte[]      outBytes,
+            int         outOff)
+        {
+            if (len < 0)
+            {
+            throw new ArgumentException("Can't have a negative input length!");
+            }
+
+            int resultLen = 0;
+            int gapLen = Buffer.Length - bufOff;
+
+            if (len > gapLen)
+            {
+                Array.Copy(input, inOff, Buffer, bufOff, gapLen);
+
+                resultLen += translator.Encode(Buffer, 0, Buffer.Length, outBytes, outOff);
+
+                bufOff = 0;
+
+                len -= gapLen;
+                inOff += gapLen;
+                outOff += resultLen;
+
+                int chunkSize = len - (len % Buffer.Length);
+
+                resultLen += translator.Encode(input, inOff, chunkSize, outBytes, outOff);
+
+                len -= chunkSize;
+                inOff += chunkSize;
+            }
+
+            if (len != 0)
+            {
+                Array.Copy(input, inOff, Buffer, bufOff, len);
+
+                bufOff += len;
+            }
+
+            return resultLen;
+        }
+    }
+
+}
diff --git a/Crypto/src/util/encoders/Hex.cs b/Crypto/src/util/encoders/Hex.cs
new file mode 100644
index 000000000..fbe475991
--- /dev/null
+++ b/Crypto/src/util/encoders/Hex.cs
@@ -0,0 +1,131 @@
+using System;
+using System.IO;
+using System.Text;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+	/// <summary>
+	/// Class to decode and encode Hex.
+	/// </summary>
+	public sealed class Hex
+	{
+		private static readonly IEncoder encoder = new HexEncoder();
+
+		private Hex()
+		{
+		}
+
+		public static string ToHexString(
+			byte[] data)
+		{
+			byte[] hex = Encode(data, 0, data.Length);
+            return Strings.FromAsciiByteArray(hex);
+		}
+
+		public static string ToHexString(
+			byte[]	data,
+			int		off,
+			int		length)
+		{
+			byte[] hex = Encode(data, off, length);
+            return Strings.FromAsciiByteArray(hex);
+        }
+
+		/**
+		 * encode the input data producing a Hex encoded byte array.
+		 *
+		 * @return a byte array containing the Hex encoded data.
+		 */
+		public static byte[] Encode(
+			byte[] data)
+		{
+			return Encode(data, 0, data.Length);
+		}
+
+		/**
+		 * encode the input data producing a Hex encoded byte array.
+		 *
+		 * @return a byte array containing the Hex encoded data.
+		 */
+		public static byte[] Encode(
+			byte[]	data,
+			int		off,
+			int		length)
+		{
+			MemoryStream bOut = new MemoryStream(length * 2);
+
+			encoder.Encode(data, off, length, bOut);
+
+			return bOut.ToArray();
+		}
+
+		/**
+		 * Hex encode the byte data writing it to the given output stream.
+		 *
+		 * @return the number of bytes produced.
+		 */
+		public static int Encode(
+			byte[]	data,
+			Stream	outStream)
+		{
+			return encoder.Encode(data, 0, data.Length, outStream);
+		}
+
+		/**
+		 * Hex encode the byte data writing it to the given output stream.
+		 *
+		 * @return the number of bytes produced.
+		 */
+		public static int Encode(
+			byte[]	data,
+			int		off,
+			int		length,
+			Stream	outStream)
+		{
+			return encoder.Encode(data, off, length, outStream);
+		}
+
+		/**
+		 * decode the Hex encoded input data. It is assumed the input data is valid.
+		 *
+		 * @return a byte array representing the decoded data.
+		 */
+		public static byte[] Decode(
+			byte[] data)
+		{
+			MemoryStream bOut = new MemoryStream((data.Length + 1) / 2);
+
+			encoder.Decode(data, 0, data.Length, bOut);
+
+			return bOut.ToArray();
+		}
+
+		/**
+		 * decode the Hex encoded string data - whitespace will be ignored.
+		 *
+		 * @return a byte array representing the decoded data.
+		 */
+		public static byte[] Decode(
+			string data)
+		{
+			MemoryStream bOut = new MemoryStream((data.Length + 1) / 2);
+
+			encoder.DecodeString(data, bOut);
+
+			return bOut.ToArray();
+		}
+
+		/**
+		 * decode the Hex encoded string data writing it to the given output stream,
+		 * whitespace characters will be ignored.
+		 *
+		 * @return the number of bytes produced.
+		 */
+		public static int Decode(
+			string	data,
+			Stream	outStream)
+		{
+			return encoder.DecodeString(data, outStream);
+		}
+	}
+}
diff --git a/Crypto/src/util/encoders/HexEncoder.cs b/Crypto/src/util/encoders/HexEncoder.cs
new file mode 100644
index 000000000..c47d6219b
--- /dev/null
+++ b/Crypto/src/util/encoders/HexEncoder.cs
@@ -0,0 +1,164 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+	public class HexEncoder
+		: IEncoder
+	{
+		private static readonly byte[] encodingTable =
+		{
+			(byte)'0', (byte)'1', (byte)'2', (byte)'3', (byte)'4', (byte)'5', (byte)'6', (byte)'7',
+			(byte)'8', (byte)'9', (byte)'a', (byte)'b', (byte)'c', (byte)'d', (byte)'e', (byte)'f'
+		};
+
+		/*
+		* set up the decoding table.
+		*/
+		internal static readonly byte[] decodingTable = new byte[128];
+
+		static HexEncoder()
+		{
+			for (int i = 0; i < encodingTable.Length; i++)
+			{
+				decodingTable[encodingTable[i]] = (byte)i;
+			}
+
+			decodingTable['A'] = decodingTable['a'];
+			decodingTable['B'] = decodingTable['b'];
+			decodingTable['C'] = decodingTable['c'];
+			decodingTable['D'] = decodingTable['d'];
+			decodingTable['E'] = decodingTable['e'];
+			decodingTable['F'] = decodingTable['f'];
+		}
+
+		/**
+		* encode the input data producing a Hex output stream.
+		*
+		* @return the number of bytes produced.
+		*/
+		public int Encode(
+			byte[]	data,
+			int		off,
+			int		length,
+			Stream	outStream)
+		{
+			for (int i = off; i < (off + length); i++)
+			{
+				int v = data[i];
+
+				outStream.WriteByte(encodingTable[v >> 4]);
+				outStream.WriteByte(encodingTable[v & 0xf]);
+			}
+
+			return length * 2;
+		}
+
+		private bool ignore(
+			char c)
+		{
+			return (c == '\n' || c =='\r' || c == '\t' || c == ' ');
+		}
+
+		/**
+		* decode the Hex encoded byte data writing it to the given output stream,
+		* whitespace characters will be ignored.
+		*
+		* @return the number of bytes produced.
+		*/
+		public int Decode(
+			byte[]	data,
+			int		off,
+			int		length,
+			Stream	outStream)
+		{
+			byte b1, b2;
+			int outLen = 0;
+			int end = off + length;
+
+			while (end > off)
+			{
+				if (!ignore((char)data[end - 1]))
+				{
+					break;
+				}
+
+				end--;
+			}
+
+			int i = off;
+			while (i < end)
+			{
+				while (i < end && ignore((char)data[i]))
+				{
+					i++;
+				}
+
+				b1 = decodingTable[data[i++]];
+
+				while (i < end && ignore((char)data[i]))
+				{
+					i++;
+				}
+
+				b2 = decodingTable[data[i++]];
+
+				outStream.WriteByte((byte)((b1 << 4) | b2));
+
+				outLen++;
+			}
+
+			return outLen;
+		}
+
+		/**
+		* decode the Hex encoded string data writing it to the given output stream,
+		* whitespace characters will be ignored.
+		*
+		* @return the number of bytes produced.
+		*/
+		public int DecodeString(
+			string	data,
+			Stream	outStream)
+		{
+			byte    b1, b2;
+			int     length = 0;
+
+			int     end = data.Length;
+
+			while (end > 0)
+			{
+				if (!ignore(data[end - 1]))
+				{
+					break;
+				}
+
+				end--;
+			}
+
+			int i = 0;
+			while (i < end)
+			{
+				while (i < end && ignore(data[i]))
+				{
+					i++;
+				}
+
+				b1 = decodingTable[data[i++]];
+
+				while (i < end && ignore(data[i]))
+				{
+					i++;
+				}
+
+				b2 = decodingTable[data[i++]];
+
+				outStream.WriteByte((byte)((b1 << 4) | b2));
+
+				length++;
+			}
+
+			return length;
+		}
+	}
+}
diff --git a/Crypto/src/util/encoders/HexTranslator.cs b/Crypto/src/util/encoders/HexTranslator.cs
new file mode 100644
index 000000000..9775b6948
--- /dev/null
+++ b/Crypto/src/util/encoders/HexTranslator.cs
@@ -0,0 +1,108 @@
+using System;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+    /// <summary>
+    /// A hex translator.
+    /// </summary>
+    public class HexTranslator : ITranslator
+    {
+        private static readonly byte[]   hexTable =
+            {
+                (byte)'0', (byte)'1', (byte)'2', (byte)'3', (byte)'4', (byte)'5', (byte)'6', (byte)'7',
+                (byte)'8', (byte)'9', (byte)'a', (byte)'b', (byte)'c', (byte)'d', (byte)'e', (byte)'f'
+            };
+
+        /// <summary>
+        /// Return encoded block size.
+        /// </summary>
+        /// <returns>2</returns>
+        public int GetEncodedBlockSize()
+        {
+            return 2;
+        }
+
+        /// <summary>
+        /// Encode some data.
+        /// </summary>
+        /// <param name="input">Input data array.</param>
+        /// <param name="inOff">Start position within input data array.</param>
+        /// <param name="length">The amount of data to process.</param>
+        /// <param name="outBytes">The output data array.</param>
+        /// <param name="outOff">The offset within the output data array to start writing from.</param>
+        /// <returns>Amount of data encoded.</returns>
+        public int Encode(
+            byte[]  input,
+            int     inOff,
+            int     length,
+            byte[]  outBytes,
+            int     outOff)
+        {
+            for (int i = 0, j = 0; i < length; i++, j += 2)
+            {
+                outBytes[outOff + j] = hexTable[(input[inOff] >> 4) & 0x0f];
+                outBytes[outOff + j + 1] = hexTable[input[inOff] & 0x0f];
+
+                inOff++;
+            }
+
+            return length * 2;
+        }
+
+        /// <summary>
+        /// Returns the decoded block size.
+        /// </summary>
+        /// <returns>1</returns>
+        public int GetDecodedBlockSize()
+        {
+            return 1;
+        }
+
+        /// <summary>
+        /// Decode data from a byte array.
+        /// </summary>
+        /// <param name="input">The input data array.</param>
+        /// <param name="inOff">Start position within input data array.</param>
+        /// <param name="length">The amounty of data to process.</param>
+        /// <param name="outBytes">The output data array.</param>
+        /// <param name="outOff">The position within the output data array to start writing from.</param>
+        /// <returns>The amount of data written.</returns>
+        public int Decode(
+            byte[]  input,
+            int     inOff,
+            int     length,
+            byte[]  outBytes,
+            int     outOff)
+        {
+            int halfLength = length / 2;
+            byte left, right;
+            for (int i = 0; i < halfLength; i++)
+            {
+                left  = input[inOff + i * 2];
+                right = input[inOff + i * 2 + 1];
+
+                if (left < (byte)'a')
+                {
+                    outBytes[outOff] = (byte)((left - '0') << 4);
+                }
+                else
+                {
+                    outBytes[outOff] = (byte)((left - 'a' + 10) << 4);
+                }
+                if (right < (byte)'a')
+                {
+                    outBytes[outOff] += (byte)(right - '0');
+                }
+                else
+                {
+                    outBytes[outOff] += (byte)(right - 'a' + 10);
+                }
+
+                outOff++;
+            }
+
+            return halfLength;
+        }
+    }
+
+}
diff --git a/Crypto/src/util/encoders/IEncoder.cs b/Crypto/src/util/encoders/IEncoder.cs
new file mode 100644
index 000000000..5887d5daa
--- /dev/null
+++ b/Crypto/src/util/encoders/IEncoder.cs
@@ -0,0 +1,18 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+	/**
+	 * Encode and decode byte arrays (typically from binary to 7-bit ASCII
+	 * encodings).
+	 */
+	public interface IEncoder
+	{
+		int Encode(byte[] data, int off, int length, Stream outStream);
+
+		int Decode(byte[] data, int off, int length, Stream outStream);
+
+		int DecodeString(string data, Stream outStream);
+	}
+}
diff --git a/Crypto/src/util/encoders/Translator.cs b/Crypto/src/util/encoders/Translator.cs
new file mode 100644
index 000000000..10bd24b63
--- /dev/null
+++ b/Crypto/src/util/encoders/Translator.cs
@@ -0,0 +1,19 @@
+using System;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+    /// <summary>
+    /// Translator interface.
+    /// </summary>
+    public interface ITranslator
+    {
+        int GetEncodedBlockSize();
+
+        int Encode(byte[] input, int inOff, int length, byte[] outBytes, int outOff);
+
+        int GetDecodedBlockSize();
+
+        int Decode(byte[] input, int inOff, int length, byte[] outBytes, int outOff);
+    }
+
+}
diff --git a/Crypto/src/util/encoders/UrlBase64.cs b/Crypto/src/util/encoders/UrlBase64.cs
new file mode 100644
index 000000000..94195ef5e
--- /dev/null
+++ b/Crypto/src/util/encoders/UrlBase64.cs
@@ -0,0 +1,127 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+	/**
+	* Convert binary data to and from UrlBase64 encoding.  This is identical to
+	* Base64 encoding, except that the padding character is "." and the other 
+	* non-alphanumeric characters are "-" and "_" instead of "+" and "/".
+	* <p>
+	* The purpose of UrlBase64 encoding is to provide a compact encoding of binary
+	* data that is safe for use as an URL parameter. Base64 encoding does not
+	* produce encoded values that are safe for use in URLs, since "/" can be 
+	* interpreted as a path delimiter; "+" is the encoded form of a space; and
+	* "=" is used to separate a name from the corresponding value in an URL 
+	* parameter.
+	* </p>
+	*/
+	public class UrlBase64
+	{
+		private static readonly IEncoder encoder = new UrlBase64Encoder();
+
+		/**
+		* Encode the input data producing a URL safe base 64 encoded byte array.
+		*
+		* @return a byte array containing the URL safe base 64 encoded data.
+		*/
+		public static byte[] Encode(
+			byte[] data)
+		{
+			MemoryStream bOut = new MemoryStream();
+
+			try
+			{
+				encoder.Encode(data, 0, data.Length, bOut);
+			}
+			catch (IOException e)
+			{
+				throw new Exception("exception encoding URL safe base64 string: " + e.Message, e);
+			}
+
+			return bOut.ToArray();
+		}
+
+		/**
+		* Encode the byte data writing it to the given output stream.
+		*
+		* @return the number of bytes produced.
+		*/
+		public static int Encode(
+			byte[]	data,
+			Stream	outStr)
+		{
+			return encoder.Encode(data, 0, data.Length, outStr);
+		}
+
+		/**
+		* Decode the URL safe base 64 encoded input data - white space will be ignored.
+		*
+		* @return a byte array representing the decoded data.
+		*/
+		public static byte[] Decode(
+			byte[] data)
+		{
+			MemoryStream bOut = new MemoryStream();
+
+			try
+			{
+				encoder.Decode(data, 0, data.Length, bOut);
+			}
+			catch (IOException e)
+			{
+				throw new Exception("exception decoding URL safe base64 string: " + e.Message, e);
+			}
+
+			return bOut.ToArray();
+		}
+
+		/**
+		* decode the URL safe base 64 encoded byte data writing it to the given output stream,
+		* whitespace characters will be ignored.
+		*
+		* @return the number of bytes produced.
+		*/
+		public static int Decode(
+			byte[]	data,
+			Stream	outStr)
+		{
+			return encoder.Decode(data, 0, data.Length, outStr);
+		}
+
+		/**
+		* decode the URL safe base 64 encoded string data - whitespace will be ignored.
+		*
+		* @return a byte array representing the decoded data.
+		*/
+		public static byte[] Decode(
+			string data)
+		{
+			MemoryStream bOut = new MemoryStream();
+
+			try
+			{
+				encoder.DecodeString(data, bOut);
+			}
+			catch (IOException e)
+			{
+				throw new Exception("exception decoding URL safe base64 string: " + e.Message, e);
+			}
+	        
+			return bOut.ToArray();
+		}
+	    
+		/**
+		* Decode the URL safe base 64 encoded string data writing it to the given output stream,
+		* whitespace characters will be ignored.
+		*
+		* @return the number of bytes produced.
+		*/
+		public static int Decode(
+			string	data,
+			Stream	outStr)
+		{
+			return encoder.DecodeString(data, outStr);
+		}
+	}
+}
diff --git a/Crypto/src/util/encoders/UrlBase64Encoder.cs b/Crypto/src/util/encoders/UrlBase64Encoder.cs
new file mode 100644
index 000000000..5611a831c
--- /dev/null
+++ b/Crypto/src/util/encoders/UrlBase64Encoder.cs
@@ -0,0 +1,31 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.Encoders
+{
+	/**
+	* Convert binary data to and from UrlBase64 encoding.  This is identical to
+	* Base64 encoding, except that the padding character is "." and the other 
+	* non-alphanumeric characters are "-" and "_" instead of "+" and "/".
+	* <p>
+	* The purpose of UrlBase64 encoding is to provide a compact encoding of binary
+	* data that is safe for use as an URL parameter. Base64 encoding does not
+	* produce encoded values that are safe for use in URLs, since "/" can be 
+	* interpreted as a path delimiter; "+" is the encoded form of a space; and
+	* "=" is used to separate a name from the corresponding value in an URL 
+	* parameter.
+	* </p>
+	*/
+	public class UrlBase64Encoder
+		: Base64Encoder
+	{
+		public UrlBase64Encoder()
+		{
+			encodingTable[encodingTable.Length - 2] = (byte) '-';
+			encodingTable[encodingTable.Length - 1] = (byte) '_';
+			padding = (byte) '.';
+			// we must re-create the decoding table with the new encoded values.
+			InitialiseDecodingTable();
+		}
+	}
+}
\ No newline at end of file
diff --git a/Crypto/src/util/io/BaseInputStream.cs b/Crypto/src/util/io/BaseInputStream.cs
new file mode 100644
index 000000000..08eedb160
--- /dev/null
+++ b/Crypto/src/util/io/BaseInputStream.cs
@@ -0,0 +1,54 @@
+using System;
+using System.Diagnostics;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.IO
+{
+    public abstract class BaseInputStream : Stream
+    {
+		private bool closed;
+
+		public sealed override bool CanRead { get { return !closed; } }
+        public sealed override bool CanSeek { get { return false; } }
+        public sealed override bool CanWrite { get { return false; } }
+
+        protected override void Dispose(bool disposing)
+        {
+            base.Dispose(disposing);
+
+            closed = true;
+        }
+
+        public sealed override void Flush() {}
+        public sealed override long Length { get { throw new NotSupportedException(); } }
+        public sealed override long Position
+        {
+            get { throw new NotSupportedException(); }
+            set { throw new NotSupportedException(); }
+        }
+
+        public override int Read(byte[] buffer, int offset, int count)
+        {
+            int pos = offset;
+            try
+            {
+                int end = offset + count;
+                while (pos < end)
+                {
+                    int b = ReadByte();
+                    if (b == -1) break;
+                    buffer[pos++] = (byte) b;
+                }
+            }
+            catch (IOException)
+            {
+                if (pos == offset) throw;
+            }
+            return pos - offset;
+        }
+
+        public sealed override long Seek(long offset, SeekOrigin origin) { throw new NotSupportedException(); }
+        public sealed override void SetLength(long value) { throw new NotSupportedException(); }
+        public sealed override void Write(byte[] buffer, int offset, int count) { throw new NotSupportedException(); }
+    }
+}
diff --git a/Crypto/src/util/io/BaseOutputStream.cs b/Crypto/src/util/io/BaseOutputStream.cs
new file mode 100644
index 000000000..77233f68c
--- /dev/null
+++ b/Crypto/src/util/io/BaseOutputStream.cs
@@ -0,0 +1,54 @@
+using System;
+using System.Diagnostics;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.IO
+{
+    public abstract class BaseOutputStream : Stream
+    {
+		private bool closed;
+
+		public sealed override bool CanRead { get { return false; } }
+        public sealed override bool CanSeek { get { return false; } }
+        public sealed override bool CanWrite { get { return !closed; } }
+		
+        protected override void Dispose(bool disposing)
+        {
+            base.Dispose(disposing);
+
+            closed = true;
+        }
+        
+        public override void Flush() {}
+        public sealed override long Length { get { throw new NotSupportedException(); } }
+        public sealed override long Position
+        {
+            get { throw new NotSupportedException(); }
+            set { throw new NotSupportedException(); }
+        }
+        public sealed override int Read(byte[] buffer, int offset, int count) { throw new NotSupportedException(); }
+        public sealed override long Seek(long offset, SeekOrigin origin) { throw new NotSupportedException(); }
+        public sealed override void SetLength(long value) { throw new NotSupportedException(); }
+
+        public override void Write(byte[] buffer, int offset, int count)
+        {
+            Debug.Assert(buffer != null);
+            Debug.Assert(0 <= offset && offset <= buffer.Length);
+            Debug.Assert(count >= 0);
+
+            int end = offset + count;
+
+            Debug.Assert(0 <= end && end <= buffer.Length);
+
+            for (int i = offset; i < end; ++i)
+            {
+                this.WriteByte(buffer[i]);
+            }
+        }
+
+		public virtual void Write(params byte[] buffer)
+		{
+			Write(buffer, 0, buffer.Length);
+		}
+	}
+}
diff --git a/Crypto/src/util/io/PushbackStream.cs b/Crypto/src/util/io/PushbackStream.cs
new file mode 100644
index 000000000..954694259
--- /dev/null
+++ b/Crypto/src/util/io/PushbackStream.cs
@@ -0,0 +1,52 @@
+using System;
+using System.IO;
+
+using Org.BouncyCastle.Asn1.Utilities;
+
+namespace Org.BouncyCastle.Utilities.IO
+{
+	public class PushbackStream
+		: FilterStream
+	{
+		private int buf = -1;
+
+		public PushbackStream(
+			Stream s)
+			: base(s)
+		{
+		}
+
+		public override int ReadByte()
+		{
+			if (buf != -1)
+			{
+				int tmp = buf;
+				buf = -1;
+				return tmp;
+			}
+
+			return base.ReadByte();
+		}
+
+		public override int Read(byte[] buffer, int offset, int count)
+		{
+			if (buf != -1 && count > 0)
+			{
+				// TODO Can this case be made more efficient?
+				buffer[offset] = (byte) buf;
+				buf = -1;
+				return 1;
+			}
+
+			return base.Read(buffer, offset, count);
+		}
+
+		public virtual void Unread(int b)
+		{
+			if (buf != -1)
+				throw new InvalidOperationException("Can only push back one byte");
+
+			buf = b & 0xFF;
+		}
+	}
+}
diff --git a/Crypto/src/util/io/StreamOverflowException.cs b/Crypto/src/util/io/StreamOverflowException.cs
new file mode 100644
index 000000000..a8e7432fa
--- /dev/null
+++ b/Crypto/src/util/io/StreamOverflowException.cs
@@ -0,0 +1,27 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.IO
+{
+	public class StreamOverflowException
+		: IOException
+	{
+		public StreamOverflowException()
+			: base()
+		{
+		}
+
+		public StreamOverflowException(
+			string message)
+			: base(message)
+		{
+		}
+
+		public StreamOverflowException(
+			string		message,
+			Exception	exception)
+			: base(message, exception)
+		{
+		}
+	}
+}
diff --git a/Crypto/src/util/io/Streams.cs b/Crypto/src/util/io/Streams.cs
new file mode 100644
index 000000000..ee95d3b01
--- /dev/null
+++ b/Crypto/src/util/io/Streams.cs
@@ -0,0 +1,94 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.IO
+{
+	public sealed class Streams
+	{
+		private const int BufferSize = 512;
+
+		private Streams()
+		{
+		}
+
+		public static void Drain(Stream inStr)
+		{
+			byte[] bs = new byte[BufferSize];
+			while (inStr.Read(bs, 0, bs.Length) > 0)
+			{
+			}
+		}
+
+		public static byte[] ReadAll(Stream inStr)
+		{
+			MemoryStream buf = new MemoryStream();
+			PipeAll(inStr, buf);
+			return buf.ToArray();
+		}
+
+		public static byte[] ReadAllLimited(Stream inStr, int limit)
+		{
+			MemoryStream buf = new MemoryStream();
+			PipeAllLimited(inStr, limit, buf);
+			return buf.ToArray();
+		}
+
+		public static int ReadFully(Stream inStr, byte[] buf)
+		{
+			return ReadFully(inStr, buf, 0, buf.Length);
+		}
+
+		public static int ReadFully(Stream inStr, byte[] buf, int off, int len)
+		{
+			int totalRead = 0;
+			while (totalRead < len)
+			{
+				int numRead = inStr.Read(buf, off + totalRead, len - totalRead);
+				if (numRead < 1)
+					break;
+				totalRead += numRead;
+			}
+			return totalRead;
+		}
+
+		public static void PipeAll(Stream inStr, Stream outStr)
+		{
+			byte[] bs = new byte[BufferSize];
+			int numRead;
+			while ((numRead = inStr.Read(bs, 0, bs.Length)) > 0)
+			{
+				outStr.Write(bs, 0, numRead);
+			}
+		}
+
+		/// <summary>
+		/// Pipe all bytes from <c>inStr</c> to <c>outStr</c>, throwing <c>StreamFlowException</c> if greater
+		/// than <c>limit</c> bytes in <c>inStr</c>.
+		/// </summary>
+		/// <param name="inStr">
+		/// A <see cref="Stream"/>
+		/// </param>
+		/// <param name="limit">
+		/// A <see cref="System.Int64"/>
+		/// </param>
+		/// <param name="outStr">
+		/// A <see cref="Stream"/>
+		/// </param>
+		/// <returns>The number of bytes actually transferred, if not greater than <c>limit</c></returns>
+		/// <exception cref="IOException"></exception>
+		public static long PipeAllLimited(Stream inStr, long limit, Stream outStr)
+		{
+			byte[] bs = new byte[BufferSize];
+			long total = 0;
+			int numRead;
+			while ((numRead = inStr.Read(bs, 0, bs.Length)) > 0)
+			{
+				total += numRead;
+				if (total > limit)
+					throw new StreamOverflowException("Data Overflow");
+				outStr.Write(bs, 0, numRead);
+			}
+			return total;
+		}
+	}
+}
diff --git a/Crypto/src/util/io/TeeInputStream.cs b/Crypto/src/util/io/TeeInputStream.cs
new file mode 100644
index 000000000..fed9823f0
--- /dev/null
+++ b/Crypto/src/util/io/TeeInputStream.cs
@@ -0,0 +1,54 @@
+using System;
+using System.Diagnostics;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.IO
+{
+	public class TeeInputStream
+		: BaseInputStream
+	{
+		private readonly Stream input, tee;
+
+		public TeeInputStream(Stream input, Stream tee)
+		{
+			Debug.Assert(input.CanRead);
+			Debug.Assert(tee.CanWrite);
+
+			this.input = input;
+			this.tee = tee;
+		}
+
+        protected override void Dispose(bool disposing)
+        {
+            if (disposing)
+            {
+                input.Dispose();
+                tee.Dispose();
+            }
+        }
+
+		public override int Read(byte[] buf, int off, int len)
+		{
+			int i = input.Read(buf, off, len);
+
+			if (i > 0)
+			{
+				tee.Write(buf, off, i);
+			}
+
+			return i;
+		}
+
+		public override int ReadByte()
+		{
+			int i = input.ReadByte();
+
+			if (i >= 0)
+			{
+				tee.WriteByte((byte)i);
+			}
+
+			return i;
+		}
+	}
+}
diff --git a/Crypto/src/util/io/TeeOutputStream.cs b/Crypto/src/util/io/TeeOutputStream.cs
new file mode 100644
index 000000000..965ef23c8
--- /dev/null
+++ b/Crypto/src/util/io/TeeOutputStream.cs
@@ -0,0 +1,42 @@
+using System;
+using System.Diagnostics;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.IO
+{
+    public class TeeOutputStream
+		: BaseOutputStream
+	{
+		private readonly Stream output, tee;
+
+		public TeeOutputStream(Stream output, Stream tee)
+		{
+			Debug.Assert(output.CanWrite);
+			Debug.Assert(tee.CanWrite);
+
+			this.output = output;
+			this.tee = tee;
+		}
+
+        protected override void Dispose(bool disposing)
+        {
+            if (disposing)
+            {
+                output.Dispose();
+                tee.Dispose();
+            }
+        }
+
+		public override void Write(byte[] buffer, int offset, int count)
+		{
+			output.Write(buffer, offset, count);
+			tee.Write(buffer, offset, count);
+		}
+
+		public override void WriteByte(byte b)
+		{
+			output.WriteByte(b);
+			tee.WriteByte(b);
+		}
+	}
+}
diff --git a/Crypto/src/util/io/pem/PemGenerationException.cs b/Crypto/src/util/io/pem/PemGenerationException.cs
new file mode 100644
index 000000000..22e83ac2d
--- /dev/null
+++ b/Crypto/src/util/io/pem/PemGenerationException.cs
@@ -0,0 +1,26 @@
+using System;
+
+namespace Org.BouncyCastle.Utilities.IO.Pem
+{
+	public class PemGenerationException
+		: Exception
+	{
+		public PemGenerationException()
+			: base()
+		{
+		}
+
+		public PemGenerationException(
+			string message)
+			: base(message)
+		{
+		}
+
+		public PemGenerationException(
+			string		message,
+			Exception	exception)
+			: base(message, exception)
+		{
+		}
+	}
+}
diff --git a/Crypto/src/util/io/pem/PemHeader.cs b/Crypto/src/util/io/pem/PemHeader.cs
new file mode 100644
index 000000000..72da8a4f7
--- /dev/null
+++ b/Crypto/src/util/io/pem/PemHeader.cs
@@ -0,0 +1,55 @@
+using System;
+
+namespace Org.BouncyCastle.Utilities.IO.Pem
+{
+	public class PemHeader
+	{
+		private string name;
+		private string val;
+
+		public PemHeader(string name, string val)
+		{
+			this.name = name;
+			this.val = val;
+		}
+
+		public virtual string Name
+		{
+			get { return name; }
+		}
+
+		public virtual string Value
+		{
+			get { return val; }
+		}
+
+		public override int GetHashCode()
+		{
+			return GetHashCode(this.name) + 31 * GetHashCode(this.val);
+		}
+
+		public override bool Equals(object obj)
+		{
+			if (obj == this)
+				return true;
+
+			if (!(obj is PemHeader))
+				return false;
+
+			PemHeader other = (PemHeader)obj;
+
+			return Platform.Equals(this.name, other.name)
+				&& Platform.Equals(this.val, other.val);
+		}
+
+		private int GetHashCode(string s)
+		{
+			if (s == null)
+			{
+				return 1;
+			}
+
+			return s.GetHashCode();
+		}
+	}
+}
diff --git a/Crypto/src/util/io/pem/PemObject.cs b/Crypto/src/util/io/pem/PemObject.cs
new file mode 100644
index 000000000..41212f997
--- /dev/null
+++ b/Crypto/src/util/io/pem/PemObject.cs
@@ -0,0 +1,47 @@
+using System;
+using System.Collections;
+
+using Org.BouncyCastle.Utilities.Collections;
+
+namespace Org.BouncyCastle.Utilities.IO.Pem
+{
+	public class PemObject
+		: PemObjectGenerator
+	{
+		private string		type;
+		private IList		headers;
+		private byte[]		content;
+
+		public PemObject(string type, byte[] content)
+			: this(type, Platform.CreateArrayList(), content)
+		{
+		}
+
+		public PemObject(String type, IList headers, byte[] content)
+		{
+			this.type = type;
+            this.headers = Platform.CreateArrayList(headers);
+			this.content = content;
+		}
+
+		public string Type
+		{
+			get { return type; }
+		}
+
+		public IList Headers
+		{
+			get { return headers; }
+		}
+
+		public byte[] Content
+		{
+			get { return content; }
+		}
+
+		public PemObject Generate()
+		{
+			return this;
+		}
+	}
+}
diff --git a/Crypto/src/util/io/pem/PemObjectGenerator.cs b/Crypto/src/util/io/pem/PemObjectGenerator.cs
new file mode 100644
index 000000000..6f9bfc191
--- /dev/null
+++ b/Crypto/src/util/io/pem/PemObjectGenerator.cs
@@ -0,0 +1,13 @@
+using System;
+
+namespace Org.BouncyCastle.Utilities.IO.Pem
+{
+	public interface PemObjectGenerator
+	{
+		/// <returns>
+		/// A <see cref="PemObject"/>
+		/// </returns>
+		/// <exception cref="PemGenerationException"></exception>
+		PemObject Generate();
+	}
+}
diff --git a/Crypto/src/util/io/pem/PemObjectParser.cs b/Crypto/src/util/io/pem/PemObjectParser.cs
new file mode 100644
index 000000000..91d26dc3a
--- /dev/null
+++ b/Crypto/src/util/io/pem/PemObjectParser.cs
@@ -0,0 +1,17 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.IO.Pem
+{
+	public interface PemObjectParser
+	{
+		/// <param name="obj">
+		/// A <see cref="PemObject"/>
+		/// </param>
+		/// <returns>
+		/// A <see cref="System.Object"/>
+		/// </returns>
+		/// <exception cref="IOException"></exception>
+		object ParseObject(PemObject obj);
+	}
+}
diff --git a/Crypto/src/util/io/pem/PemReader.cs b/Crypto/src/util/io/pem/PemReader.cs
new file mode 100644
index 000000000..b3284705d
--- /dev/null
+++ b/Crypto/src/util/io/pem/PemReader.cs
@@ -0,0 +1,94 @@
+using System;
+using System.Collections;
+using System.IO;
+using System.Text;
+
+using Org.BouncyCastle.Utilities.Encoders;
+
+namespace Org.BouncyCastle.Utilities.IO.Pem
+{
+	public class PemReader
+	{
+		private const string BeginString = "-----BEGIN ";
+		private const string EndString = "-----END ";
+
+		private readonly TextReader reader;
+
+		public PemReader(TextReader reader)
+		{
+			if (reader == null)
+				throw new ArgumentNullException("reader");
+
+			this.reader = reader;
+		}
+
+		public TextReader Reader
+		{
+			get { return reader; }
+		}
+
+		/// <returns>
+		/// A <see cref="PemObject"/>
+		/// </returns>
+		/// <exception cref="IOException"></exception>
+		public PemObject ReadPemObject()
+		{
+			string line = reader.ReadLine();
+
+			if (line != null && line.StartsWith(BeginString))
+			{
+				line = line.Substring(BeginString.Length);
+				int index = line.IndexOf('-');
+				string type = line.Substring(0, index);
+
+				if (index > 0)
+					return LoadObject(type);
+			}
+
+			return null;
+		}
+
+		private PemObject LoadObject(string type)
+		{
+			string endMarker = EndString + type;
+			IList headers = Platform.CreateArrayList();
+			StringBuilder buf = new StringBuilder();
+
+			string line;
+			while ((line = reader.ReadLine()) != null
+				&& line.IndexOf(endMarker) == -1)
+			{
+				int colonPos = line.IndexOf(':');
+
+				if (colonPos == -1)
+				{
+					buf.Append(line.Trim());
+				}
+				else
+				{
+					// Process field
+					string fieldName = line.Substring(0, colonPos).Trim();
+
+					if (fieldName.StartsWith("X-"))
+						fieldName = fieldName.Substring(2);
+
+					string fieldValue = line.Substring(colonPos + 1).Trim();
+
+					headers.Add(new PemHeader(fieldName, fieldValue));
+				}
+			}
+
+			if (line == null)
+			{
+				throw new IOException(endMarker + " not found");
+			}
+
+			if (buf.Length % 4 != 0)
+			{
+				throw new IOException("base64 data appears to be truncated");
+			}
+
+			return new PemObject(type, headers, Base64.Decode(buf.ToString()));
+		}
+	}
+}
diff --git a/Crypto/src/util/io/pem/PemWriter.cs b/Crypto/src/util/io/pem/PemWriter.cs
new file mode 100644
index 000000000..e85b31543
--- /dev/null
+++ b/Crypto/src/util/io/pem/PemWriter.cs
@@ -0,0 +1,120 @@
+using System;
+using System.Collections;
+using System.IO;
+
+using Org.BouncyCastle.Utilities.Encoders;
+
+namespace Org.BouncyCastle.Utilities.IO.Pem
+{
+	/**
+	* A generic PEM writer, based on RFC 1421
+	*/
+	public class PemWriter
+	{
+		private const int LineLength = 64;
+
+		private readonly TextWriter	writer;
+		private readonly int		nlLength;
+		private char[]				buf = new char[LineLength];
+		
+		/**
+		 * Base constructor.
+		 *
+		 * @param out output stream to use.
+		 */
+		public PemWriter(TextWriter writer)
+		{
+			if (writer == null)
+				throw new ArgumentNullException("writer");
+
+			this.writer = writer;
+			this.nlLength = Platform.NewLine.Length;
+		}
+
+		public TextWriter Writer
+		{
+			get { return writer; }
+		}
+
+		/**
+		 * Return the number of bytes or characters required to contain the
+		 * passed in object if it is PEM encoded.
+		 *
+		 * @param obj pem object to be output
+		 * @return an estimate of the number of bytes
+		 */
+		public int GetOutputSize(PemObject obj)
+		{
+			// BEGIN and END boundaries.
+			int size = (2 * (obj.Type.Length + 10 + nlLength)) + 6 + 4;
+
+			if (obj.Headers.Count > 0)
+			{
+				foreach (PemHeader header in obj.Headers)
+				{
+					size += header.Name.Length + ": ".Length + header.Value.Length + nlLength;
+				}
+
+				size += nlLength;
+			}
+
+			// base64 encoding
+			int dataLen = ((obj.Content.Length + 2) / 3) * 4;
+
+			size += dataLen + (((dataLen + LineLength - 1) / LineLength) * nlLength);
+
+			return size;
+		}
+
+		public void WriteObject(PemObjectGenerator objGen)
+		{
+			PemObject obj = objGen.Generate();
+
+			WritePreEncapsulationBoundary(obj.Type);
+
+			if (obj.Headers.Count > 0)
+			{
+				foreach (PemHeader header in obj.Headers)
+				{
+					writer.Write(header.Name);
+					writer.Write(": ");
+					writer.WriteLine(header.Value);
+				}
+
+				writer.WriteLine();
+			}
+
+			WriteEncoded(obj.Content);
+			WritePostEncapsulationBoundary(obj.Type);
+		}
+
+		private void WriteEncoded(byte[] bytes)
+		{
+			bytes = Base64.Encode(bytes);
+
+			for (int i = 0; i < bytes.Length; i += buf.Length)
+			{
+				int index = 0;
+				while (index != buf.Length)
+				{
+					if ((i + index) >= bytes.Length)
+						break;
+
+					buf[index] = (char)bytes[i + index];
+					index++;
+				}
+				writer.WriteLine(buf, 0, index);
+			}
+		}
+
+		private void WritePreEncapsulationBoundary(string type)
+		{
+			writer.WriteLine("-----BEGIN " + type + "-----");
+		}
+
+		private void WritePostEncapsulationBoundary(string type)
+		{
+			writer.WriteLine("-----END " + type + "-----");
+		}
+	}
+}
diff --git a/Crypto/src/util/net/IPAddress.cs b/Crypto/src/util/net/IPAddress.cs
new file mode 100644
index 000000000..2a30a15f0
--- /dev/null
+++ b/Crypto/src/util/net/IPAddress.cs
@@ -0,0 +1,197 @@
+using System;
+using System.Globalization;
+
+using Org.BouncyCastle.Math;
+
+namespace Org.BouncyCastle.Utilities.Net
+{
+	public class IPAddress
+	{
+		/**
+		 * Validate the given IPv4 or IPv6 address.
+		 *
+		 * @param address the IP address as a string.
+		 *
+		 * @return true if a valid address, false otherwise
+		 */
+		public static bool IsValid(
+			string address)
+		{
+			return IsValidIPv4(address) || IsValidIPv6(address);
+		}
+
+		/**
+		 * Validate the given IPv4 or IPv6 address and netmask.
+		 *
+		 * @param address the IP address as a string.
+		 *
+		 * @return true if a valid address with netmask, false otherwise
+		 */
+		public static bool IsValidWithNetMask(
+			string address)
+		{
+			return IsValidIPv4WithNetmask(address) || IsValidIPv6WithNetmask(address);
+		}
+
+		/**
+		 * Validate the given IPv4 address.
+		 * 
+		 * @param address the IP address as a string.
+		 *
+		 * @return true if a valid IPv4 address, false otherwise
+		 */
+		public static bool IsValidIPv4(
+			string address)
+		{
+			try
+			{
+				return unsafeIsValidIPv4(address);
+			}
+			catch (FormatException) {}
+			catch (OverflowException) {}
+			return false;
+		}
+
+		private static bool unsafeIsValidIPv4(
+			string address)
+		{
+			if (address.Length == 0)
+				return false;
+
+			int octets = 0;
+			string temp = address + ".";
+
+			int pos;
+			int start = 0;
+			while (start < temp.Length
+				&& (pos = temp.IndexOf('.', start)) > start)
+			{
+				if (octets == 4)
+					return false;
+
+				string octetStr = temp.Substring(start, pos - start);
+				int octet = Int32.Parse(octetStr);
+
+				if (octet < 0 || octet > 255)
+					return false;
+
+				start = pos + 1;
+				octets++;
+			}
+
+			return octets == 4;
+		}
+
+		public static bool IsValidIPv4WithNetmask(
+			string address)
+		{
+			int index = address.IndexOf("/");
+			string mask = address.Substring(index + 1);
+
+			return (index > 0) && IsValidIPv4(address.Substring(0, index))
+				&& (IsValidIPv4(mask) || IsMaskValue(mask, 32));
+		}
+
+		public static bool IsValidIPv6WithNetmask(
+			string address)
+		{
+			int index = address.IndexOf("/");
+			string mask = address.Substring(index + 1);
+
+			return (index > 0) && (IsValidIPv6(address.Substring(0, index))
+				&& (IsValidIPv6(mask) || IsMaskValue(mask, 128)));
+		}
+
+		private static bool IsMaskValue(
+			string	component,
+			int		size)
+		{
+			int val = Int32.Parse(component);
+			try
+			{
+				return val >= 0 && val <= size;
+			}
+			catch (FormatException) {}
+			catch (OverflowException) {}
+			return false;
+		}
+
+		/**
+		 * Validate the given IPv6 address.
+		 *
+		 * @param address the IP address as a string.
+		 *
+		 * @return true if a valid IPv4 address, false otherwise
+		 */
+		public static bool IsValidIPv6(
+			string address)
+		{
+			try
+			{
+				return unsafeIsValidIPv6(address);
+			}
+			catch (FormatException) {}
+			catch (OverflowException) {}
+			return false;
+		}
+
+		private static bool unsafeIsValidIPv6(
+			string address)
+		{
+			if (address.Length == 0)
+			{
+				return false;
+			}
+
+			int octets = 0;
+
+			string temp = address + ":";
+			bool doubleColonFound = false;
+			int pos;
+			int start = 0;
+			while (start < temp.Length
+				&& (pos = temp.IndexOf(':', start)) >= start)
+			{
+				if (octets == 8)
+				{
+					return false;
+				}
+
+				if (start != pos)
+				{
+					string value = temp.Substring(start, pos - start);
+
+					if (pos == (temp.Length - 1) && value.IndexOf('.') > 0)
+					{
+						if (!IsValidIPv4(value))
+						{
+							return false;
+						}
+
+						octets++; // add an extra one as address covers 2 words.
+					}
+					else
+					{
+						string octetStr = temp.Substring(start, pos - start);
+						int octet = Int32.Parse(octetStr, NumberStyles.AllowHexSpecifier);
+
+						if (octet < 0 || octet > 0xffff)
+							return false;
+					}
+				}
+				else
+				{
+					if (pos != 1 && pos != temp.Length - 1 && doubleColonFound)
+					{
+						return false;
+					}
+					doubleColonFound = true;
+				}
+				start = pos + 1;
+				octets++;
+			}
+
+			return octets == 8 || doubleColonFound;
+		}
+	}
+}
diff --git a/Crypto/src/util/zlib/Adler32.cs b/Crypto/src/util/zlib/Adler32.cs
new file mode 100644
index 000000000..c38258f2a
--- /dev/null
+++ b/Crypto/src/util/zlib/Adler32.cs
@@ -0,0 +1,88 @@
+using System;
+/*
+ * $Id: Adler32.cs,v 1.1 2006-07-31 13:59:25 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    internal sealed class Adler32{
+
+        // largest prime smaller than 65536
+        private const int BASE=65521; 
+        // NMAX is the largest n such that 255n(n+1)/2 + (n+1)(BASE-1) <= 2^32-1
+        private const int NMAX=5552;
+
+        internal long adler32(long adler, byte[] buf, int index, int len){
+            if(buf == null){ return 1L; }
+
+            long s1=adler&0xffff;
+            long s2=(adler>>16)&0xffff;
+            int k;
+
+            while(len > 0) {
+                k=len<NMAX?len:NMAX;
+                len-=k;
+                while(k>=16){
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    s1+=buf[index++]&0xff; s2+=s1;
+                    k-=16;
+                }
+                if(k!=0){
+                    do{
+                        s1+=buf[index++]&0xff; s2+=s1;
+                    }
+                    while(--k!=0);
+                }
+                s1%=BASE;
+                s2%=BASE;
+            }
+            return (s2<<16)|s1;
+        }
+
+    }
+}
\ No newline at end of file
diff --git a/Crypto/src/util/zlib/Deflate.cs b/Crypto/src/util/zlib/Deflate.cs
new file mode 100644
index 000000000..ca0430939
--- /dev/null
+++ b/Crypto/src/util/zlib/Deflate.cs
@@ -0,0 +1,1640 @@
+using System;
+/*
+ * $Id: Deflate.cs,v 1.2 2008-05-10 09:35:40 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    public sealed class Deflate{
+
+        private const int MAX_MEM_LEVEL=9;
+
+        private const int Z_DEFAULT_COMPRESSION=-1;
+
+        private const int MAX_WBITS=15;            // 32K LZ77 window
+        private const int DEF_MEM_LEVEL=8;
+
+        internal class Config{
+            internal int good_length; // reduce lazy search above this match length
+            internal int max_lazy;    // do not perform lazy search above this match length
+            internal int nice_length; // quit search above this match length
+            internal int max_chain;
+            internal int func;
+            internal Config(int good_length, int max_lazy, 
+                int nice_length, int max_chain, int func){
+                this.good_length=good_length;
+                this.max_lazy=max_lazy;
+                this.nice_length=nice_length;
+                this.max_chain=max_chain;
+                this.func=func;
+            }
+        }
+  
+        private const int STORED=0;
+        private const int FAST=1;
+        private const int SLOW=2;
+        private static readonly Config[] config_table;
+
+		static Deflate(){
+            config_table=new Config[10];
+            //                         good  lazy  nice  chain
+            config_table[0]=new Config(0,    0,    0,    0, STORED);
+            config_table[1]=new Config(4,    4,    8,    4, FAST);
+            config_table[2]=new Config(4,    5,   16,    8, FAST);
+            config_table[3]=new Config(4,    6,   32,   32, FAST);
+
+            config_table[4]=new Config(4,    4,   16,   16, SLOW);
+            config_table[5]=new Config(8,   16,   32,   32, SLOW);
+            config_table[6]=new Config(8,   16,  128,  128, SLOW);
+            config_table[7]=new Config(8,   32,  128,  256, SLOW);
+            config_table[8]=new Config(32, 128,  258, 1024, SLOW);
+            config_table[9]=new Config(32, 258,  258, 4096, SLOW);
+        }
+
+        private static readonly String[] z_errmsg = {
+                                               "need dictionary",     // Z_NEED_DICT       2
+                                               "stream end",          // Z_STREAM_END      1
+                                               "",                    // Z_OK              0
+                                               "file error",          // Z_ERRNO         (-1)
+                                               "stream error",        // Z_STREAM_ERROR  (-2)
+                                               "data error",          // Z_DATA_ERROR    (-3)
+                                               "insufficient memory", // Z_MEM_ERROR     (-4)
+                                               "buffer error",        // Z_BUF_ERROR     (-5)
+                                               "incompatible version",// Z_VERSION_ERROR (-6)
+                                               ""
+                                           };
+
+        // block not completed, need more input or more output
+        private const int NeedMore=0; 
+
+        // block flush performed
+        private const int BlockDone=1; 
+
+        // finish started, need only more output at next deflate
+        private const int FinishStarted=2;
+
+        // finish done, accept no more input or output
+        private const int FinishDone=3;
+
+        // preset dictionary flag in zlib header
+        private const int PRESET_DICT=0x20;
+
+        private const int Z_FILTERED=1;
+        private const int Z_HUFFMAN_ONLY=2;
+        private const int Z_DEFAULT_STRATEGY=0;
+
+        private const int Z_NO_FLUSH=0;
+        private const int Z_PARTIAL_FLUSH=1;
+        private const int Z_SYNC_FLUSH=2;
+        private const int Z_FULL_FLUSH=3;
+        private const int Z_FINISH=4;
+
+        private const int Z_OK=0;
+        private const int Z_STREAM_END=1;
+        private const int Z_NEED_DICT=2;
+        private const int Z_ERRNO=-1;
+        private const int Z_STREAM_ERROR=-2;
+        private const int Z_DATA_ERROR=-3;
+        private const int Z_MEM_ERROR=-4;
+        private const int Z_BUF_ERROR=-5;
+        private const int Z_VERSION_ERROR=-6;
+
+        private const int INIT_STATE=42;
+        private const int BUSY_STATE=113;
+        private const int FINISH_STATE=666;
+
+        // The deflate compression method
+        private const int Z_DEFLATED=8;
+
+        private const int STORED_BLOCK=0;
+        private const int STATIC_TREES=1;
+        private const int DYN_TREES=2;
+
+        // The three kinds of block type
+        private const int Z_BINARY=0;
+        private const int Z_ASCII=1;
+        private const int Z_UNKNOWN=2;
+
+        private const int Buf_size=8*2;
+
+        // repeat previous bit length 3-6 times (2 bits of repeat count)
+        private const int REP_3_6=16; 
+
+        // repeat a zero length 3-10 times  (3 bits of repeat count)
+        private const int REPZ_3_10=17; 
+
+        // repeat a zero length 11-138 times  (7 bits of repeat count)
+        private const int REPZ_11_138=18; 
+
+        private const int MIN_MATCH=3;
+        private const int MAX_MATCH=258;
+        private const int MIN_LOOKAHEAD=(MAX_MATCH+MIN_MATCH+1);
+
+        private const int MAX_BITS=15;
+        private const int D_CODES=30;
+        private const int BL_CODES=19;
+        private const int LENGTH_CODES=29;
+        private const int LITERALS=256;
+        private const int L_CODES=(LITERALS+1+LENGTH_CODES);
+        private const int HEAP_SIZE=(2*L_CODES+1);
+
+        private const int END_BLOCK=256;
+
+        internal ZStream strm;         // pointer back to this zlib stream
+        internal int status;           // as the name implies
+        internal byte[] pending_buf;   // output still pending
+        internal int pending_buf_size; // size of pending_buf
+        internal int pending_out;      // next pending byte to output to the stream
+        internal int pending;          // nb of bytes in the pending buffer
+        internal int noheader;         // suppress zlib header and adler32
+        internal byte data_type;       // UNKNOWN, BINARY or ASCII
+        internal byte method;          // STORED (for zip only) or DEFLATED
+        internal int last_flush;       // value of flush param for previous deflate call
+
+        internal int w_size;           // LZ77 window size (32K by default)
+        internal int w_bits;           // log2(w_size)  (8..16)
+        internal int w_mask;           // w_size - 1
+
+        internal byte[] window;
+        // Sliding window. Input bytes are read into the second half of the window,
+        // and move to the first half later to keep a dictionary of at least wSize
+        // bytes. With this organization, matches are limited to a distance of
+        // wSize-MAX_MATCH bytes, but this ensures that IO is always
+        // performed with a length multiple of the block size. Also, it limits
+        // the window size to 64K, which is quite useful on MSDOS.
+        // To do: use the user input buffer as sliding window.
+
+        internal int window_size;
+        // Actual size of window: 2*wSize, except when the user input buffer
+        // is directly used as sliding window.
+
+        internal short[] prev;
+        // Link to older string with same hash index. To limit the size of this
+        // array to 64K, this link is maintained only for the last 32K strings.
+        // An index in this array is thus a window index modulo 32K.
+
+        internal short[] head; // Heads of the hash chains or NIL.
+
+        internal int ins_h;          // hash index of string to be inserted
+        internal int hash_size;      // number of elements in hash table
+        internal int hash_bits;      // log2(hash_size)
+        internal int hash_mask;      // hash_size-1
+
+        // Number of bits by which ins_h must be shifted at each input
+        // step. It must be such that after MIN_MATCH steps, the oldest
+        // byte no longer takes part in the hash key, that is:
+        // hash_shift * MIN_MATCH >= hash_bits
+        internal int hash_shift;
+
+        // Window position at the beginning of the current output block. Gets
+        // negative when the window is moved backwards.
+
+        internal int block_start;
+
+        internal int match_length;           // length of best match
+        internal int prev_match;             // previous match
+        internal int match_available;        // set if previous match exists
+        internal int strstart;               // start of string to insert
+        internal int match_start;            // start of matching string
+        internal int lookahead;              // number of valid bytes ahead in window
+
+        // Length of the best match at previous step. Matches not greater than this
+        // are discarded. This is used in the lazy match evaluation.
+        internal int prev_length;
+
+        // To speed up deflation, hash chains are never searched beyond this
+        // length.  A higher limit improves compression ratio but degrades the speed.
+        internal int max_chain_length;
+
+        // Attempt to find a better match only when the current match is strictly
+        // smaller than this value. This mechanism is used only for compression
+        // levels >= 4.
+        internal int max_lazy_match;
+
+        // Insert new strings in the hash table only if the match length is not
+        // greater than this length. This saves time but degrades compression.
+        // max_insert_length is used only for compression levels <= 3.
+
+        internal int level;    // compression level (1..9)
+        internal int strategy; // favor or force Huffman coding
+
+        // Use a faster search when the previous match is longer than this
+        internal int good_match;
+
+        // Stop searching when current match exceeds this
+        internal int nice_match;
+
+        internal short[] dyn_ltree;       // literal and length tree
+        internal short[] dyn_dtree;       // distance tree
+        internal short[] bl_tree;         // Huffman tree for bit lengths
+
+        internal Tree l_desc=new Tree();  // desc for literal tree
+        internal Tree d_desc=new Tree();  // desc for distance tree
+        internal Tree bl_desc=new Tree(); // desc for bit length tree
+
+        // number of codes at each bit length for an optimal tree
+        internal short[] bl_count=new short[MAX_BITS+1];
+
+        // heap used to build the Huffman trees
+        internal int[] heap=new int[2*L_CODES+1];
+
+        internal int heap_len;               // number of elements in the heap
+        internal int heap_max;               // element of largest frequency
+        // The sons of heap[n] are heap[2*n] and heap[2*n+1]. heap[0] is not used.
+        // The same heap array is used to build all trees.
+
+        // Depth of each subtree used as tie breaker for trees of equal frequency
+        internal byte[] depth=new byte[2*L_CODES+1];
+
+        internal int l_buf;               // index for literals or lengths */
+
+        // Size of match buffer for literals/lengths.  There are 4 reasons for
+        // limiting lit_bufsize to 64K:
+        //   - frequencies can be kept in 16 bit counters
+        //   - if compression is not successful for the first block, all input
+        //     data is still in the window so we can still emit a stored block even
+        //     when input comes from standard input.  (This can also be done for
+        //     all blocks if lit_bufsize is not greater than 32K.)
+        //   - if compression is not successful for a file smaller than 64K, we can
+        //     even emit a stored file instead of a stored block (saving 5 bytes).
+        //     This is applicable only for zip (not gzip or zlib).
+        //   - creating new Huffman trees less frequently may not provide fast
+        //     adaptation to changes in the input data statistics. (Take for
+        //     example a binary file with poorly compressible code followed by
+        //     a highly compressible string table.) Smaller buffer sizes give
+        //     fast adaptation but have of course the overhead of transmitting
+        //     trees more frequently.
+        //   - I can't count above 4
+        internal int lit_bufsize;
+
+        internal int last_lit;      // running index in l_buf
+
+        // Buffer for distances. To simplify the code, d_buf and l_buf have
+        // the same number of elements. To use different lengths, an extra flag
+        // array would be necessary.
+
+        internal int d_buf;         // index of pendig_buf
+
+        internal int opt_len;        // bit length of current block with optimal trees
+        internal int static_len;     // bit length of current block with static trees
+        internal int matches;        // number of string matches in current block
+        internal int last_eob_len;   // bit length of EOB code for last block
+
+        // Output buffer. bits are inserted starting at the bottom (least
+        // significant bits).
+        internal uint bi_buf;
+
+        // Number of valid bits in bi_buf.  All bits above the last valid bit
+        // are always zero.
+        internal int bi_valid;
+
+        internal Deflate(){
+            dyn_ltree=new short[HEAP_SIZE*2];
+            dyn_dtree=new short[(2*D_CODES+1)*2]; // distance tree
+            bl_tree=new short[(2*BL_CODES+1)*2];  // Huffman tree for bit lengths
+        }
+
+        internal void lm_init() {
+            window_size=2*w_size;
+
+            head[hash_size-1]=0;
+            for(int i=0; i<hash_size-1; i++){
+                head[i]=0;
+            }
+
+            // Set the default configuration parameters:
+            max_lazy_match   = Deflate.config_table[level].max_lazy;
+            good_match       = Deflate.config_table[level].good_length;
+            nice_match       = Deflate.config_table[level].nice_length;
+            max_chain_length = Deflate.config_table[level].max_chain;
+
+            strstart = 0;
+            block_start = 0;
+            lookahead = 0;
+            match_length = prev_length = MIN_MATCH-1;
+            match_available = 0;
+            ins_h = 0;
+        }
+
+        // Initialize the tree data structures for a new zlib stream.
+        internal void tr_init(){
+
+            l_desc.dyn_tree = dyn_ltree;
+            l_desc.stat_desc = StaticTree.static_l_desc;
+
+            d_desc.dyn_tree = dyn_dtree;
+            d_desc.stat_desc = StaticTree.static_d_desc;
+
+            bl_desc.dyn_tree = bl_tree;
+            bl_desc.stat_desc = StaticTree.static_bl_desc;
+
+            bi_buf = 0;
+            bi_valid = 0;
+            last_eob_len = 8; // enough lookahead for inflate
+
+            // Initialize the first block of the first file:
+            init_block();
+        }
+
+        internal void init_block(){
+            // Initialize the trees.
+            for(int i = 0; i < L_CODES; i++) dyn_ltree[i*2] = 0;
+            for(int i= 0; i < D_CODES; i++) dyn_dtree[i*2] = 0;
+            for(int i= 0; i < BL_CODES; i++) bl_tree[i*2] = 0;
+
+            dyn_ltree[END_BLOCK*2] = 1;
+            opt_len = static_len = 0;
+            last_lit = matches = 0;
+        }
+
+        // Restore the heap property by moving down the tree starting at node k,
+        // exchanging a node with the smallest of its two sons if necessary, stopping
+        // when the heap property is re-established (each father smaller than its
+        // two sons).
+        internal void pqdownheap(short[] tree,  // the tree to restore
+            int k          // node to move down
+            ){
+            int v = heap[k];
+            int j = k << 1;  // left son of k
+            while (j <= heap_len) {
+                // Set j to the smallest of the two sons:
+                if (j < heap_len &&
+                    smaller(tree, heap[j+1], heap[j], depth)){
+                    j++;
+                }
+                // Exit if v is smaller than both sons
+                if(smaller(tree, v, heap[j], depth)) break;
+
+                // Exchange v with the smallest son
+                heap[k]=heap[j];  k = j;
+                // And continue down the tree, setting j to the left son of k
+                j <<= 1;
+            }
+            heap[k] = v;
+        }
+
+        internal static bool smaller(short[] tree, int n, int m, byte[] depth){
+            short tn2=tree[n*2];
+            short tm2=tree[m*2];
+            return (tn2<tm2 ||
+                (tn2==tm2 && depth[n] <= depth[m]));
+        }
+
+        // Scan a literal or distance tree to determine the frequencies of the codes
+        // in the bit length tree.
+        internal void scan_tree (short[] tree,// the tree to be scanned
+            int max_code // and its largest code of non zero frequency
+            ){
+            int n;                     // iterates over all tree elements
+            int prevlen = -1;          // last emitted length
+            int curlen;                // length of current code
+            int nextlen = tree[0*2+1]; // length of next code
+            int count = 0;             // repeat count of the current code
+            int max_count = 7;         // max repeat count
+            int min_count = 4;         // min repeat count
+
+            if (nextlen == 0){ max_count = 138; min_count = 3; }
+            tree[(max_code+1)*2+1] = -1; // guard
+
+            for(n = 0; n <= max_code; n++) {
+                curlen = nextlen; nextlen = tree[(n+1)*2+1];
+                if(++count < max_count && curlen == nextlen) {
+                    continue;
+                }
+                else if(count < min_count) {
+                    bl_tree[curlen*2] += (short)count;
+                }
+                else if(curlen != 0) {
+                    if(curlen != prevlen) bl_tree[curlen*2]++;
+                    bl_tree[REP_3_6*2]++;
+                }
+                else if(count <= 10) {
+                    bl_tree[REPZ_3_10*2]++;
+                }
+                else{
+                    bl_tree[REPZ_11_138*2]++;
+                }
+                count = 0; prevlen = curlen;
+                if(nextlen == 0) {
+                    max_count = 138; min_count = 3;
+                }
+                else if(curlen == nextlen) {
+                    max_count = 6; min_count = 3;
+                }
+                else{
+                    max_count = 7; min_count = 4;
+                }
+            }
+        }
+
+        // Construct the Huffman tree for the bit lengths and return the index in
+        // bl_order of the last bit length code to send.
+        internal int build_bl_tree(){
+            int max_blindex;  // index of last bit length code of non zero freq
+
+            // Determine the bit length frequencies for literal and distance trees
+            scan_tree(dyn_ltree, l_desc.max_code);
+            scan_tree(dyn_dtree, d_desc.max_code);
+
+            // Build the bit length tree:
+            bl_desc.build_tree(this);
+            // opt_len now includes the length of the tree representations, except
+            // the lengths of the bit lengths codes and the 5+5+4 bits for the counts.
+
+            // Determine the number of bit length codes to send. The pkzip format
+            // requires that at least 4 bit length codes be sent. (appnote.txt says
+            // 3 but the actual value used is 4.)
+            for (max_blindex = BL_CODES-1; max_blindex >= 3; max_blindex--) {
+                if (bl_tree[Tree.bl_order[max_blindex]*2+1] != 0) break;
+            }
+            // Update opt_len to include the bit length tree and counts
+            opt_len += 3*(max_blindex+1) + 5+5+4;
+
+            return max_blindex;
+        }
+
+
+        // Send the header for a block using dynamic Huffman trees: the counts, the
+        // lengths of the bit length codes, the literal tree and the distance tree.
+        // IN assertion: lcodes >= 257, dcodes >= 1, blcodes >= 4.
+        internal void send_all_trees(int lcodes, int dcodes, int blcodes){
+            int rank;                    // index in bl_order
+
+            send_bits(lcodes-257, 5); // not +255 as stated in appnote.txt
+            send_bits(dcodes-1,   5);
+            send_bits(blcodes-4,  4); // not -3 as stated in appnote.txt
+            for (rank = 0; rank < blcodes; rank++) {
+                send_bits(bl_tree[Tree.bl_order[rank]*2+1], 3);
+            }
+            send_tree(dyn_ltree, lcodes-1); // literal tree
+            send_tree(dyn_dtree, dcodes-1); // distance tree
+        }
+
+        // Send a literal or distance tree in compressed form, using the codes in
+        // bl_tree.
+        internal void send_tree (short[] tree,// the tree to be sent
+            int max_code // and its largest code of non zero frequency
+            ){
+            int n;                     // iterates over all tree elements
+            int prevlen = -1;          // last emitted length
+            int curlen;                // length of current code
+            int nextlen = tree[0*2+1]; // length of next code
+            int count = 0;             // repeat count of the current code
+            int max_count = 7;         // max repeat count
+            int min_count = 4;         // min repeat count
+
+            if (nextlen == 0){ max_count = 138; min_count = 3; }
+
+            for (n = 0; n <= max_code; n++) {
+                curlen = nextlen; nextlen = tree[(n+1)*2+1];
+                if(++count < max_count && curlen == nextlen) {
+                    continue;
+                }
+                else if(count < min_count) {
+                    do { send_code(curlen, bl_tree); } while (--count != 0);
+                }
+                else if(curlen != 0){
+                    if(curlen != prevlen){
+                        send_code(curlen, bl_tree); count--;
+                    }
+                    send_code(REP_3_6, bl_tree); 
+                    send_bits(count-3, 2);
+                }
+                else if(count <= 10){
+                    send_code(REPZ_3_10, bl_tree); 
+                    send_bits(count-3, 3);
+                }
+                else{
+                    send_code(REPZ_11_138, bl_tree);
+                    send_bits(count-11, 7);
+                }
+                count = 0; prevlen = curlen;
+                if(nextlen == 0){
+                    max_count = 138; min_count = 3;
+                }
+                else if(curlen == nextlen){
+                    max_count = 6; min_count = 3;
+                }
+                else{
+                    max_count = 7; min_count = 4;
+                }
+            }
+        }
+
+        // Output a byte on the stream.
+        // IN assertion: there is enough room in pending_buf.
+        internal void put_byte(byte[] p, int start, int len){
+            System.Array.Copy(p, start, pending_buf, pending, len);
+            pending+=len;
+        }
+
+        internal void put_byte(byte c){
+            pending_buf[pending++]=c;
+        }
+        internal void put_short(int w) {
+            pending_buf[pending++]=(byte)(w/*&0xff*/);
+            pending_buf[pending++]=(byte)(w>>8);
+        }
+        internal void putShortMSB(int b){
+            pending_buf[pending++]=(byte)(b>>8);
+            pending_buf[pending++]=(byte)(b/*&0xff*/);
+        }   
+
+        internal void send_code(int c, short[] tree){
+            int c2=c*2;
+            send_bits((tree[c2]&0xffff), (tree[c2+1]&0xffff));
+        }
+
+        internal void send_bits(int val, int length){
+            if (bi_valid > Buf_size - length) {
+                bi_buf |= (uint)(val << bi_valid);
+                pending_buf[pending++]=(byte)(bi_buf/*&0xff*/);
+                pending_buf[pending++]=(byte)(bi_buf>>8);
+                bi_buf = ((uint)val) >> (Buf_size - bi_valid);
+                bi_valid += length - Buf_size;
+            } else {
+                bi_buf |= (uint)(val << bi_valid);
+                bi_valid += length;
+            }
+//            int len = length;
+//            if (bi_valid > (int)Buf_size - len) {
+//                int val = value;
+//                //      bi_buf |= (val << bi_valid);
+//                bi_buf = (short)((ushort)bi_buf | (ushort)((val << bi_valid)&0xffff));
+//                put_short(bi_buf);
+//                bi_buf = (short)(((uint)val) >> (Buf_size - bi_valid));
+//                bi_valid += len - Buf_size;
+//            } else {
+//                //      bi_buf |= (value) << bi_valid;
+//                bi_buf = (short)((ushort)bi_buf | (ushort)(((value) << bi_valid)&0xffff));
+//                bi_valid += len;
+//            }
+        }
+
+        // Send one empty static block to give enough lookahead for inflate.
+        // This takes 10 bits, of which 7 may remain in the bit buffer.
+        // The current inflate code requires 9 bits of lookahead. If the
+        // last two codes for the previous block (real code plus EOB) were coded
+        // on 5 bits or less, inflate may have only 5+3 bits of lookahead to decode
+        // the last real code. In this case we send two empty static blocks instead
+        // of one. (There are no problems if the previous block is stored or fixed.)
+        // To simplify the code, we assume the worst case of last real code encoded
+        // on one bit only.
+        internal void _tr_align(){
+            send_bits(STATIC_TREES<<1, 3);
+            send_code(END_BLOCK, StaticTree.static_ltree);
+
+            bi_flush();
+
+            // Of the 10 bits for the empty block, we have already sent
+            // (10 - bi_valid) bits. The lookahead for the last real code (before
+            // the EOB of the previous block) was thus at least one plus the length
+            // of the EOB plus what we have just sent of the empty static block.
+            if (1 + last_eob_len + 10 - bi_valid < 9) {
+                send_bits(STATIC_TREES<<1, 3);
+                send_code(END_BLOCK, StaticTree.static_ltree);
+                bi_flush();
+            }
+            last_eob_len = 7;
+        }
+
+
+        // Save the match info and tally the frequency counts. Return true if
+        // the current block must be flushed.
+        internal bool _tr_tally (int dist, // distance of matched string
+            int lc // match length-MIN_MATCH or unmatched char (if dist==0)
+            ){
+
+            pending_buf[d_buf+last_lit*2] = (byte)(dist>>8);
+            pending_buf[d_buf+last_lit*2+1] = (byte)dist;
+
+            pending_buf[l_buf+last_lit] = (byte)lc; last_lit++;
+
+            if (dist == 0) {
+                // lc is the unmatched char
+                dyn_ltree[lc*2]++;
+            } 
+            else {
+                matches++;
+                // Here, lc is the match length - MIN_MATCH
+                dist--;             // dist = match distance - 1
+                dyn_ltree[(Tree._length_code[lc]+LITERALS+1)*2]++;
+                dyn_dtree[Tree.d_code(dist)*2]++;
+            }
+
+            if ((last_lit & 0x1fff) == 0 && level > 2) {
+                // Compute an upper bound for the compressed length
+                int out_length = last_lit*8;
+                int in_length = strstart - block_start;
+                int dcode;
+                for (dcode = 0; dcode < D_CODES; dcode++) {
+                    out_length += (int)((int)dyn_dtree[dcode*2] *
+                        (5L+Tree.extra_dbits[dcode]));
+                }
+                out_length >>= 3;
+                if ((matches < (last_lit/2)) && out_length < in_length/2) return true;
+            }
+
+            return (last_lit == lit_bufsize-1);
+            // We avoid equality with lit_bufsize because of wraparound at 64K
+            // on 16 bit machines and because stored blocks are restricted to
+            // 64K-1 bytes.
+        }
+
+        // Send the block data compressed using the given Huffman trees
+        internal void compress_block(short[] ltree, short[] dtree){
+            int  dist;      // distance of matched string
+            int lc;         // match length or unmatched char (if dist == 0)
+            int lx = 0;     // running index in l_buf
+            int code;       // the code to send
+            int extra;      // number of extra bits to send
+
+            if (last_lit != 0){
+                do{
+                    dist=((pending_buf[d_buf+lx*2]<<8)&0xff00)|
+                        (pending_buf[d_buf+lx*2+1]&0xff);
+                    lc=(pending_buf[l_buf+lx])&0xff; lx++;
+
+                    if(dist == 0){
+                        send_code(lc, ltree); // send a literal byte
+                    } 
+                    else{
+                        // Here, lc is the match length - MIN_MATCH
+                        code = Tree._length_code[lc];
+
+                        send_code(code+LITERALS+1, ltree); // send the length code
+                        extra = Tree.extra_lbits[code];
+                        if(extra != 0){
+                            lc -= Tree.base_length[code];
+                            send_bits(lc, extra);       // send the extra length bits
+                        }
+                        dist--; // dist is now the match distance - 1
+                        code = Tree.d_code(dist);
+
+                        send_code(code, dtree);       // send the distance code
+                        extra = Tree.extra_dbits[code];
+                        if (extra != 0) {
+                            dist -= Tree.base_dist[code];
+                            send_bits(dist, extra);   // send the extra distance bits
+                        }
+                    } // literal or match pair ?
+
+                    // Check that the overlay between pending_buf and d_buf+l_buf is ok:
+                }
+                while (lx < last_lit);
+            }
+
+            send_code(END_BLOCK, ltree);
+            last_eob_len = ltree[END_BLOCK*2+1];
+        }
+
+        // Set the data type to ASCII or BINARY, using a crude approximation:
+        // binary if more than 20% of the bytes are <= 6 or >= 128, ascii otherwise.
+        // IN assertion: the fields freq of dyn_ltree are set and the total of all
+        // frequencies does not exceed 64K (to fit in an int on 16 bit machines).
+        internal void set_data_type(){
+            int n = 0;
+            int  ascii_freq = 0;
+            int  bin_freq = 0;
+            while(n<7){ bin_freq += dyn_ltree[n*2]; n++;}
+            while(n<128){ ascii_freq += dyn_ltree[n*2]; n++;}
+            while(n<LITERALS){ bin_freq += dyn_ltree[n*2]; n++;}
+            data_type=(byte)(bin_freq > (ascii_freq >> 2) ? Z_BINARY : Z_ASCII);
+        }
+
+        // Flush the bit buffer, keeping at most 7 bits in it.
+        internal void bi_flush(){
+            if (bi_valid == 16) {
+                pending_buf[pending++]=(byte)(bi_buf/*&0xff*/);
+                pending_buf[pending++]=(byte)(bi_buf>>8);
+                bi_buf=0;
+                bi_valid=0;
+            }
+            else if (bi_valid >= 8) {
+                pending_buf[pending++]=(byte)(bi_buf);
+                bi_buf>>=8;
+                bi_buf &= 0x00ff;
+                bi_valid-=8;
+            }
+        }
+
+        // Flush the bit buffer and align the output on a byte boundary
+        internal void bi_windup(){
+            if (bi_valid > 8) {
+                pending_buf[pending++]=(byte)(bi_buf);
+                pending_buf[pending++]=(byte)(bi_buf>>8);
+            } else if (bi_valid > 0) {
+                pending_buf[pending++]=(byte)(bi_buf);
+            }
+            bi_buf = 0;
+            bi_valid = 0;
+        }
+
+        // Copy a stored block, storing first the length and its
+        // one's complement if requested.
+        internal void copy_block(int buf,         // the input data
+            int len,         // its length
+            bool header   // true if block header must be written
+            ){
+            //int index=0;
+            bi_windup();      // align on byte boundary
+            last_eob_len = 8; // enough lookahead for inflate
+
+            if (header) {
+                put_short((short)len);   
+                put_short((short)~len);
+            }
+
+            //  while(len--!=0) {
+            //    put_byte(window[buf+index]);
+            //    index++;
+            //  }
+            put_byte(window, buf, len);
+        }
+
+        internal void flush_block_only(bool eof){
+            _tr_flush_block(block_start>=0 ? block_start : -1,
+                strstart-block_start,
+                eof);
+            block_start=strstart;
+            strm.flush_pending();
+        }
+
+        // Copy without compression as much as possible from the input stream, return
+        // the current block state.
+        // This function does not insert new strings in the dictionary since
+        // uncompressible data is probably not useful. This function is used
+        // only for the level=0 compression option.
+        // NOTE: this function should be optimized to avoid extra copying from
+        // window to pending_buf.
+        internal int deflate_stored(int flush){
+            // Stored blocks are limited to 0xffff bytes, pending_buf is limited
+            // to pending_buf_size, and each stored block has a 5 byte header:
+
+            int max_block_size = 0xffff;
+            int max_start;
+
+            if(max_block_size > pending_buf_size - 5) {
+                max_block_size = pending_buf_size - 5;
+            }
+
+            // Copy as much as possible from input to output:
+            while(true){
+                // Fill the window as much as possible:
+                if(lookahead<=1){
+                    fill_window();
+                    if(lookahead==0 && flush==Z_NO_FLUSH) return NeedMore;
+                    if(lookahead==0) break; // flush the current block
+                }
+
+                strstart+=lookahead;
+                lookahead=0;
+
+                // Emit a stored block if pending_buf will be full:
+                max_start=block_start+max_block_size;
+                if(strstart==0|| strstart>=max_start) {
+                    // strstart == 0 is possible when wraparound on 16-bit machine
+                    lookahead = (int)(strstart-max_start);
+                    strstart = (int)max_start;
+      
+                    flush_block_only(false);
+                    if(strm.avail_out==0) return NeedMore;
+
+                }
+
+                // Flush if we may have to slide, otherwise block_start may become
+                // negative and the data will be gone:
+                if(strstart-block_start >= w_size-MIN_LOOKAHEAD) {
+                    flush_block_only(false);
+                    if(strm.avail_out==0) return NeedMore;
+                }
+            }
+
+            flush_block_only(flush == Z_FINISH);
+            if(strm.avail_out==0)
+                return (flush == Z_FINISH) ? FinishStarted : NeedMore;
+
+            return flush == Z_FINISH ? FinishDone : BlockDone;
+        }
+
+        // Send a stored block
+        internal void _tr_stored_block(int buf,        // input block
+            int stored_len, // length of input block
+            bool eof     // true if this is the last block for a file
+            ){
+            send_bits((STORED_BLOCK<<1)+(eof?1:0), 3);  // send block type
+            copy_block(buf, stored_len, true);          // with header
+        }
+
+        // Determine the best encoding for the current block: dynamic trees, static
+        // trees or store, and output the encoded block to the zip file.
+        internal void _tr_flush_block(int buf,        // input block, or NULL if too old
+            int stored_len, // length of input block
+            bool eof     // true if this is the last block for a file
+            ) {
+            int opt_lenb, static_lenb;// opt_len and static_len in bytes
+            int max_blindex = 0;      // index of last bit length code of non zero freq
+
+            // Build the Huffman trees unless a stored block is forced
+            if(level > 0) {
+                // Check if the file is ascii or binary
+                if(data_type == Z_UNKNOWN) set_data_type();
+
+                // Construct the literal and distance trees
+                l_desc.build_tree(this);
+
+                d_desc.build_tree(this);
+
+                // At this point, opt_len and static_len are the total bit lengths of
+                // the compressed block data, excluding the tree representations.
+
+                // Build the bit length tree for the above two trees, and get the index
+                // in bl_order of the last bit length code to send.
+                max_blindex=build_bl_tree();
+
+                // Determine the best encoding. Compute first the block length in bytes
+                opt_lenb=(opt_len+3+7)>>3;
+                static_lenb=(static_len+3+7)>>3;
+
+                if(static_lenb<=opt_lenb) opt_lenb=static_lenb;
+            }
+            else {
+                opt_lenb=static_lenb=stored_len+5; // force a stored block
+            }
+
+            if(stored_len+4<=opt_lenb && buf != -1){
+                // 4: two words for the lengths
+                // The test buf != NULL is only necessary if LIT_BUFSIZE > WSIZE.
+                // Otherwise we can't have processed more than WSIZE input bytes since
+                // the last block flush, because compression would have been
+                // successful. If LIT_BUFSIZE <= WSIZE, it is never too late to
+                // transform a block into a stored block.
+                _tr_stored_block(buf, stored_len, eof);
+            }
+            else if(static_lenb == opt_lenb){
+                send_bits((STATIC_TREES<<1)+(eof?1:0), 3);
+                compress_block(StaticTree.static_ltree, StaticTree.static_dtree);
+            }
+            else{
+                send_bits((DYN_TREES<<1)+(eof?1:0), 3);
+                send_all_trees(l_desc.max_code+1, d_desc.max_code+1, max_blindex+1);
+                compress_block(dyn_ltree, dyn_dtree);
+            }
+
+            // The above check is made mod 2^32, for files larger than 512 MB
+            // and uLong implemented on 32 bits.
+
+            init_block();
+
+            if(eof){
+                bi_windup();
+            }
+        }
+
+        // Fill the window when the lookahead becomes insufficient.
+        // Updates strstart and lookahead.
+        //
+        // IN assertion: lookahead < MIN_LOOKAHEAD
+        // OUT assertions: strstart <= window_size-MIN_LOOKAHEAD
+        //    At least one byte has been read, or avail_in == 0; reads are
+        //    performed for at least two bytes (required for the zip translate_eol
+        //    option -- not supported here).
+        internal void fill_window(){
+            int n, m;
+            int p;
+            int more;    // Amount of free space at the end of the window.
+
+            do{
+                more = (window_size-lookahead-strstart);
+
+                // Deal with !@#$% 64K limit:
+                if(more==0 && strstart==0 && lookahead==0){
+                    more = w_size;
+                } 
+                else if(more==-1) {
+                    // Very unlikely, but possible on 16 bit machine if strstart == 0
+                    // and lookahead == 1 (input done one byte at time)
+                    more--;
+
+                    // If the window is almost full and there is insufficient lookahead,
+                    // move the upper half to the lower one to make room in the upper half.
+                }
+                else if(strstart >= w_size+ w_size-MIN_LOOKAHEAD) {
+                    System.Array.Copy(window, w_size, window, 0, w_size);
+                    match_start-=w_size;
+                    strstart-=w_size; // we now have strstart >= MAX_DIST
+                    block_start-=w_size;
+
+                    // Slide the hash table (could be avoided with 32 bit values
+                    // at the expense of memory usage). We slide even when level == 0
+                    // to keep the hash table consistent if we switch back to level > 0
+                    // later. (Using level 0 permanently is not an optimal usage of
+                    // zlib, so we don't care about this pathological case.)
+
+                    n = hash_size;
+                    p=n;
+                    do {
+                        m = (head[--p]&0xffff);
+                        head[p]=(short)(m>=w_size ? (m-w_size) : 0);
+                    }
+                    while (--n != 0);
+
+                    n = w_size;
+                    p = n;
+                    do {
+                        m = (prev[--p]&0xffff);
+                        prev[p] = (short)(m >= w_size ? (m-w_size) : 0);
+                        // If n is not on any hash chain, prev[n] is garbage but
+                        // its value will never be used.
+                    }
+                    while (--n!=0);
+                    more += w_size;
+                }
+
+                if (strm.avail_in == 0) return;
+
+                // If there was no sliding:
+                //    strstart <= WSIZE+MAX_DIST-1 && lookahead <= MIN_LOOKAHEAD - 1 &&
+                //    more == window_size - lookahead - strstart
+                // => more >= window_size - (MIN_LOOKAHEAD-1 + WSIZE + MAX_DIST-1)
+                // => more >= window_size - 2*WSIZE + 2
+                // In the BIG_MEM or MMAP case (not yet supported),
+                //   window_size == input_size + MIN_LOOKAHEAD  &&
+                //   strstart + s->lookahead <= input_size => more >= MIN_LOOKAHEAD.
+                // Otherwise, window_size == 2*WSIZE so more >= 2.
+                // If there was sliding, more >= WSIZE. So in all cases, more >= 2.
+
+                n = strm.read_buf(window, strstart + lookahead, more);
+                lookahead += n;
+
+                // Initialize the hash value now that we have some input:
+                if(lookahead >= MIN_MATCH) {
+                    ins_h = window[strstart]&0xff;
+                    ins_h=(((ins_h)<<hash_shift)^(window[strstart+1]&0xff))&hash_mask;
+                }
+                // If the whole input has less than MIN_MATCH bytes, ins_h is garbage,
+                // but this is not important since only literal bytes will be emitted.
+            }
+            while (lookahead < MIN_LOOKAHEAD && strm.avail_in != 0);
+        }
+
+        // Compress as much as possible from the input stream, return the current
+        // block state.
+        // This function does not perform lazy evaluation of matches and inserts
+        // new strings in the dictionary only for unmatched strings or for short
+        // matches. It is used only for the fast compression options.
+        internal int deflate_fast(int flush){
+            //    short hash_head = 0; // head of the hash chain
+            int hash_head = 0; // head of the hash chain
+            bool bflush;      // set if current block must be flushed
+
+            while(true){
+                // Make sure that we always have enough lookahead, except
+                // at the end of the input file. We need MAX_MATCH bytes
+                // for the next match, plus MIN_MATCH bytes to insert the
+                // string following the next match.
+                if(lookahead < MIN_LOOKAHEAD){
+                    fill_window();
+                    if(lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH){
+                        return NeedMore;
+                    }
+                    if(lookahead == 0) break; // flush the current block
+                }
+
+                // Insert the string window[strstart .. strstart+2] in the
+                // dictionary, and set hash_head to the head of the hash chain:
+                if(lookahead >= MIN_MATCH){
+                    ins_h=(((ins_h)<<hash_shift)^(window[(strstart)+(MIN_MATCH-1)]&0xff))&hash_mask;
+
+                    //  prev[strstart&w_mask]=hash_head=head[ins_h];
+                    hash_head=(head[ins_h]&0xffff);
+                    prev[strstart&w_mask]=head[ins_h];
+                    head[ins_h]=(short)strstart;
+                }
+
+                // Find the longest match, discarding those <= prev_length.
+                // At this point we have always match_length < MIN_MATCH
+
+                if(hash_head!=0L && 
+                    ((strstart-hash_head)&0xffff) <= w_size-MIN_LOOKAHEAD
+                    ){
+                    // To simplify the code, we prevent matches with the string
+                    // of window index 0 (in particular we have to avoid a match
+                    // of the string with itself at the start of the input file).
+                    if(strategy != Z_HUFFMAN_ONLY){
+                        match_length=longest_match (hash_head);
+                    }
+                    // longest_match() sets match_start
+                }
+                if(match_length>=MIN_MATCH){
+                    //        check_match(strstart, match_start, match_length);
+
+                    bflush=_tr_tally(strstart-match_start, match_length-MIN_MATCH);
+
+                    lookahead -= match_length;
+
+                    // Insert new strings in the hash table only if the match length
+                    // is not too large. This saves time but degrades compression.
+                    if(match_length <= max_lazy_match &&
+                        lookahead >= MIN_MATCH) {
+                        match_length--; // string at strstart already in hash table
+                        do{
+                            strstart++;
+
+                            ins_h=((ins_h<<hash_shift)^(window[(strstart)+(MIN_MATCH-1)]&0xff))&hash_mask;
+                            //      prev[strstart&w_mask]=hash_head=head[ins_h];
+                            hash_head=(head[ins_h]&0xffff);
+                            prev[strstart&w_mask]=head[ins_h];
+                            head[ins_h]=(short)strstart;
+
+                            // strstart never exceeds WSIZE-MAX_MATCH, so there are
+                            // always MIN_MATCH bytes ahead.
+                        }
+                        while (--match_length != 0);
+                        strstart++; 
+                    }
+                    else{
+                        strstart += match_length;
+                        match_length = 0;
+                        ins_h = window[strstart]&0xff;
+
+                        ins_h=(((ins_h)<<hash_shift)^(window[strstart+1]&0xff))&hash_mask;
+                        // If lookahead < MIN_MATCH, ins_h is garbage, but it does not
+                        // matter since it will be recomputed at next deflate call.
+                    }
+                }
+                else {
+                    // No match, output a literal byte
+
+                    bflush=_tr_tally(0, window[strstart]&0xff);
+                    lookahead--;
+                    strstart++; 
+                }
+                if (bflush){
+
+                    flush_block_only(false);
+                    if(strm.avail_out==0) return NeedMore;
+                }
+            }
+
+            flush_block_only(flush == Z_FINISH);
+            if(strm.avail_out==0){
+                if(flush == Z_FINISH) return FinishStarted;
+                else return NeedMore;
+            }
+            return flush==Z_FINISH ? FinishDone : BlockDone;
+        }
+
+        // Same as above, but achieves better compression. We use a lazy
+        // evaluation for matches: a match is finally adopted only if there is
+        // no better match at the next window position.
+        internal int deflate_slow(int flush){
+            //    short hash_head = 0;    // head of hash chain
+            int hash_head = 0;    // head of hash chain
+            bool bflush;         // set if current block must be flushed
+
+            // Process the input block.
+            while(true){
+                // Make sure that we always have enough lookahead, except
+                // at the end of the input file. We need MAX_MATCH bytes
+                // for the next match, plus MIN_MATCH bytes to insert the
+                // string following the next match.
+
+                if (lookahead < MIN_LOOKAHEAD) {
+                    fill_window();
+                    if(lookahead < MIN_LOOKAHEAD && flush == Z_NO_FLUSH) {
+                        return NeedMore;
+                    }
+                    if(lookahead == 0) break; // flush the current block
+                }
+
+                // Insert the string window[strstart .. strstart+2] in the
+                // dictionary, and set hash_head to the head of the hash chain:
+
+                if(lookahead >= MIN_MATCH) {
+                    ins_h=(((ins_h)<<hash_shift)^(window[(strstart)+(MIN_MATCH-1)]&0xff)) & hash_mask;
+                    //  prev[strstart&w_mask]=hash_head=head[ins_h];
+                    hash_head=(head[ins_h]&0xffff);
+                    prev[strstart&w_mask]=head[ins_h];
+                    head[ins_h]=(short)strstart;
+                }
+
+                // Find the longest match, discarding those <= prev_length.
+                prev_length = match_length; prev_match = match_start;
+                match_length = MIN_MATCH-1;
+
+                if (hash_head != 0 && prev_length < max_lazy_match &&
+                    ((strstart-hash_head)&0xffff) <= w_size-MIN_LOOKAHEAD
+                    ){
+                    // To simplify the code, we prevent matches with the string
+                    // of window index 0 (in particular we have to avoid a match
+                    // of the string with itself at the start of the input file).
+
+                    if(strategy != Z_HUFFMAN_ONLY) {
+                        match_length = longest_match(hash_head);
+                    }
+                    // longest_match() sets match_start
+
+                    if (match_length <= 5 && (strategy == Z_FILTERED ||
+                        (match_length == MIN_MATCH &&
+                        strstart - match_start > 4096))) {
+
+                        // If prev_match is also MIN_MATCH, match_start is garbage
+                        // but we will ignore the current match anyway.
+                        match_length = MIN_MATCH-1;
+                    }
+                }
+
+                // If there was a match at the previous step and the current
+                // match is not better, output the previous match:
+                if(prev_length >= MIN_MATCH && match_length <= prev_length) {
+                    int max_insert = strstart + lookahead - MIN_MATCH;
+                    // Do not insert strings in hash table beyond this.
+
+                    //          check_match(strstart-1, prev_match, prev_length);
+
+                    bflush=_tr_tally(strstart-1-prev_match, prev_length - MIN_MATCH);
+
+                    // Insert in hash table all strings up to the end of the match.
+                    // strstart-1 and strstart are already inserted. If there is not
+                    // enough lookahead, the last two strings are not inserted in
+                    // the hash table.
+                    lookahead -= prev_length-1;
+                    prev_length -= 2;
+                    do{
+                        if(++strstart <= max_insert) {
+                            ins_h=(((ins_h)<<hash_shift)^(window[(strstart)+(MIN_MATCH-1)]&0xff))&hash_mask;
+                            //prev[strstart&w_mask]=hash_head=head[ins_h];
+                            hash_head=(head[ins_h]&0xffff);
+                            prev[strstart&w_mask]=head[ins_h];
+                            head[ins_h]=(short)strstart;
+                        }
+                    }
+                    while(--prev_length != 0);
+                    match_available = 0;
+                    match_length = MIN_MATCH-1;
+                    strstart++;
+
+                    if (bflush){
+                        flush_block_only(false);
+                        if(strm.avail_out==0) return NeedMore;
+                    }
+                } else if (match_available!=0) {
+
+                    // If there was no match at the previous position, output a
+                    // single literal. If there was a match but the current match
+                    // is longer, truncate the previous match to a single literal.
+
+                    bflush=_tr_tally(0, window[strstart-1]&0xff);
+
+                    if (bflush) {
+                        flush_block_only(false);
+                    }
+                    strstart++;
+                    lookahead--;
+                    if(strm.avail_out == 0) return NeedMore;
+                } else {
+                    // There is no previous match to compare with, wait for
+                    // the next step to decide.
+
+                    match_available = 1;
+                    strstart++;
+                    lookahead--;
+                }
+            }
+
+            if(match_available!=0) {
+                bflush=_tr_tally(0, window[strstart-1]&0xff);
+                match_available = 0;
+            }
+            flush_block_only(flush == Z_FINISH);
+
+            if(strm.avail_out==0){
+                if(flush == Z_FINISH) return FinishStarted;
+                else return NeedMore;
+            }
+
+            return flush == Z_FINISH ? FinishDone : BlockDone;
+        }
+
+        internal int longest_match(int cur_match){
+            int chain_length = max_chain_length; // max hash chain length
+            int scan = strstart;                 // current string
+            int match;                           // matched string
+            int len;                             // length of current match
+            int best_len = prev_length;          // best match length so far
+            int limit = strstart>(w_size-MIN_LOOKAHEAD) ?
+                strstart-(w_size-MIN_LOOKAHEAD) : 0;
+            int nice_match=this.nice_match;
+
+            // Stop when cur_match becomes <= limit. To simplify the code,
+            // we prevent matches with the string of window index 0.
+
+            int wmask = w_mask;
+
+            int strend = strstart + MAX_MATCH;
+            byte scan_end1 = window[scan+best_len-1];
+            byte scan_end = window[scan+best_len];
+
+            // The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16.
+            // It is easy to get rid of this optimization if necessary.
+
+            // Do not waste too much time if we already have a good match:
+            if (prev_length >= good_match) {
+                chain_length >>= 2;
+            }
+
+            // Do not look for matches beyond the end of the input. This is necessary
+            // to make deflate deterministic.
+            if (nice_match > lookahead) nice_match = lookahead;
+
+            do {
+                match = cur_match;
+
+                // Skip to next match if the match length cannot increase
+                // or if the match length is less than 2:
+                if (window[match+best_len]   != scan_end  ||
+                    window[match+best_len-1] != scan_end1 ||
+                    window[match]       != window[scan]     ||
+                    window[++match]     != window[scan+1])      continue;
+
+                // The check at best_len-1 can be removed because it will be made
+                // again later. (This heuristic is not always a win.)
+                // It is not necessary to compare scan[2] and match[2] since they
+                // are always equal when the other bytes match, given that
+                // the hash keys are equal and that HASH_BITS >= 8.
+                scan += 2; match++;
+
+                // We check for insufficient lookahead only every 8th comparison;
+                // the 256th check will be made at strstart+258.
+                do {
+                } while (window[++scan] == window[++match] &&
+                    window[++scan] == window[++match] &&
+                    window[++scan] == window[++match] &&
+                    window[++scan] == window[++match] &&
+                    window[++scan] == window[++match] &&
+                    window[++scan] == window[++match] &&
+                    window[++scan] == window[++match] &&
+                    window[++scan] == window[++match] &&
+                    scan < strend);
+
+                len = MAX_MATCH - (int)(strend - scan);
+                scan = strend - MAX_MATCH;
+
+                if(len>best_len) {
+                    match_start = cur_match;
+                    best_len = len;
+                    if (len >= nice_match) break;
+                    scan_end1  = window[scan+best_len-1];
+                    scan_end   = window[scan+best_len];
+                }
+
+            } while ((cur_match = (prev[cur_match & wmask]&0xffff)) > limit
+                && --chain_length != 0);
+
+            if (best_len <= lookahead) return best_len;
+            return lookahead;
+        }
+    
+        internal int deflateInit(ZStream strm, int level, int bits){
+            return deflateInit2(strm, level, Z_DEFLATED, bits, DEF_MEM_LEVEL,
+                Z_DEFAULT_STRATEGY);
+        }
+        internal int deflateInit(ZStream strm, int level){
+            return deflateInit(strm, level, MAX_WBITS);
+        }
+        internal int deflateInit2(ZStream strm, int level, int method,  int windowBits,
+            int memLevel, int strategy){
+            int noheader = 0;
+            //    byte[] my_version=ZLIB_VERSION;
+
+            //
+            //  if (version == null || version[0] != my_version[0]
+            //  || stream_size != sizeof(z_stream)) {
+            //  return Z_VERSION_ERROR;
+            //  }
+
+            strm.msg = null;
+
+            if (level == Z_DEFAULT_COMPRESSION) level = 6;
+
+            if (windowBits < 0) { // undocumented feature: suppress zlib header
+                noheader = 1;
+                windowBits = -windowBits;
+            }
+
+            if (memLevel < 1 || memLevel > MAX_MEM_LEVEL || 
+                method != Z_DEFLATED ||
+                windowBits < 9 || windowBits > 15 || level < 0 || level > 9 ||
+                strategy < 0 || strategy > Z_HUFFMAN_ONLY) {
+                return Z_STREAM_ERROR;
+            }
+
+            strm.dstate = (Deflate)this;
+
+            this.noheader = noheader;
+            w_bits = windowBits;
+            w_size = 1 << w_bits;
+            w_mask = w_size - 1;
+
+            hash_bits = memLevel + 7;
+            hash_size = 1 << hash_bits;
+            hash_mask = hash_size - 1;
+            hash_shift = ((hash_bits+MIN_MATCH-1)/MIN_MATCH);
+
+            window = new byte[w_size*2];
+            prev = new short[w_size];
+            head = new short[hash_size];
+
+            lit_bufsize = 1 << (memLevel + 6); // 16K elements by default
+
+            // We overlay pending_buf and d_buf+l_buf. This works since the average
+            // output size for (length,distance) codes is <= 24 bits.
+            pending_buf = new byte[lit_bufsize*4];
+            pending_buf_size = lit_bufsize*4;
+
+            d_buf = lit_bufsize/2;
+            l_buf = (1+2)*lit_bufsize;
+
+            this.level = level;
+
+            //System.out.println("level="+level);
+
+            this.strategy = strategy;
+            this.method = (byte)method;
+
+            return deflateReset(strm);
+        }
+
+        internal int deflateReset(ZStream strm){
+            strm.total_in = strm.total_out = 0;
+            strm.msg = null; //
+            strm.data_type = Z_UNKNOWN;
+
+            pending = 0;
+            pending_out = 0;
+
+            if(noheader < 0) {
+                noheader = 0; // was set to -1 by deflate(..., Z_FINISH);
+            }
+            status = (noheader!=0) ? BUSY_STATE : INIT_STATE;
+            strm.adler=strm._adler.adler32(0, null, 0, 0);
+
+            last_flush = Z_NO_FLUSH;
+
+            tr_init();
+            lm_init();
+            return Z_OK;
+        }
+
+        internal int deflateEnd(){
+            if(status!=INIT_STATE && status!=BUSY_STATE && status!=FINISH_STATE){
+                return Z_STREAM_ERROR;
+            }
+            // Deallocate in reverse order of allocations:
+            pending_buf=null;
+            head=null;
+            prev=null;
+            window=null;
+            // free
+            // dstate=null;
+            return status == BUSY_STATE ? Z_DATA_ERROR : Z_OK;
+        }
+
+        internal int deflateParams(ZStream strm, int _level, int _strategy){
+            int err=Z_OK;
+
+            if(_level == Z_DEFAULT_COMPRESSION){
+                _level = 6;
+            }
+            if(_level < 0 || _level > 9 || 
+                _strategy < 0 || _strategy > Z_HUFFMAN_ONLY) {
+                return Z_STREAM_ERROR;
+            }
+
+            if(config_table[level].func!=config_table[_level].func &&
+                strm.total_in != 0) {
+                // Flush the last buffer:
+                err = strm.deflate(Z_PARTIAL_FLUSH);
+            }
+
+            if(level != _level) {
+                level = _level;
+                max_lazy_match   = config_table[level].max_lazy;
+                good_match       = config_table[level].good_length;
+                nice_match       = config_table[level].nice_length;
+                max_chain_length = config_table[level].max_chain;
+            }
+            strategy = _strategy;
+            return err;
+        }
+
+        internal int deflateSetDictionary (ZStream strm, byte[] dictionary, int dictLength){
+            int length = dictLength;
+            int index=0;
+
+            if(dictionary == null || status != INIT_STATE)
+                return Z_STREAM_ERROR;
+
+            strm.adler=strm._adler.adler32(strm.adler, dictionary, 0, dictLength);
+
+            if(length < MIN_MATCH) return Z_OK;
+            if(length > w_size-MIN_LOOKAHEAD){
+                length = w_size-MIN_LOOKAHEAD;
+                index=dictLength-length; // use the tail of the dictionary
+            }
+            System.Array.Copy(dictionary, index, window, 0, length);
+            strstart = length;
+            block_start = length;
+
+            // Insert all strings in the hash table (except for the last two bytes).
+            // s->lookahead stays null, so s->ins_h will be recomputed at the next
+            // call of fill_window.
+
+            ins_h = window[0]&0xff;
+            ins_h=(((ins_h)<<hash_shift)^(window[1]&0xff))&hash_mask;
+
+            for(int n=0; n<=length-MIN_MATCH; n++){
+                ins_h=(((ins_h)<<hash_shift)^(window[(n)+(MIN_MATCH-1)]&0xff))&hash_mask;
+                prev[n&w_mask]=head[ins_h];
+                head[ins_h]=(short)n;
+            }
+            return Z_OK;
+        }
+
+        internal int deflate(ZStream strm, int flush){
+            int old_flush;
+
+            if(flush>Z_FINISH || flush<0){
+                return Z_STREAM_ERROR;
+            }
+
+            if(strm.next_out == null ||
+                (strm.next_in == null && strm.avail_in != 0) ||
+                (status == FINISH_STATE && flush != Z_FINISH)) {
+                strm.msg=z_errmsg[Z_NEED_DICT-(Z_STREAM_ERROR)];
+                return Z_STREAM_ERROR;
+            }
+            if(strm.avail_out == 0){
+                strm.msg=z_errmsg[Z_NEED_DICT-(Z_BUF_ERROR)];
+                return Z_BUF_ERROR;
+            }
+
+            this.strm = strm; // just in case
+            old_flush = last_flush;
+            last_flush = flush;
+
+            // Write the zlib header
+            if(status == INIT_STATE) {
+                int header = (Z_DEFLATED+((w_bits-8)<<4))<<8;
+                int level_flags=((level-1)&0xff)>>1;
+
+                if(level_flags>3) level_flags=3;
+                header |= (level_flags<<6);
+                if(strstart!=0) header |= PRESET_DICT;
+                header+=31-(header % 31);
+
+                status=BUSY_STATE;
+                putShortMSB(header);
+
+
+                // Save the adler32 of the preset dictionary:
+                if(strstart!=0){
+                    putShortMSB((int)(strm.adler>>16));
+                    putShortMSB((int)(strm.adler&0xffff));
+                }
+                strm.adler=strm._adler.adler32(0, null, 0, 0);
+            }
+
+            // Flush as much pending output as possible
+            if(pending != 0) {
+                strm.flush_pending();
+                if(strm.avail_out == 0) {
+                    //System.out.println("  avail_out==0");
+                    // Since avail_out is 0, deflate will be called again with
+                    // more output space, but possibly with both pending and
+                    // avail_in equal to zero. There won't be anything to do,
+                    // but this is not an error situation so make sure we
+                    // return OK instead of BUF_ERROR at next call of deflate:
+                    last_flush = -1;
+                    return Z_OK;
+                }
+
+                // Make sure there is something to do and avoid duplicate consecutive
+                // flushes. For repeated and useless calls with Z_FINISH, we keep
+                // returning Z_STREAM_END instead of Z_BUFF_ERROR.
+            }
+            else if(strm.avail_in==0 && flush <= old_flush &&
+                flush != Z_FINISH) {
+                strm.msg=z_errmsg[Z_NEED_DICT-(Z_BUF_ERROR)];
+                return Z_BUF_ERROR;
+            }
+
+            // User must not provide more input after the first FINISH:
+            if(status == FINISH_STATE && strm.avail_in != 0) {
+                strm.msg=z_errmsg[Z_NEED_DICT-(Z_BUF_ERROR)];
+                return Z_BUF_ERROR;
+            }
+
+            // Start a new block or continue the current one.
+            if(strm.avail_in!=0 || lookahead!=0 ||
+                (flush != Z_NO_FLUSH && status != FINISH_STATE)) {
+                int bstate=-1;
+                switch(config_table[level].func){
+                    case STORED: 
+                        bstate = deflate_stored(flush);
+                        break;
+                    case FAST: 
+                        bstate = deflate_fast(flush);
+                        break;
+                    case SLOW: 
+                        bstate = deflate_slow(flush);
+                        break;
+                    default:
+                        break;
+                }
+
+                if (bstate==FinishStarted || bstate==FinishDone) {
+                    status = FINISH_STATE;
+                }
+                if (bstate==NeedMore || bstate==FinishStarted) {
+                    if(strm.avail_out == 0) {
+                        last_flush = -1; // avoid BUF_ERROR next call, see above
+                    }
+                    return Z_OK;
+                    // If flush != Z_NO_FLUSH && avail_out == 0, the next call
+                    // of deflate should use the same flush parameter to make sure
+                    // that the flush is complete. So we don't have to output an
+                    // empty block here, this will be done at next call. This also
+                    // ensures that for a very small output buffer, we emit at most
+                    // one empty block.
+                }
+
+                if (bstate==BlockDone) {
+                    if(flush == Z_PARTIAL_FLUSH) {
+                        _tr_align();
+                    } 
+                    else { // FULL_FLUSH or SYNC_FLUSH
+                        _tr_stored_block(0, 0, false);
+                        // For a full flush, this empty block will be recognized
+                        // as a special marker by inflate_sync().
+                        if(flush == Z_FULL_FLUSH) {
+                            //state.head[s.hash_size-1]=0;
+                            for(int i=0; i<hash_size/*-1*/; i++)  // forget history
+                                head[i]=0;
+                        }
+                    }
+                    strm.flush_pending();
+                    if(strm.avail_out == 0) {
+                        last_flush = -1; // avoid BUF_ERROR at next call, see above
+                        return Z_OK;
+                    }
+                }
+            }
+
+            if(flush!=Z_FINISH) return Z_OK;
+            if(noheader!=0) return Z_STREAM_END;
+
+            // Write the zlib trailer (adler32)
+            putShortMSB((int)(strm.adler>>16));
+            putShortMSB((int)(strm.adler&0xffff));
+            strm.flush_pending();
+
+            // If avail_out is zero, the application will call deflate again
+            // to flush the rest.
+            noheader = -1; // write the trailer only once!
+            return pending != 0 ? Z_OK : Z_STREAM_END;
+        }
+    }
+}
\ No newline at end of file
diff --git a/Crypto/src/util/zlib/InfBlocks.cs b/Crypto/src/util/zlib/InfBlocks.cs
new file mode 100644
index 000000000..479d9b5c9
--- /dev/null
+++ b/Crypto/src/util/zlib/InfBlocks.cs
@@ -0,0 +1,618 @@
+using System;
+/*
+ * $Id: InfBlocks.cs,v 1.2 2008-05-10 09:35:40 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    internal sealed class InfBlocks{
+        private const int MANY=1440;
+
+        // And'ing with mask[n] masks the lower n bits
+        private static readonly int[] inflate_mask = {
+                                                0x00000000, 0x00000001, 0x00000003, 0x00000007, 0x0000000f,
+                                                0x0000001f, 0x0000003f, 0x0000007f, 0x000000ff, 0x000001ff,
+                                                0x000003ff, 0x000007ff, 0x00000fff, 0x00001fff, 0x00003fff,
+                                                0x00007fff, 0x0000ffff
+                                            };
+
+        // Table for deflate from PKZIP's appnote.txt.
+        static readonly int[] border = { // Order of the bit length code lengths
+                                  16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15
+                              };
+
+        private const int Z_OK=0;
+        private const int Z_STREAM_END=1;
+        private const int Z_NEED_DICT=2;
+        private const int Z_ERRNO=-1;
+        private const int Z_STREAM_ERROR=-2;
+        private const int Z_DATA_ERROR=-3;
+        private const int Z_MEM_ERROR=-4;
+        private const int Z_BUF_ERROR=-5;
+        private const int Z_VERSION_ERROR=-6;
+
+        private const int TYPE=0;  // get type bits (3, including end bit)
+        private const int LENS=1;  // get lengths for stored
+        private const int STORED=2;// processing stored block
+        private const int TABLE=3; // get table lengths
+        private const int BTREE=4; // get bit lengths tree for a dynamic block
+        private const int DTREE=5; // get length, distance trees for a dynamic block
+        private const int CODES=6; // processing fixed or dynamic block
+        private const int DRY=7;   // output remaining window bytes
+        private const int DONE=8;  // finished last block, done
+        private const int BAD=9;   // ot a data error--stuck here
+
+        internal int mode;            // current inflate_block mode 
+
+        internal int left;            // if STORED, bytes left to copy 
+
+        internal int table;           // table lengths (14 bits) 
+        internal int index;           // index into blens (or border) 
+        internal int[] blens;         // bit lengths of codes 
+        internal int[] bb=new int[1]; // bit length tree depth 
+        internal int[] tb=new int[1]; // bit length decoding tree 
+
+        internal InfCodes codes=new InfCodes();      // if CODES, current state 
+
+        int last;            // true if this block is the last block 
+
+        // mode independent information 
+        internal int bitk;            // bits in bit buffer 
+        internal int bitb;            // bit buffer 
+        internal int[] hufts;         // single malloc for tree space 
+        internal byte[] window;       // sliding window 
+        internal int end;             // one byte after sliding window 
+        internal int read;            // window read pointer 
+        internal int write;           // window write pointer 
+        internal Object checkfn;      // check function 
+        internal long check;          // check on output 
+
+        internal InfTree inftree=new InfTree();
+
+        internal InfBlocks(ZStream z, Object checkfn, int w){
+            hufts=new int[MANY*3];
+            window=new byte[w];
+            end=w;
+            this.checkfn = checkfn;
+            mode = TYPE;
+            reset(z, null);
+        }
+
+        internal void reset(ZStream z, long[] c){
+            if(c!=null) c[0]=check;
+            if(mode==BTREE || mode==DTREE){
+            }
+            if(mode==CODES){
+                codes.free(z);
+            }
+            mode=TYPE;
+            bitk=0;
+            bitb=0;
+            read=write=0;
+
+            if(checkfn != null)
+                z.adler=check=z._adler.adler32(0L, null, 0, 0);
+        }
+
+        internal int proc(ZStream z, int r){
+            int t;              // temporary storage
+            int b;              // bit buffer
+            int k;              // bits in bit buffer
+            int p;              // input data pointer
+            int n;              // bytes available there
+            int q;              // output window write pointer
+            int m; {              // bytes to end of window or read pointer
+
+            // copy input/output information to locals (UPDATE macro restores)
+     p=z.next_in_index;n=z.avail_in;b=bitb;k=bitk;} {
+     q=write;m=(int)(q<read?read-q-1:end-q);}
+
+            // process input based on current state
+            while(true){
+                switch (mode){
+                    case TYPE:
+
+                        while(k<(3)){
+                            if(n!=0){
+                                r=Z_OK;
+                            }
+                            else{
+                                bitb=b; bitk=k; 
+                                z.avail_in=n;
+                                z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                write=q;
+                                return inflate_flush(z,r);
+                            };
+                            n--;
+                            b|=(z.next_in[p++]&0xff)<<k;
+                            k+=8;
+                        }
+                        t = (int)(b & 7);
+                        last = t & 1;
+
+                    switch (t >> 1){
+                        case 0: {                         // stored 
+           b>>=(3);k-=(3);}
+                            t = k & 7; {                    // go to byte boundary
+
+           b>>=(t);k-=(t);}
+                            mode = LENS;                  // get length of stored block
+                            break;
+                        case 1: {                         // fixed
+                            int[] bl=new int[1];
+                            int[] bd=new int[1];
+                            int[][] tl=new int[1][];
+                            int[][] td=new int[1][];
+
+                            InfTree.inflate_trees_fixed(bl, bd, tl, td, z);
+                            codes.init(bl[0], bd[0], tl[0], 0, td[0], 0, z);
+                        } {
+
+           b>>=(3);k-=(3);}
+
+                            mode = CODES;
+                            break;
+                        case 2: {                         // dynamic
+
+           b>>=(3);k-=(3);}
+
+                            mode = TABLE;
+                            break;
+                        case 3: {                         // illegal
+
+           b>>=(3);k-=(3);}
+                            mode = BAD;
+                            z.msg = "invalid block type";
+                            r = Z_DATA_ERROR;
+
+                            bitb=b; bitk=k; 
+                            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                            write=q;
+                            return inflate_flush(z,r);
+                    }
+                        break;
+                    case LENS:
+
+                        while(k<(32)){
+                            if(n!=0){
+                                r=Z_OK;
+                            }
+                            else{
+                                bitb=b; bitk=k; 
+                                z.avail_in=n;
+                                z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                write=q;
+                                return inflate_flush(z,r);
+                            };
+                            n--;
+                            b|=(z.next_in[p++]&0xff)<<k;
+                            k+=8;
+                        }
+
+                        if ((((~b) >> 16) & 0xffff) != (b & 0xffff)){
+                            mode = BAD;
+                            z.msg = "invalid stored block lengths";
+                            r = Z_DATA_ERROR;
+
+                            bitb=b; bitk=k; 
+                            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                            write=q;
+                            return inflate_flush(z,r);
+                        }
+                        left = (b & 0xffff);
+                        b = k = 0;                       // dump bits
+                        mode = left!=0 ? STORED : (last!=0 ? DRY : TYPE);
+                        break;
+                    case STORED:
+                        if (n == 0){
+                            bitb=b; bitk=k; 
+                            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                            write=q;
+                            return inflate_flush(z,r);
+                        }
+
+                        if(m==0){
+                            if(q==end&&read!=0){
+                                q=0; m=(int)(q<read?read-q-1:end-q);
+                            }
+                            if(m==0){
+                                write=q; 
+                                r=inflate_flush(z,r);
+                                q=write;m=(int)(q<read?read-q-1:end-q);
+                                if(q==end&&read!=0){
+                                    q=0; m=(int)(q<read?read-q-1:end-q);
+                                }
+                                if(m==0){
+                                    bitb=b; bitk=k; 
+                                    z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                    write=q;
+                                    return inflate_flush(z,r);
+                                }
+                            }
+                        }
+                        r=Z_OK;
+
+                        t = left;
+                        if(t>n) t = n;
+                        if(t>m) t = m;
+                        System.Array.Copy(z.next_in, p, window, q, t);
+                        p += t;  n -= t;
+                        q += t;  m -= t;
+                        if ((left -= t) != 0)
+                            break;
+                        mode = last!=0 ? DRY : TYPE;
+                        break;
+                    case TABLE:
+
+                        while(k<(14)){
+                            if(n!=0){
+                                r=Z_OK;
+                            }
+                            else{
+                                bitb=b; bitk=k; 
+                                z.avail_in=n;
+                                z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                write=q;
+                                return inflate_flush(z,r);
+                            };
+                            n--;
+                            b|=(z.next_in[p++]&0xff)<<k;
+                            k+=8;
+                        }
+
+                        table = t = (b & 0x3fff);
+                        if ((t & 0x1f) > 29 || ((t >> 5) & 0x1f) > 29) {
+                            mode = BAD;
+                            z.msg = "too many length or distance symbols";
+                            r = Z_DATA_ERROR;
+
+                            bitb=b; bitk=k; 
+                            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                            write=q;
+                            return inflate_flush(z,r);
+                        }
+                        t = 258 + (t & 0x1f) + ((t >> 5) & 0x1f);
+                        if(blens==null || blens.Length<t){
+                            blens=new int[t];
+                        }
+                        else{
+                            for(int i=0; i<t; i++){blens[i]=0;}
+                        } {
+
+	 b>>=(14);k-=(14);}
+
+                        index = 0;
+                        mode = BTREE;
+                        goto case BTREE;
+                    case BTREE:
+                        while (index < 4 + (table >> 10)){
+                            while(k<(3)){
+                                if(n!=0){
+                                    r=Z_OK;
+                                }
+                                else{
+                                    bitb=b; bitk=k; 
+                                    z.avail_in=n;
+                                    z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                    write=q;
+                                    return inflate_flush(z,r);
+                                };
+                                n--;
+                                b|=(z.next_in[p++]&0xff)<<k;
+                                k+=8;
+                            }
+
+                            blens[border[index++]] = b&7; {
+
+	   b>>=(3);k-=(3);}
+                        }
+
+                        while(index < 19){
+                            blens[border[index++]] = 0;
+                        }
+
+                        bb[0] = 7;
+                        t = inftree.inflate_trees_bits(blens, bb, tb, hufts, z);
+                        if (t != Z_OK){
+                            r = t;
+                            if (r == Z_DATA_ERROR){
+                                blens=null;
+                                mode = BAD;
+                            }
+
+                            bitb=b; bitk=k; 
+                            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                            write=q;
+                            return inflate_flush(z,r);
+                        }
+
+                        index = 0;
+                        mode = DTREE;
+                        goto case DTREE;
+                    case DTREE:
+                        while (true){
+                            t = table;
+                            if(!(index < 258 + (t & 0x1f) + ((t >> 5) & 0x1f))){
+                                break;
+                            }
+
+                            int i, j, c;
+
+                            t = bb[0];
+
+                            while(k<(t)){
+                                if(n!=0){
+                                    r=Z_OK;
+                                }
+                                else{
+                                    bitb=b; bitk=k; 
+                                    z.avail_in=n;
+                                    z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                    write=q;
+                                    return inflate_flush(z,r);
+                                };
+                                n--;
+                                b|=(z.next_in[p++]&0xff)<<k;
+                                k+=8;
+                            }
+
+                            if(tb[0]==-1){
+                                //System.err.println("null...");
+                            }
+
+                            t=hufts[(tb[0]+(b&inflate_mask[t]))*3+1];
+                            c=hufts[(tb[0]+(b&inflate_mask[t]))*3+2];
+
+                            if (c < 16){
+                                b>>=(t);k-=(t);
+                                blens[index++] = c;
+                            }
+                            else { // c == 16..18
+                                i = c == 18 ? 7 : c - 14;
+                                j = c == 18 ? 11 : 3;
+
+                                while(k<(t+i)){
+                                    if(n!=0){
+                                        r=Z_OK;
+                                    }
+                                    else{
+                                        bitb=b; bitk=k; 
+                                        z.avail_in=n;
+                                        z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                        write=q;
+                                        return inflate_flush(z,r);
+                                    };
+                                    n--;
+                                    b|=(z.next_in[p++]&0xff)<<k;
+                                    k+=8;
+                                }
+
+                                b>>=(t);k-=(t);
+
+                                j += (b & inflate_mask[i]);
+
+                                b>>=(i);k-=(i);
+
+                                i = index;
+                                t = table;
+                                if (i + j > 258 + (t & 0x1f) + ((t >> 5) & 0x1f) ||
+                                    (c == 16 && i < 1)){
+                                    blens=null;
+                                    mode = BAD;
+                                    z.msg = "invalid bit length repeat";
+                                    r = Z_DATA_ERROR;
+
+                                    bitb=b; bitk=k; 
+                                    z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                    write=q;
+                                    return inflate_flush(z,r);
+                                }
+
+                                c = c == 16 ? blens[i-1] : 0;
+                                do{
+                                    blens[i++] = c;
+                                }
+                                while (--j!=0);
+                                index = i;
+                            }
+                        }
+
+                        tb[0]=-1; {
+                        int[] bl=new int[1];
+                        int[] bd=new int[1];
+                        int[] tl=new int[1];
+                        int[] td=new int[1];
+                        bl[0] = 9;         // must be <= 9 for lookahead assumptions
+                        bd[0] = 6;         // must be <= 9 for lookahead assumptions
+
+                        t = table;
+                        t = inftree.inflate_trees_dynamic(257 + (t & 0x1f), 
+                            1 + ((t >> 5) & 0x1f),
+                            blens, bl, bd, tl, td, hufts, z);
+
+                        if (t != Z_OK){
+                            if (t == Z_DATA_ERROR){
+                                blens=null;
+                                mode = BAD;
+                            }
+                            r = t;
+
+                            bitb=b; bitk=k; 
+                            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                            write=q;
+                            return inflate_flush(z,r);
+                        }
+                        codes.init(bl[0], bd[0], hufts, tl[0], hufts, td[0], z);
+                    }
+                        mode = CODES;
+                        goto case CODES;
+                    case CODES:
+                        bitb=b; bitk=k;
+                        z.avail_in=n; z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        write=q;
+
+                        if ((r = codes.proc(this, z, r)) != Z_STREAM_END){
+                            return inflate_flush(z, r);
+                        }
+                        r = Z_OK;
+                        codes.free(z);
+
+                        p=z.next_in_index; n=z.avail_in;b=bitb;k=bitk;
+                        q=write;m=(int)(q<read?read-q-1:end-q);
+
+                        if (last==0){
+                            mode = TYPE;
+                            break;
+                        }
+                        mode = DRY;
+                        goto case DRY;
+                    case DRY:
+                        write=q; 
+                        r=inflate_flush(z, r); 
+                        q=write; m=(int)(q<read?read-q-1:end-q);
+                        if (read != write){
+                            bitb=b; bitk=k; 
+                            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                            write=q;
+                            return inflate_flush(z, r);
+                        }
+                        mode = DONE;
+                        goto case DONE;
+                    case DONE:
+                        r = Z_STREAM_END;
+
+                        bitb=b; bitk=k; 
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        write=q;
+                        return inflate_flush(z, r);
+                    case BAD:
+                        r = Z_DATA_ERROR;
+
+                        bitb=b; bitk=k; 
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        write=q;
+                        return inflate_flush(z, r);
+
+                    default:
+                        r = Z_STREAM_ERROR;
+
+                        bitb=b; bitk=k; 
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        write=q;
+                        return inflate_flush(z, r);
+                }
+            }
+        }
+
+        internal void free(ZStream z){
+            reset(z, null);
+            window=null;
+            hufts=null;
+            //ZFREE(z, s);
+        }
+
+        internal void set_dictionary(byte[] d, int start, int n){
+            System.Array.Copy(d, start, window, 0, n);
+            read = write = n;
+        }
+
+        // Returns true if inflate is currently at the end of a block generated
+        // by Z_SYNC_FLUSH or Z_FULL_FLUSH. 
+        internal int sync_point(){
+            return mode == LENS ? 1 : 0;
+        }
+
+        // copy as much as possible from the sliding window to the output area
+        internal int inflate_flush(ZStream z, int r){
+            int n;
+            int p;
+            int q;
+
+            // local copies of source and destination pointers
+            p = z.next_out_index;
+            q = read;
+
+            // compute number of bytes to copy as far as end of window
+            n = (int)((q <= write ? write : end) - q);
+            if (n > z.avail_out) n = z.avail_out;
+            if (n!=0 && r == Z_BUF_ERROR) r = Z_OK;
+
+            // update counters
+            z.avail_out -= n;
+            z.total_out += n;
+
+            // update check information
+            if(checkfn != null)
+                z.adler=check=z._adler.adler32(check, window, q, n);
+
+            // copy as far as end of window
+            System.Array.Copy(window, q, z.next_out, p, n);
+            p += n;
+            q += n;
+
+            // see if more to copy at beginning of window
+            if (q == end){
+                // wrap pointers
+                q = 0;
+                if (write == end)
+                    write = 0;
+
+                // compute bytes to copy
+                n = write - q;
+                if (n > z.avail_out) n = z.avail_out;
+                if (n!=0 && r == Z_BUF_ERROR) r = Z_OK;
+
+                // update counters
+                z.avail_out -= n;
+                z.total_out += n;
+
+                // update check information
+                if(checkfn != null)
+                    z.adler=check=z._adler.adler32(check, window, q, n);
+
+                // copy
+                System.Array.Copy(window, q, z.next_out, p, n);
+                p += n;
+                q += n;
+            }
+
+            // update pointers
+            z.next_out_index = p;
+            read = q;
+
+            // done
+            return r;
+        }
+    }
+}
\ No newline at end of file
diff --git a/Crypto/src/util/zlib/InfCodes.cs b/Crypto/src/util/zlib/InfCodes.cs
new file mode 100644
index 000000000..6fcafe458
--- /dev/null
+++ b/Crypto/src/util/zlib/InfCodes.cs
@@ -0,0 +1,611 @@
+using System;
+/*
+ * $Id: InfCodes.cs,v 1.2 2008-05-10 09:35:40 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    internal sealed class InfCodes{
+
+        private static readonly int[] inflate_mask = {
+                                                0x00000000, 0x00000001, 0x00000003, 0x00000007, 0x0000000f,
+                                                0x0000001f, 0x0000003f, 0x0000007f, 0x000000ff, 0x000001ff,
+                                                0x000003ff, 0x000007ff, 0x00000fff, 0x00001fff, 0x00003fff,
+                                                0x00007fff, 0x0000ffff
+                                            };
+
+        private const int Z_OK=0;
+        private const int Z_STREAM_END=1;
+        private const int Z_NEED_DICT=2;
+        private const int Z_ERRNO=-1;
+        private const int Z_STREAM_ERROR=-2;
+        private const int Z_DATA_ERROR=-3;
+        private const int Z_MEM_ERROR=-4;
+        private const int Z_BUF_ERROR=-5;
+        private const int Z_VERSION_ERROR=-6;
+
+        // waiting for "i:"=input,
+        //             "o:"=output,
+        //             "x:"=nothing
+        private const int START=0;  // x: set up for LEN
+        private const int LEN=1;    // i: get length/literal/eob next
+        private const int LENEXT=2; // i: getting length extra (have base)
+        private const int DIST=3;   // i: get distance next
+        private const int DISTEXT=4;// i: getting distance extra
+        private const int COPY=5;   // o: copying bytes in window, waiting for space
+        private const int LIT=6;    // o: got literal, waiting for output space
+        private const int WASH=7;   // o: got eob, possibly still output waiting
+        private const int END=8;    // x: got eob and all data flushed
+        private const int BADCODE=9;// x: got error
+
+        int mode;      // current inflate_codes mode
+
+        // mode dependent information
+        int len;
+
+        int[] tree; // pointer into tree
+        int tree_index=0;
+        int need;   // bits needed
+
+        int lit;
+
+        // if EXT or COPY, where and how much
+        int get;              // bits to get for extra
+        int dist;             // distance back to copy from
+
+        byte lbits;           // ltree bits decoded per branch
+        byte dbits;           // dtree bits decoder per branch
+        int[] ltree;          // literal/length/eob tree
+        int ltree_index;      // literal/length/eob tree
+        int[] dtree;          // distance tree
+        int dtree_index;      // distance tree
+
+        internal InfCodes(){
+        }
+        internal void init(int bl, int bd,
+            int[] tl, int tl_index,
+            int[] td, int td_index, ZStream z){
+            mode=START;
+            lbits=(byte)bl;
+            dbits=(byte)bd;
+            ltree=tl;
+            ltree_index=tl_index;
+            dtree = td;
+            dtree_index=td_index;
+            tree=null;
+        }
+
+        internal int proc(InfBlocks s, ZStream z, int r){ 
+            int j;              // temporary storage
+            int tindex;         // temporary pointer
+            int e;              // extra bits or operation
+            int b=0;            // bit buffer
+            int k=0;            // bits in bit buffer
+            int p=0;            // input data pointer
+            int n;              // bytes available there
+            int q;              // output window write pointer
+            int m;              // bytes to end of window or read pointer
+            int f;              // pointer to copy strings from
+
+            // copy input/output information to locals (UPDATE macro restores)
+            p=z.next_in_index;n=z.avail_in;b=s.bitb;k=s.bitk;
+            q=s.write;m=q<s.read?s.read-q-1:s.end-q;
+
+            // process input and output based on current state
+            while (true){
+                switch (mode){
+                        // waiting for "i:"=input, "o:"=output, "x:"=nothing
+                    case START:         // x: set up for LEN
+                        if (m >= 258 && n >= 10){
+
+                            s.bitb=b;s.bitk=k;
+                            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                            s.write=q;
+                            r = inflate_fast(lbits, dbits, 
+                                ltree, ltree_index, 
+                                dtree, dtree_index,
+                                s, z);
+
+                            p=z.next_in_index;n=z.avail_in;b=s.bitb;k=s.bitk;
+                            q=s.write;m=q<s.read?s.read-q-1:s.end-q;
+
+                            if (r != Z_OK){
+                                mode = r == Z_STREAM_END ? WASH : BADCODE;
+                                break;
+                            }
+                        }
+                        need = lbits;
+                        tree = ltree;
+                        tree_index=ltree_index;
+
+                        mode = LEN;
+                        goto case LEN;
+                    case LEN:           // i: get length/literal/eob next
+                        j = need;
+
+                        while(k<(j)){
+                            if(n!=0)r=Z_OK;
+                            else{
+
+                                s.bitb=b;s.bitk=k;
+                                z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                s.write=q;
+                                return s.inflate_flush(z,r);
+                            }
+                            n--;
+                            b|=(z.next_in[p++]&0xff)<<k;
+                            k+=8;
+                        }
+
+                        tindex=(tree_index+(b&inflate_mask[j]))*3;
+
+                        b>>=(tree[tindex+1]);
+                        k-=(tree[tindex+1]);
+
+                        e=tree[tindex];
+
+                        if(e == 0){               // literal
+                            lit = tree[tindex+2];
+                            mode = LIT;
+                            break;
+                        }
+                        if((e & 16)!=0 ){          // length
+                            get = e & 15;
+                            len = tree[tindex+2];
+                            mode = LENEXT;
+                            break;
+                        }
+                        if ((e & 64) == 0){        // next table
+                            need = e;
+                            tree_index = tindex/3+tree[tindex+2];
+                            break;
+                        }
+                        if ((e & 32)!=0){               // end of block
+                            mode = WASH;
+                            break;
+                        }
+                        mode = BADCODE;        // invalid code
+                        z.msg = "invalid literal/length code";
+                        r = Z_DATA_ERROR;
+
+                        s.bitb=b;s.bitk=k;
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        s.write=q;
+                        return s.inflate_flush(z,r);
+
+                    case LENEXT:        // i: getting length extra (have base)
+                        j = get;
+
+                        while(k<(j)){
+                            if(n!=0)r=Z_OK;
+                            else{
+
+                                s.bitb=b;s.bitk=k;
+                                z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                s.write=q;
+                                return s.inflate_flush(z,r);
+                            }
+                            n--; b|=(z.next_in[p++]&0xff)<<k;
+                            k+=8;
+                        }
+
+                        len += (b & inflate_mask[j]);
+
+                        b>>=j;
+                        k-=j;
+
+                        need = dbits;
+                        tree = dtree;
+                        tree_index=dtree_index;
+                        mode = DIST;
+                        goto case DIST;
+                    case DIST:          // i: get distance next
+                        j = need;
+
+                        while(k<(j)){
+                            if(n!=0)r=Z_OK;
+                            else{
+
+                                s.bitb=b;s.bitk=k;
+                                z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                s.write=q;
+                                return s.inflate_flush(z,r);
+                            }
+                            n--; b|=(z.next_in[p++]&0xff)<<k;
+                            k+=8;
+                        }
+
+                        tindex=(tree_index+(b & inflate_mask[j]))*3;
+
+                        b>>=tree[tindex+1];
+                        k-=tree[tindex+1];
+
+                        e = (tree[tindex]);
+                        if((e & 16)!=0){               // distance
+                            get = e & 15;
+                            dist = tree[tindex+2];
+                            mode = DISTEXT;
+                            break;
+                        }
+                        if ((e & 64) == 0){        // next table
+                            need = e;
+                            tree_index = tindex/3 + tree[tindex+2];
+                            break;
+                        }
+                        mode = BADCODE;        // invalid code
+                        z.msg = "invalid distance code";
+                        r = Z_DATA_ERROR;
+
+                        s.bitb=b;s.bitk=k;
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        s.write=q;
+                        return s.inflate_flush(z,r);
+
+                    case DISTEXT:       // i: getting distance extra
+                        j = get;
+
+                        while(k<(j)){
+                            if(n!=0)r=Z_OK;
+                            else{
+
+                                s.bitb=b;s.bitk=k;
+                                z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                s.write=q;
+                                return s.inflate_flush(z,r);
+                            }
+                            n--; b|=(z.next_in[p++]&0xff)<<k;
+                            k+=8;
+                        }
+
+                        dist += (b & inflate_mask[j]);
+
+                        b>>=j;
+                        k-=j;
+
+                        mode = COPY;
+                        goto case COPY;
+                    case COPY:          // o: copying bytes in window, waiting for space
+                        f = q - dist;
+                        while(f < 0){     // modulo window size-"while" instead
+                            f += s.end;     // of "if" handles invalid distances
+                        }
+                        while (len!=0){
+
+                            if(m==0){
+                                if(q==s.end&&s.read!=0){q=0;m=q<s.read?s.read-q-1:s.end-q;}
+                                if(m==0){
+                                    s.write=q; r=s.inflate_flush(z,r);
+                                    q=s.write;m=q<s.read?s.read-q-1:s.end-q;
+
+                                    if(q==s.end&&s.read!=0){q=0;m=q<s.read?s.read-q-1:s.end-q;}
+
+                                    if(m==0){
+                                        s.bitb=b;s.bitk=k;
+                                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                        s.write=q;
+                                        return s.inflate_flush(z,r);
+                                    }  
+                                }
+                            }
+
+                            s.window[q++]=s.window[f++]; m--;
+
+                            if (f == s.end)
+                                f = 0;
+                            len--;
+                        }
+                        mode = START;
+                        break;
+                    case LIT:           // o: got literal, waiting for output space
+                        if(m==0){
+                            if(q==s.end&&s.read!=0){q=0;m=q<s.read?s.read-q-1:s.end-q;}
+                            if(m==0){
+                                s.write=q; r=s.inflate_flush(z,r);
+                                q=s.write;m=q<s.read?s.read-q-1:s.end-q;
+
+                                if(q==s.end&&s.read!=0){q=0;m=q<s.read?s.read-q-1:s.end-q;}
+                                if(m==0){
+                                    s.bitb=b;s.bitk=k;
+                                    z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                    s.write=q;
+                                    return s.inflate_flush(z,r);
+                                }
+                            }
+                        }
+                        r=Z_OK;
+
+                        s.window[q++]=(byte)lit; m--;
+
+                        mode = START;
+                        break;
+                    case WASH:           // o: got eob, possibly more output
+                        if (k > 7){        // return unused byte, if any
+                            k -= 8;
+                            n++;
+                            p--;             // can always return one
+                        }
+
+                        s.write=q; r=s.inflate_flush(z,r);
+                        q=s.write;m=q<s.read?s.read-q-1:s.end-q;
+
+                        if (s.read != s.write){
+                            s.bitb=b;s.bitk=k;
+                            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                            s.write=q;
+                            return s.inflate_flush(z,r);
+                        }
+                        mode = END;
+                        goto case END;
+                    case END:
+                        r = Z_STREAM_END;
+                        s.bitb=b;s.bitk=k;
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        s.write=q;
+                        return s.inflate_flush(z,r);
+
+                    case BADCODE:       // x: got error
+
+                        r = Z_DATA_ERROR;
+
+                        s.bitb=b;s.bitk=k;
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        s.write=q;
+                        return s.inflate_flush(z,r);
+
+                    default:
+                        r = Z_STREAM_ERROR;
+
+                        s.bitb=b;s.bitk=k;
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        s.write=q;
+                        return s.inflate_flush(z,r);
+                }
+            }
+        }
+
+        internal void free(ZStream z){
+            //  ZFREE(z, c);
+        }
+
+        // Called with number of bytes left to write in window at least 258
+        // (the maximum string length) and number of input bytes available
+        // at least ten.  The ten bytes are six bytes for the longest length/
+        // distance pair plus four bytes for overloading the bit buffer.
+
+        internal int inflate_fast(int bl, int bd, 
+            int[] tl, int tl_index,
+            int[] td, int td_index,
+            InfBlocks s, ZStream z){
+            int t;                // temporary pointer
+            int[] tp;             // temporary pointer
+            int tp_index;         // temporary pointer
+            int e;                // extra bits or operation
+            int b;                // bit buffer
+            int k;                // bits in bit buffer
+            int p;                // input data pointer
+            int n;                // bytes available there
+            int q;                // output window write pointer
+            int m;                // bytes to end of window or read pointer
+            int ml;               // mask for literal/length tree
+            int md;               // mask for distance tree
+            int c;                // bytes to copy
+            int d;                // distance back to copy from
+            int r;                // copy source pointer
+
+            int tp_index_t_3;     // (tp_index+t)*3
+
+            // load input, output, bit values
+            p=z.next_in_index;n=z.avail_in;b=s.bitb;k=s.bitk;
+            q=s.write;m=q<s.read?s.read-q-1:s.end-q;
+
+            // initialize masks
+            ml = inflate_mask[bl];
+            md = inflate_mask[bd];
+
+            // do until not enough input or output space for fast loop
+            do {                          // assume called with m >= 258 && n >= 10
+                // get literal/length code
+            while(k<(20)){              // max bits for literal/length code
+                n--;
+                b|=(z.next_in[p++]&0xff)<<k;k+=8;
+            }
+
+                t= b&ml;
+                tp=tl; 
+                tp_index=tl_index;
+                tp_index_t_3=(tp_index+t)*3;
+                if ((e = tp[tp_index_t_3]) == 0){
+                    b>>=(tp[tp_index_t_3+1]); k-=(tp[tp_index_t_3+1]);
+
+                    s.window[q++] = (byte)tp[tp_index_t_3+2];
+                    m--;
+                    continue;
+                }
+                do {
+
+                    b>>=(tp[tp_index_t_3+1]); k-=(tp[tp_index_t_3+1]);
+
+                    if((e&16)!=0){
+                        e &= 15;
+                        c = tp[tp_index_t_3+2] + ((int)b & inflate_mask[e]);
+
+                        b>>=e; k-=e;
+
+                        // decode distance base of block to copy
+                        while(k<(15)){           // max bits for distance code
+                            n--;
+                            b|=(z.next_in[p++]&0xff)<<k;k+=8;
+                        }
+
+                        t= b&md;
+                        tp=td;
+                        tp_index=td_index;
+                        tp_index_t_3=(tp_index+t)*3;
+                        e = tp[tp_index_t_3];
+
+                        do {
+
+                            b>>=(tp[tp_index_t_3+1]); k-=(tp[tp_index_t_3+1]);
+
+                            if((e&16)!=0){
+                                // get extra bits to add to distance base
+                                e &= 15;
+                                while(k<(e)){         // get extra bits (up to 13)
+                                    n--;
+                                    b|=(z.next_in[p++]&0xff)<<k;k+=8;
+                                }
+
+                                d = tp[tp_index_t_3+2] + (b&inflate_mask[e]);
+
+                                b>>=(e); k-=(e);
+
+                                // do the copy
+                                m -= c;
+                                if (q >= d){                // offset before dest
+                                    //  just copy
+                                    r=q-d;
+                                    if(q-r>0 && 2>(q-r)){           
+                                        s.window[q++]=s.window[r++]; // minimum count is three,
+                                        s.window[q++]=s.window[r++]; // so unroll loop a little
+                                        c-=2;
+                                    }
+                                    else{
+                                        System.Array.Copy(s.window, r, s.window, q, 2);
+                                        q+=2; r+=2; c-=2;
+                                    }
+                                }
+                                else{                  // else offset after destination
+                                    r=q-d;
+                                    do{
+                                        r+=s.end;          // force pointer in window
+                                    }while(r<0);         // covers invalid distances
+                                    e=s.end-r;
+                                    if(c>e){             // if source crosses,
+                                        c-=e;              // wrapped copy
+                                        if(q-r>0 && e>(q-r)){           
+                                            do{s.window[q++] = s.window[r++];}
+                                            while(--e!=0);
+                                        }
+                                        else{
+                                            System.Array.Copy(s.window, r, s.window, q, e);
+                                            q+=e; r+=e; e=0;
+                                        }
+                                        r = 0;                  // copy rest from start of window
+                                    }
+
+                                }
+
+                                // copy all or what's left
+                                if(q-r>0 && c>(q-r)){           
+                                    do{s.window[q++] = s.window[r++];}
+                                    while(--c!=0);
+                                }
+                                else{
+                                    System.Array.Copy(s.window, r, s.window, q, c);
+                                    q+=c; r+=c; c=0;
+                                }
+                                break;
+                            }
+                            else if((e&64)==0){
+                                t+=tp[tp_index_t_3+2];
+                                t+=(b&inflate_mask[e]);
+                                tp_index_t_3=(tp_index+t)*3;
+                                e=tp[tp_index_t_3];
+                            }
+                            else{
+                                z.msg = "invalid distance code";
+
+                                c=z.avail_in-n;c=(k>>3)<c?k>>3:c;n+=c;p-=c;k-=c<<3;
+
+                                s.bitb=b;s.bitk=k;
+                                z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                                s.write=q;
+
+                                return Z_DATA_ERROR;
+                            }
+                        }
+                        while(true);
+                        break;
+                    }
+
+                    if((e&64)==0){
+                        t+=tp[tp_index_t_3+2];
+                        t+=(b&inflate_mask[e]);
+                        tp_index_t_3=(tp_index+t)*3;
+                        if((e=tp[tp_index_t_3])==0){
+
+                            b>>=(tp[tp_index_t_3+1]); k-=(tp[tp_index_t_3+1]);
+
+                            s.window[q++]=(byte)tp[tp_index_t_3+2];
+                            m--;
+                            break;
+                        }
+                    }
+                    else if((e&32)!=0){
+
+                        c=z.avail_in-n;c=(k>>3)<c?k>>3:c;n+=c;p-=c;k-=c<<3;
+ 
+                        s.bitb=b;s.bitk=k;
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        s.write=q;
+
+                        return Z_STREAM_END;
+                    }
+                    else{
+                        z.msg="invalid literal/length code";
+
+                        c=z.avail_in-n;c=(k>>3)<c?k>>3:c;n+=c;p-=c;k-=c<<3;
+
+                        s.bitb=b;s.bitk=k;
+                        z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+                        s.write=q;
+
+                        return Z_DATA_ERROR;
+                    }
+                } 
+                while(true);
+            } 
+            while(m>=258 && n>= 10);
+
+            // not enough input or output--restore pointers and return
+            c=z.avail_in-n;c=(k>>3)<c?k>>3:c;n+=c;p-=c;k-=c<<3;
+
+            s.bitb=b;s.bitk=k;
+            z.avail_in=n;z.total_in+=p-z.next_in_index;z.next_in_index=p;
+            s.write=q;
+
+            return Z_OK;
+        }
+    }
+}
\ No newline at end of file
diff --git a/Crypto/src/util/zlib/InfTree.cs b/Crypto/src/util/zlib/InfTree.cs
new file mode 100644
index 000000000..6ed7d1920
--- /dev/null
+++ b/Crypto/src/util/zlib/InfTree.cs
@@ -0,0 +1,523 @@
+using System;
+/*
+ * $Id: InfTree.cs,v 1.2 2008-05-10 09:35:40 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    internal sealed class InfTree{
+
+        private const int MANY=1440;
+
+        private const int Z_OK=0;
+        private const int Z_STREAM_END=1;
+        private const int Z_NEED_DICT=2;
+        private const int Z_ERRNO=-1;
+        private const int Z_STREAM_ERROR=-2;
+        private const int Z_DATA_ERROR=-3;
+        private const int Z_MEM_ERROR=-4;
+        private const int Z_BUF_ERROR=-5;
+        private const int Z_VERSION_ERROR=-6;
+
+        private const int fixed_bl = 9;
+        private const int fixed_bd = 5;
+
+		static readonly int[] fixed_tl = {
+                                    96,7,256, 0,8,80, 0,8,16, 84,8,115,
+                                    82,7,31, 0,8,112, 0,8,48, 0,9,192,
+                                    80,7,10, 0,8,96, 0,8,32, 0,9,160,
+                                    0,8,0, 0,8,128, 0,8,64, 0,9,224,
+                                    80,7,6, 0,8,88, 0,8,24, 0,9,144,
+                                    83,7,59, 0,8,120, 0,8,56, 0,9,208,
+                                    81,7,17, 0,8,104, 0,8,40, 0,9,176,
+                                    0,8,8, 0,8,136, 0,8,72, 0,9,240,
+                                    80,7,4, 0,8,84, 0,8,20, 85,8,227,
+                                    83,7,43, 0,8,116, 0,8,52, 0,9,200,
+                                    81,7,13, 0,8,100, 0,8,36, 0,9,168,
+                                    0,8,4, 0,8,132, 0,8,68, 0,9,232,
+                                    80,7,8, 0,8,92, 0,8,28, 0,9,152,
+                                    84,7,83, 0,8,124, 0,8,60, 0,9,216,
+                                    82,7,23, 0,8,108, 0,8,44, 0,9,184,
+                                    0,8,12, 0,8,140, 0,8,76, 0,9,248,
+                                    80,7,3, 0,8,82, 0,8,18, 85,8,163,
+                                    83,7,35, 0,8,114, 0,8,50, 0,9,196,
+                                    81,7,11, 0,8,98, 0,8,34, 0,9,164,
+                                    0,8,2, 0,8,130, 0,8,66, 0,9,228,
+                                    80,7,7, 0,8,90, 0,8,26, 0,9,148,
+                                    84,7,67, 0,8,122, 0,8,58, 0,9,212,
+                                    82,7,19, 0,8,106, 0,8,42, 0,9,180,
+                                    0,8,10, 0,8,138, 0,8,74, 0,9,244,
+                                    80,7,5, 0,8,86, 0,8,22, 192,8,0,
+                                    83,7,51, 0,8,118, 0,8,54, 0,9,204,
+                                    81,7,15, 0,8,102, 0,8,38, 0,9,172,
+                                    0,8,6, 0,8,134, 0,8,70, 0,9,236,
+                                    80,7,9, 0,8,94, 0,8,30, 0,9,156,
+                                    84,7,99, 0,8,126, 0,8,62, 0,9,220,
+                                    82,7,27, 0,8,110, 0,8,46, 0,9,188,
+                                    0,8,14, 0,8,142, 0,8,78, 0,9,252,
+                                    96,7,256, 0,8,81, 0,8,17, 85,8,131,
+                                    82,7,31, 0,8,113, 0,8,49, 0,9,194,
+                                    80,7,10, 0,8,97, 0,8,33, 0,9,162,
+                                    0,8,1, 0,8,129, 0,8,65, 0,9,226,
+                                    80,7,6, 0,8,89, 0,8,25, 0,9,146,
+                                    83,7,59, 0,8,121, 0,8,57, 0,9,210,
+                                    81,7,17, 0,8,105, 0,8,41, 0,9,178,
+                                    0,8,9, 0,8,137, 0,8,73, 0,9,242,
+                                    80,7,4, 0,8,85, 0,8,21, 80,8,258,
+                                    83,7,43, 0,8,117, 0,8,53, 0,9,202,
+                                    81,7,13, 0,8,101, 0,8,37, 0,9,170,
+                                    0,8,5, 0,8,133, 0,8,69, 0,9,234,
+                                    80,7,8, 0,8,93, 0,8,29, 0,9,154,
+                                    84,7,83, 0,8,125, 0,8,61, 0,9,218,
+                                    82,7,23, 0,8,109, 0,8,45, 0,9,186,
+                                    0,8,13, 0,8,141, 0,8,77, 0,9,250,
+                                    80,7,3, 0,8,83, 0,8,19, 85,8,195,
+                                    83,7,35, 0,8,115, 0,8,51, 0,9,198,
+                                    81,7,11, 0,8,99, 0,8,35, 0,9,166,
+                                    0,8,3, 0,8,131, 0,8,67, 0,9,230,
+                                    80,7,7, 0,8,91, 0,8,27, 0,9,150,
+                                    84,7,67, 0,8,123, 0,8,59, 0,9,214,
+                                    82,7,19, 0,8,107, 0,8,43, 0,9,182,
+                                    0,8,11, 0,8,139, 0,8,75, 0,9,246,
+                                    80,7,5, 0,8,87, 0,8,23, 192,8,0,
+                                    83,7,51, 0,8,119, 0,8,55, 0,9,206,
+                                    81,7,15, 0,8,103, 0,8,39, 0,9,174,
+                                    0,8,7, 0,8,135, 0,8,71, 0,9,238,
+                                    80,7,9, 0,8,95, 0,8,31, 0,9,158,
+                                    84,7,99, 0,8,127, 0,8,63, 0,9,222,
+                                    82,7,27, 0,8,111, 0,8,47, 0,9,190,
+                                    0,8,15, 0,8,143, 0,8,79, 0,9,254,
+                                    96,7,256, 0,8,80, 0,8,16, 84,8,115,
+                                    82,7,31, 0,8,112, 0,8,48, 0,9,193,
+
+                                    80,7,10, 0,8,96, 0,8,32, 0,9,161,
+                                    0,8,0, 0,8,128, 0,8,64, 0,9,225,
+                                    80,7,6, 0,8,88, 0,8,24, 0,9,145,
+                                    83,7,59, 0,8,120, 0,8,56, 0,9,209,
+                                    81,7,17, 0,8,104, 0,8,40, 0,9,177,
+                                    0,8,8, 0,8,136, 0,8,72, 0,9,241,
+                                    80,7,4, 0,8,84, 0,8,20, 85,8,227,
+                                    83,7,43, 0,8,116, 0,8,52, 0,9,201,
+                                    81,7,13, 0,8,100, 0,8,36, 0,9,169,
+                                    0,8,4, 0,8,132, 0,8,68, 0,9,233,
+                                    80,7,8, 0,8,92, 0,8,28, 0,9,153,
+                                    84,7,83, 0,8,124, 0,8,60, 0,9,217,
+                                    82,7,23, 0,8,108, 0,8,44, 0,9,185,
+                                    0,8,12, 0,8,140, 0,8,76, 0,9,249,
+                                    80,7,3, 0,8,82, 0,8,18, 85,8,163,
+                                    83,7,35, 0,8,114, 0,8,50, 0,9,197,
+                                    81,7,11, 0,8,98, 0,8,34, 0,9,165,
+                                    0,8,2, 0,8,130, 0,8,66, 0,9,229,
+                                    80,7,7, 0,8,90, 0,8,26, 0,9,149,
+                                    84,7,67, 0,8,122, 0,8,58, 0,9,213,
+                                    82,7,19, 0,8,106, 0,8,42, 0,9,181,
+                                    0,8,10, 0,8,138, 0,8,74, 0,9,245,
+                                    80,7,5, 0,8,86, 0,8,22, 192,8,0,
+                                    83,7,51, 0,8,118, 0,8,54, 0,9,205,
+                                    81,7,15, 0,8,102, 0,8,38, 0,9,173,
+                                    0,8,6, 0,8,134, 0,8,70, 0,9,237,
+                                    80,7,9, 0,8,94, 0,8,30, 0,9,157,
+                                    84,7,99, 0,8,126, 0,8,62, 0,9,221,
+                                    82,7,27, 0,8,110, 0,8,46, 0,9,189,
+                                    0,8,14, 0,8,142, 0,8,78, 0,9,253,
+                                    96,7,256, 0,8,81, 0,8,17, 85,8,131,
+                                    82,7,31, 0,8,113, 0,8,49, 0,9,195,
+                                    80,7,10, 0,8,97, 0,8,33, 0,9,163,
+                                    0,8,1, 0,8,129, 0,8,65, 0,9,227,
+                                    80,7,6, 0,8,89, 0,8,25, 0,9,147,
+                                    83,7,59, 0,8,121, 0,8,57, 0,9,211,
+                                    81,7,17, 0,8,105, 0,8,41, 0,9,179,
+                                    0,8,9, 0,8,137, 0,8,73, 0,9,243,
+                                    80,7,4, 0,8,85, 0,8,21, 80,8,258,
+                                    83,7,43, 0,8,117, 0,8,53, 0,9,203,
+                                    81,7,13, 0,8,101, 0,8,37, 0,9,171,
+                                    0,8,5, 0,8,133, 0,8,69, 0,9,235,
+                                    80,7,8, 0,8,93, 0,8,29, 0,9,155,
+                                    84,7,83, 0,8,125, 0,8,61, 0,9,219,
+                                    82,7,23, 0,8,109, 0,8,45, 0,9,187,
+                                    0,8,13, 0,8,141, 0,8,77, 0,9,251,
+                                    80,7,3, 0,8,83, 0,8,19, 85,8,195,
+                                    83,7,35, 0,8,115, 0,8,51, 0,9,199,
+                                    81,7,11, 0,8,99, 0,8,35, 0,9,167,
+                                    0,8,3, 0,8,131, 0,8,67, 0,9,231,
+                                    80,7,7, 0,8,91, 0,8,27, 0,9,151,
+                                    84,7,67, 0,8,123, 0,8,59, 0,9,215,
+                                    82,7,19, 0,8,107, 0,8,43, 0,9,183,
+                                    0,8,11, 0,8,139, 0,8,75, 0,9,247,
+                                    80,7,5, 0,8,87, 0,8,23, 192,8,0,
+                                    83,7,51, 0,8,119, 0,8,55, 0,9,207,
+                                    81,7,15, 0,8,103, 0,8,39, 0,9,175,
+                                    0,8,7, 0,8,135, 0,8,71, 0,9,239,
+                                    80,7,9, 0,8,95, 0,8,31, 0,9,159,
+                                    84,7,99, 0,8,127, 0,8,63, 0,9,223,
+                                    82,7,27, 0,8,111, 0,8,47, 0,9,191,
+                                    0,8,15, 0,8,143, 0,8,79, 0,9,255
+                                };
+        static readonly int[] fixed_td = {
+                                    80,5,1, 87,5,257, 83,5,17, 91,5,4097,
+                                    81,5,5, 89,5,1025, 85,5,65, 93,5,16385,
+                                    80,5,3, 88,5,513, 84,5,33, 92,5,8193,
+                                    82,5,9, 90,5,2049, 86,5,129, 192,5,24577,
+                                    80,5,2, 87,5,385, 83,5,25, 91,5,6145,
+                                    81,5,7, 89,5,1537, 85,5,97, 93,5,24577,
+                                    80,5,4, 88,5,769, 84,5,49, 92,5,12289,
+                                    82,5,13, 90,5,3073, 86,5,193, 192,5,24577
+                                };
+
+        // Tables for deflate from PKZIP's appnote.txt.
+        static readonly int[] cplens = { // Copy lengths for literal codes 257..285
+                                  3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 15, 17, 19, 23, 27, 31,
+                                  35, 43, 51, 59, 67, 83, 99, 115, 131, 163, 195, 227, 258, 0, 0
+                              };
+
+        // see note #13 above about 258
+        static readonly int[] cplext = { // Extra bits for literal codes 257..285
+                                  0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2,
+                                  3, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5, 0, 112, 112  // 112==invalid
+                              };
+
+        static readonly int[] cpdist = { // Copy offsets for distance codes 0..29
+                                  1, 2, 3, 4, 5, 7, 9, 13, 17, 25, 33, 49, 65, 97, 129, 193,
+                                  257, 385, 513, 769, 1025, 1537, 2049, 3073, 4097, 6145,
+                                  8193, 12289, 16385, 24577
+                              };
+
+        static readonly int[] cpdext = { // Extra bits for distance codes
+                                  0, 0, 0, 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6,
+                                  7, 7, 8, 8, 9, 9, 10, 10, 11, 11,
+                                  12, 12, 13, 13};
+
+        // If BMAX needs to be larger than 16, then h and x[] should be uLong.
+        const int BMAX=15;         // maximum bit length of any code
+
+        int[] hn = null;  // hufts used in space
+        int[] v = null;   // work area for huft_build 
+        int[] c = null;   // bit length count table
+        int[] r = null;   // table entry for structure assignment
+        int[] u = null;   // table stack
+        int[] x = null;   // bit offsets, then code stack
+
+        private int huft_build(int[] b, // code lengths in bits (all assumed <= BMAX)
+            int bindex, 
+            int n,   // number of codes (assumed <= 288)
+            int s,   // number of simple-valued codes (0..s-1)
+            int[] d, // list of base values for non-simple codes
+            int[] e, // list of extra bits for non-simple codes
+            int[] t, // result: starting table
+            int[] m, // maximum lookup bits, returns actual
+            int[] hp,// space for trees
+            int[] hn,// hufts used in space
+            int[] v  // working area: values in order of bit length
+            ){
+            // Given a list of code lengths and a maximum table size, make a set of
+            // tables to decode that set of codes.  Return Z_OK on success, Z_BUF_ERROR
+            // if the given code set is incomplete (the tables are still built in this
+            // case), Z_DATA_ERROR if the input is invalid (an over-subscribed set of
+            // lengths), or Z_MEM_ERROR if not enough memory.
+
+            int a;                       // counter for codes of length k
+            int f;                       // i repeats in table every f entries
+            int g;                       // maximum code length
+            int h;                       // table level
+            int i;                       // counter, current code
+            int j;                       // counter
+            int k;                       // number of bits in current code
+            int l;                       // bits per table (returned in m)
+            int mask;                    // (1 << w) - 1, to avoid cc -O bug on HP
+            int p;                       // pointer into c[], b[], or v[]
+            int q;                       // points to current table
+            int w;                       // bits before this table == (l * h)
+            int xp;                      // pointer into x
+            int y;                       // number of dummy codes added
+            int z;                       // number of entries in current table
+
+            // Generate counts for each bit length
+
+            p = 0; i = n;
+            do {
+                c[b[bindex+p]]++; p++; i--;   // assume all entries <= BMAX
+            }while(i!=0);
+
+            if(c[0] == n){                // null input--all zero length codes
+                t[0] = -1;
+                m[0] = 0;
+                return Z_OK;
+            }
+
+            // Find minimum and maximum length, bound *m by those
+            l = m[0];
+            for (j = 1; j <= BMAX; j++)
+                if(c[j]!=0) break;
+            k = j;                        // minimum code length
+            if(l < j){
+                l = j;
+            }
+            for (i = BMAX; i!=0; i--){
+                if(c[i]!=0) break;
+            }
+            g = i;                        // maximum code length
+            if(l > i){
+                l = i;
+            }
+            m[0] = l;
+
+            // Adjust last length count to fill out codes, if needed
+            for (y = 1 << j; j < i; j++, y <<= 1){
+                if ((y -= c[j]) < 0){
+                    return Z_DATA_ERROR;
+                }
+            }
+            if ((y -= c[i]) < 0){
+                return Z_DATA_ERROR;
+            }
+            c[i] += y;
+
+            // Generate starting offsets into the value table for each length
+            x[1] = j = 0;
+            p = 1;  xp = 2;
+            while (--i!=0) {                 // note that i == g from above
+                x[xp] = (j += c[p]);
+                xp++;
+                p++;
+            }
+
+            // Make a table of values in order of bit lengths
+            i = 0; p = 0;
+            do {
+                if ((j = b[bindex+p]) != 0){
+                    v[x[j]++] = i;
+                }
+                p++;
+            }
+            while (++i < n);
+            n = x[g];                     // set n to length of v
+
+            // Generate the Huffman codes and for each, make the table entries
+            x[0] = i = 0;                 // first Huffman code is zero
+            p = 0;                        // grab values in bit order
+            h = -1;                       // no tables yet--level -1
+            w = -l;                       // bits decoded == (l * h)
+            u[0] = 0;                     // just to keep compilers happy
+            q = 0;                        // ditto
+            z = 0;                        // ditto
+
+            // go through the bit lengths (k already is bits in shortest code)
+            for (; k <= g; k++){
+                a = c[k];
+                while (a--!=0){
+                    // here i is the Huffman code of length k bits for value *p
+                    // make tables up to required level
+                    while (k > w + l){
+                        h++;
+                        w += l;                 // previous table always l bits
+                        // compute minimum size table less than or equal to l bits
+                        z = g - w;
+                        z = (z > l) ? l : z;        // table size upper limit
+                        if((f=1<<(j=k-w))>a+1){     // try a k-w bit table
+                            // too few codes for k-w bit table
+                            f -= a + 1;               // deduct codes from patterns left
+                            xp = k;
+                            if(j < z){
+                                while (++j < z){        // try smaller tables up to z bits
+                                    if((f <<= 1) <= c[++xp])
+                                        break;              // enough codes to use up j bits
+                                    f -= c[xp];           // else deduct codes from patterns
+                                }
+                            }
+                        }
+                        z = 1 << j;                 // table entries for j-bit table
+
+                        // allocate new table
+                        if (hn[0] + z > MANY){       // (note: doesn't matter for fixed)
+                            return Z_DATA_ERROR;       // overflow of MANY
+                        }
+                        u[h] = q = /*hp+*/ hn[0];   // DEBUG
+                        hn[0] += z;
+ 
+                        // connect to last table, if there is one
+                        if(h!=0){
+                            x[h]=i;           // save pattern for backing up
+                            r[0]=(byte)j;     // bits in this table
+                            r[1]=(byte)l;     // bits to dump before this table
+                            j=i>>(w - l);
+                            r[2] = (int)(q - u[h-1] - j);               // offset to this table
+                            System.Array.Copy(r, 0, hp, (u[h-1]+j)*3, 3); // connect to last table
+                        }
+                        else{
+                            t[0] = q;               // first table is returned result
+                        }
+                    }
+
+                    // set up table entry in r
+                    r[1] = (byte)(k - w);
+                    if (p >= n){
+                        r[0] = 128 + 64;      // out of values--invalid code
+                    }
+                    else if (v[p] < s){
+                        r[0] = (byte)(v[p] < 256 ? 0 : 32 + 64);  // 256 is end-of-block
+                        r[2] = v[p++];          // simple code is just the value
+                    }
+                    else{
+                        r[0]=(byte)(e[v[p]-s]+16+64); // non-simple--look up in lists
+                        r[2]=d[v[p++] - s];
+                    }
+
+                    // fill code-like entries with r
+                    f=1<<(k-w);
+                    for (j=i>>w;j<z;j+=f){
+                        System.Array.Copy(r, 0, hp, (q+j)*3, 3);
+                    }
+
+                    // backwards increment the k-bit code i
+                    for (j = 1 << (k - 1); (i & j)!=0; j >>= 1){
+                        i ^= j;
+                    }
+                    i ^= j;
+
+                    // backup over finished tables
+                    mask = (1 << w) - 1;      // needed on HP, cc -O bug
+                    while ((i & mask) != x[h]){
+                        h--;                    // don't need to update q
+                        w -= l;
+                        mask = (1 << w) - 1;
+                    }
+                }
+            }
+            // Return Z_BUF_ERROR if we were given an incomplete table
+            return y != 0 && g != 1 ? Z_BUF_ERROR : Z_OK;
+        }
+
+        internal int inflate_trees_bits(int[] c,  // 19 code lengths
+            int[] bb, // bits tree desired/actual depth
+            int[] tb, // bits tree result
+            int[] hp, // space for trees
+            ZStream z // for messages
+            ){
+            int result;
+            initWorkArea(19);
+            hn[0]=0;
+            result = huft_build(c, 0, 19, 19, null, null, tb, bb, hp, hn, v);
+
+            if(result == Z_DATA_ERROR){
+                z.msg = "oversubscribed dynamic bit lengths tree";
+            }
+            else if(result == Z_BUF_ERROR || bb[0] == 0){
+                z.msg = "incomplete dynamic bit lengths tree";
+                result = Z_DATA_ERROR;
+            }
+            return result;
+        }
+
+        internal int inflate_trees_dynamic(int nl,   // number of literal/length codes
+            int nd,   // number of distance codes
+            int[] c,  // that many (total) code lengths
+            int[] bl, // literal desired/actual bit depth
+            int[] bd, // distance desired/actual bit depth 
+            int[] tl, // literal/length tree result
+            int[] td, // distance tree result
+            int[] hp, // space for trees
+            ZStream z // for messages
+            ){
+            int result;
+
+            // build literal/length tree
+            initWorkArea(288);
+            hn[0]=0;
+            result = huft_build(c, 0, nl, 257, cplens, cplext, tl, bl, hp, hn, v);
+            if (result != Z_OK || bl[0] == 0){
+                if(result == Z_DATA_ERROR){
+                    z.msg = "oversubscribed literal/length tree";
+                }
+                else if (result != Z_MEM_ERROR){
+                    z.msg = "incomplete literal/length tree";
+                    result = Z_DATA_ERROR;
+                }
+                return result;
+            }
+
+            // build distance tree
+            initWorkArea(288);
+            result = huft_build(c, nl, nd, 0, cpdist, cpdext, td, bd, hp, hn, v);
+
+            if (result != Z_OK || (bd[0] == 0 && nl > 257)){
+                if (result == Z_DATA_ERROR){
+                    z.msg = "oversubscribed distance tree";
+                }
+                else if (result == Z_BUF_ERROR) {
+                    z.msg = "incomplete distance tree";
+                    result = Z_DATA_ERROR;
+                }
+                else if (result != Z_MEM_ERROR){
+                    z.msg = "empty distance tree with lengths";
+                    result = Z_DATA_ERROR;
+                }
+                return result;
+            }
+
+            return Z_OK;
+        }
+
+        internal static int inflate_trees_fixed(int[] bl,  //literal desired/actual bit depth
+            int[] bd,  //distance desired/actual bit depth
+            int[][] tl,//literal/length tree result
+            int[][] td,//distance tree result 
+            ZStream z  //for memory allocation
+            ){
+            bl[0]=fixed_bl;
+            bd[0]=fixed_bd;
+            tl[0]=fixed_tl;
+            td[0]=fixed_td;
+            return Z_OK;
+        }
+
+        private void initWorkArea(int vsize){
+            if(hn==null){
+                hn=new int[1];
+                v=new int[vsize];
+                c=new int[BMAX+1];
+                r=new int[3];
+                u=new int[BMAX];
+                x=new int[BMAX+1];
+            }
+            if(v.Length<vsize){ v=new int[vsize]; }
+            for(int i=0; i<vsize; i++){v[i]=0;}
+            for(int i=0; i<BMAX+1; i++){c[i]=0;}
+            for(int i=0; i<3; i++){r[i]=0;}
+            //  for(int i=0; i<BMAX; i++){u[i]=0;}
+            System.Array.Copy(c, 0, u, 0, BMAX);
+            //  for(int i=0; i<BMAX+1; i++){x[i]=0;}
+            System.Array.Copy(c, 0, x, 0, BMAX+1);
+        }
+    }
+}
\ No newline at end of file
diff --git a/Crypto/src/util/zlib/Inflate.cs b/Crypto/src/util/zlib/Inflate.cs
new file mode 100644
index 000000000..ac8789332
--- /dev/null
+++ b/Crypto/src/util/zlib/Inflate.cs
@@ -0,0 +1,387 @@
+using System;
+/*
+ * $Id: Inflate.cs,v 1.2 2008-05-10 09:35:40 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    internal sealed class Inflate{
+  
+        private const int MAX_WBITS=15; // 32K LZ77 window
+
+        // preset dictionary flag in zlib header
+        private const int PRESET_DICT=0x20;
+
+        internal const int Z_NO_FLUSH=0;
+        internal const int Z_PARTIAL_FLUSH=1;
+        internal const int Z_SYNC_FLUSH=2;
+        internal const int Z_FULL_FLUSH=3;
+        internal const int Z_FINISH=4;
+
+        private const int Z_DEFLATED=8;
+
+        private const int Z_OK=0;
+        private const int Z_STREAM_END=1;
+        private const int Z_NEED_DICT=2;
+        private const int Z_ERRNO=-1;
+        private const int Z_STREAM_ERROR=-2;
+        private const int Z_DATA_ERROR=-3;
+        private const int Z_MEM_ERROR=-4;
+        private const int Z_BUF_ERROR=-5;
+        private const int Z_VERSION_ERROR=-6;
+
+        private const int METHOD=0;   // waiting for method byte
+        private const int FLAG=1;     // waiting for flag byte
+        private const int DICT4=2;    // four dictionary check bytes to go
+        private const int DICT3=3;    // three dictionary check bytes to go
+        private const int DICT2=4;    // two dictionary check bytes to go
+        private const int DICT1=5;    // one dictionary check byte to go
+        private const int DICT0=6;    // waiting for inflateSetDictionary
+        private const int BLOCKS=7;   // decompressing blocks
+        private const int CHECK4=8;   // four check bytes to go
+        private const int CHECK3=9;   // three check bytes to go
+        private const int CHECK2=10;  // two check bytes to go
+        private const int CHECK1=11;  // one check byte to go
+        private const int DONE=12;    // finished check, done
+        private const int BAD=13;     // got an error--stay here
+
+        internal int mode;                            // current inflate mode
+
+        // mode dependent information
+        internal int method;        // if FLAGS, method byte
+
+        // if CHECK, check values to compare
+        internal long[] was=new long[1] ; // computed check value
+        internal long need;               // stream check value
+
+        // if BAD, inflateSync's marker bytes count
+        internal int marker;
+
+        // mode independent information
+        internal int  nowrap;          // flag for no wrapper
+        internal int wbits;            // log2(window size)  (8..15, defaults to 15)
+
+        internal InfBlocks blocks;     // current inflate_blocks state
+
+        internal int inflateReset(ZStream z){
+            if(z == null || z.istate == null) return Z_STREAM_ERROR;
+    
+            z.total_in = z.total_out = 0;
+            z.msg = null;
+            z.istate.mode = z.istate.nowrap!=0 ? BLOCKS : METHOD;
+            z.istate.blocks.reset(z, null);
+            return Z_OK;
+        }
+
+        internal int inflateEnd(ZStream z){
+            if(blocks != null)
+                blocks.free(z);
+            blocks=null;
+            //    ZFREE(z, z->state);
+            return Z_OK;
+        }
+
+        internal int inflateInit(ZStream z, int w){
+            z.msg = null;
+            blocks = null;
+
+            // handle undocumented nowrap option (no zlib header or check)
+            nowrap = 0;
+            if(w < 0){
+                w = - w;
+                nowrap = 1;
+            }
+
+            // set window size
+            if(w<8 ||w>15){
+                inflateEnd(z);
+                return Z_STREAM_ERROR;
+            }
+            wbits=w;
+
+            z.istate.blocks=new InfBlocks(z, 
+                z.istate.nowrap!=0 ? null : this,
+                1<<w);
+
+            // reset state
+            inflateReset(z);
+            return Z_OK;
+        }
+
+        internal int inflate(ZStream z, int f){
+            int r;
+            int b;
+
+            if(z == null || z.istate == null || z.next_in == null)
+                return Z_STREAM_ERROR;
+            f = f == Z_FINISH ? Z_BUF_ERROR : Z_OK;
+            r = Z_BUF_ERROR;
+            while (true){
+                //System.out.println("mode: "+z.istate.mode);
+                switch (z.istate.mode){
+                    case METHOD:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        if(((z.istate.method = z.next_in[z.next_in_index++])&0xf)!=Z_DEFLATED){
+                            z.istate.mode = BAD;
+                            z.msg="unknown compression method";
+                            z.istate.marker = 5;       // can't try inflateSync
+                            break;
+                        }
+                        if((z.istate.method>>4)+8>z.istate.wbits){
+                            z.istate.mode = BAD;
+                            z.msg="invalid window size";
+                            z.istate.marker = 5;       // can't try inflateSync
+                            break;
+                        }
+                        z.istate.mode=FLAG;
+                        goto case FLAG;
+                    case FLAG:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        b = (z.next_in[z.next_in_index++])&0xff;
+
+                        if((((z.istate.method << 8)+b) % 31)!=0){
+                            z.istate.mode = BAD;
+                            z.msg = "incorrect header check";
+                            z.istate.marker = 5;       // can't try inflateSync
+                            break;
+                        }
+
+                        if((b&PRESET_DICT)==0){
+                            z.istate.mode = BLOCKS;
+                            break;
+                        }
+                        z.istate.mode = DICT4;
+                        goto case DICT4;
+                    case DICT4:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        z.istate.need=((z.next_in[z.next_in_index++]&0xff)<<24)&0xff000000L;
+                        z.istate.mode=DICT3;
+                        goto case DICT3;
+                    case DICT3:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        z.istate.need+=((z.next_in[z.next_in_index++]&0xff)<<16)&0xff0000L;
+                        z.istate.mode=DICT2;
+                        goto case DICT2;
+                    case DICT2:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        z.istate.need+=((z.next_in[z.next_in_index++]&0xff)<<8)&0xff00L;
+                        z.istate.mode=DICT1;
+                        goto case DICT1;
+                    case DICT1:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        z.istate.need += (z.next_in[z.next_in_index++]&0xffL);
+                        z.adler = z.istate.need;
+                        z.istate.mode = DICT0;
+                        return Z_NEED_DICT;
+                    case DICT0:
+                        z.istate.mode = BAD;
+                        z.msg = "need dictionary";
+                        z.istate.marker = 0;       // can try inflateSync
+                        return Z_STREAM_ERROR;
+                    case BLOCKS:
+
+                        r = z.istate.blocks.proc(z, r);
+                        if(r == Z_DATA_ERROR){
+                            z.istate.mode = BAD;
+                            z.istate.marker = 0;     // can try inflateSync
+                            break;
+                        }
+                        if(r == Z_OK){
+                            r = f;
+                        }
+                        if(r != Z_STREAM_END){
+                            return r;
+                        }
+                        r = f;
+                        z.istate.blocks.reset(z, z.istate.was);
+                        if(z.istate.nowrap!=0){
+                            z.istate.mode=DONE;
+                            break;
+                        }
+                        z.istate.mode=CHECK4;
+                        goto case CHECK4;
+                    case CHECK4:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        z.istate.need=((z.next_in[z.next_in_index++]&0xff)<<24)&0xff000000L;
+                        z.istate.mode=CHECK3;
+                        goto case CHECK3;
+                    case CHECK3:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        z.istate.need+=((z.next_in[z.next_in_index++]&0xff)<<16)&0xff0000L;
+                        z.istate.mode = CHECK2;
+                        goto case CHECK2;
+                    case CHECK2:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        z.istate.need+=((z.next_in[z.next_in_index++]&0xff)<<8)&0xff00L;
+                        z.istate.mode = CHECK1;
+                        goto case CHECK1;
+                    case CHECK1:
+
+                        if(z.avail_in==0)return r;r=f;
+
+                        z.avail_in--; z.total_in++;
+                        z.istate.need+=(z.next_in[z.next_in_index++]&0xffL);
+
+                        if(((int)(z.istate.was[0])) != ((int)(z.istate.need))){
+                            z.istate.mode = BAD;
+                            z.msg = "incorrect data check";
+                            z.istate.marker = 5;       // can't try inflateSync
+                            break;
+                        }
+
+                        z.istate.mode = DONE;
+                        goto case DONE;
+                    case DONE:
+                        return Z_STREAM_END;
+                    case BAD:
+                        return Z_DATA_ERROR;
+                    default:
+                        return Z_STREAM_ERROR;
+                }
+            }
+        }
+
+
+        internal int inflateSetDictionary(ZStream z, byte[] dictionary, int dictLength){
+            int index=0;
+            int length = dictLength;
+            if(z==null || z.istate == null|| z.istate.mode != DICT0)
+                return Z_STREAM_ERROR;
+
+            if(z._adler.adler32(1L, dictionary, 0, dictLength)!=z.adler){
+                return Z_DATA_ERROR;
+            }
+
+            z.adler = z._adler.adler32(0, null, 0, 0);
+
+            if(length >= (1<<z.istate.wbits)){
+                length = (1<<z.istate.wbits)-1;
+                index=dictLength - length;
+            }
+            z.istate.blocks.set_dictionary(dictionary, index, length);
+            z.istate.mode = BLOCKS;
+            return Z_OK;
+        }
+
+        private static readonly byte[] mark = {(byte)0, (byte)0, (byte)0xff, (byte)0xff};
+
+        internal int inflateSync(ZStream z){
+            int n;       // number of bytes to look at
+            int p;       // pointer to bytes
+            int m;       // number of marker bytes found in a row
+            long r, w;   // temporaries to save total_in and total_out
+
+            // set up
+            if(z == null || z.istate == null)
+                return Z_STREAM_ERROR;
+            if(z.istate.mode != BAD){
+                z.istate.mode = BAD;
+                z.istate.marker = 0;
+            }
+            if((n=z.avail_in)==0)
+                return Z_BUF_ERROR;
+            p=z.next_in_index;
+            m=z.istate.marker;
+
+            // search
+            while (n!=0 && m < 4){
+                if(z.next_in[p] == mark[m]){
+                    m++;
+                }
+                else if(z.next_in[p]!=0){
+                    m = 0;
+                }
+                else{
+                    m = 4 - m;
+                }
+                p++; n--;
+            }
+
+            // restore
+            z.total_in += p-z.next_in_index;
+            z.next_in_index = p;
+            z.avail_in = n;
+            z.istate.marker = m;
+
+            // return no joy or set up to restart on a new block
+            if(m != 4){
+                return Z_DATA_ERROR;
+            }
+            r=z.total_in;  w=z.total_out;
+            inflateReset(z);
+            z.total_in=r;  z.total_out = w;
+            z.istate.mode = BLOCKS;
+            return Z_OK;
+        }
+
+        // Returns true if inflate is currently at the end of a block generated
+        // by Z_SYNC_FLUSH or Z_FULL_FLUSH. This function is used by one PPP
+        // implementation to provide an additional safety check. PPP uses Z_SYNC_FLUSH
+        // but removes the length bytes of the resulting empty stored block. When
+        // decompressing, PPP checks that at the end of input packet, inflate is
+        // waiting for these length bytes.
+        internal int inflateSyncPoint(ZStream z){
+            if(z == null || z.istate == null || z.istate.blocks == null)
+                return Z_STREAM_ERROR;
+            return z.istate.blocks.sync_point();
+        }
+    }
+}
\ No newline at end of file
diff --git a/Crypto/src/util/zlib/JZlib.cs b/Crypto/src/util/zlib/JZlib.cs
new file mode 100644
index 000000000..4f2cfdaa9
--- /dev/null
+++ b/Crypto/src/util/zlib/JZlib.cs
@@ -0,0 +1,73 @@
+using System;
+/*
+ * $Id: JZlib.cs,v 1.3 2011-02-15 05:46:04 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    public sealed class JZlib{
+        private const String _version="1.0.7";
+        public static String version()
+		{
+			return _version;
+		}
+
+        // compression levels
+        public const int Z_NO_COMPRESSION=0;
+        public const int Z_BEST_SPEED=1;
+        public const int Z_BEST_COMPRESSION=9;
+        public const int Z_DEFAULT_COMPRESSION=-1;
+
+        // compression strategy
+        public const int Z_FILTERED=1;
+        public const int Z_HUFFMAN_ONLY=2;
+        public const int Z_DEFAULT_STRATEGY=0;
+
+        public const int Z_NO_FLUSH=0;
+        public const int Z_PARTIAL_FLUSH=1;
+        public const int Z_SYNC_FLUSH=2;
+        public const int Z_FULL_FLUSH=3;
+        public const int Z_FINISH=4;
+
+        public const int Z_OK=0;
+        public const int Z_STREAM_END=1;
+        public const int Z_NEED_DICT=2;
+        public const int Z_ERRNO=-1;
+        public const int Z_STREAM_ERROR=-2;
+        public const int Z_DATA_ERROR=-3;
+        public const int Z_MEM_ERROR=-4;
+        public const int Z_BUF_ERROR=-5;
+        public const int Z_VERSION_ERROR=-6;
+    }
+}
\ No newline at end of file
diff --git a/Crypto/src/util/zlib/StaticTree.cs b/Crypto/src/util/zlib/StaticTree.cs
new file mode 100644
index 000000000..097735db5
--- /dev/null
+++ b/Crypto/src/util/zlib/StaticTree.cs
@@ -0,0 +1,152 @@
+using System;
+/*
+ * $Id: StaticTree.cs,v 1.2 2008-05-10 09:35:40 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    internal sealed class StaticTree{
+        private const int MAX_BITS=15;
+
+        private const int BL_CODES=19;
+        private const int D_CODES=30;
+        private const int LITERALS=256;
+        private const int LENGTH_CODES=29;
+        private const int L_CODES=(LITERALS+1+LENGTH_CODES);
+
+        // Bit length codes must not exceed MAX_BL_BITS bits
+        internal const int MAX_BL_BITS=7; 
+
+        internal static readonly short[] static_ltree = {
+                                                   12,  8, 140,  8,  76,  8, 204,  8,  44,  8,
+                                                   172,  8, 108,  8, 236,  8,  28,  8, 156,  8,
+                                                   92,  8, 220,  8,  60,  8, 188,  8, 124,  8,
+                                                   252,  8,   2,  8, 130,  8,  66,  8, 194,  8,
+                                                   34,  8, 162,  8,  98,  8, 226,  8,  18,  8,
+                                                   146,  8,  82,  8, 210,  8,  50,  8, 178,  8,
+                                                   114,  8, 242,  8,  10,  8, 138,  8,  74,  8,
+                                                   202,  8,  42,  8, 170,  8, 106,  8, 234,  8,
+                                                   26,  8, 154,  8,  90,  8, 218,  8,  58,  8,
+                                                   186,  8, 122,  8, 250,  8,   6,  8, 134,  8,
+                                                   70,  8, 198,  8,  38,  8, 166,  8, 102,  8,
+                                                   230,  8,  22,  8, 150,  8,  86,  8, 214,  8,
+                                                   54,  8, 182,  8, 118,  8, 246,  8,  14,  8,
+                                                   142,  8,  78,  8, 206,  8,  46,  8, 174,  8,
+                                                   110,  8, 238,  8,  30,  8, 158,  8,  94,  8,
+                                                   222,  8,  62,  8, 190,  8, 126,  8, 254,  8,
+                                                   1,  8, 129,  8,  65,  8, 193,  8,  33,  8,
+                                                   161,  8,  97,  8, 225,  8,  17,  8, 145,  8,
+                                                   81,  8, 209,  8,  49,  8, 177,  8, 113,  8,
+                                                   241,  8,   9,  8, 137,  8,  73,  8, 201,  8,
+                                                   41,  8, 169,  8, 105,  8, 233,  8,  25,  8,
+                                                   153,  8,  89,  8, 217,  8,  57,  8, 185,  8,
+                                                   121,  8, 249,  8,   5,  8, 133,  8,  69,  8,
+                                                   197,  8,  37,  8, 165,  8, 101,  8, 229,  8,
+                                                   21,  8, 149,  8,  85,  8, 213,  8,  53,  8,
+                                                   181,  8, 117,  8, 245,  8,  13,  8, 141,  8,
+                                                   77,  8, 205,  8,  45,  8, 173,  8, 109,  8,
+                                                   237,  8,  29,  8, 157,  8,  93,  8, 221,  8,
+                                                   61,  8, 189,  8, 125,  8, 253,  8,  19,  9,
+                                                   275,  9, 147,  9, 403,  9,  83,  9, 339,  9,
+                                                   211,  9, 467,  9,  51,  9, 307,  9, 179,  9,
+                                                   435,  9, 115,  9, 371,  9, 243,  9, 499,  9,
+                                                   11,  9, 267,  9, 139,  9, 395,  9,  75,  9,
+                                                   331,  9, 203,  9, 459,  9,  43,  9, 299,  9,
+                                                   171,  9, 427,  9, 107,  9, 363,  9, 235,  9,
+                                                   491,  9,  27,  9, 283,  9, 155,  9, 411,  9,
+                                                   91,  9, 347,  9, 219,  9, 475,  9,  59,  9,
+                                                   315,  9, 187,  9, 443,  9, 123,  9, 379,  9,
+                                                   251,  9, 507,  9,   7,  9, 263,  9, 135,  9,
+                                                   391,  9,  71,  9, 327,  9, 199,  9, 455,  9,
+                                                   39,  9, 295,  9, 167,  9, 423,  9, 103,  9,
+                                                   359,  9, 231,  9, 487,  9,  23,  9, 279,  9,
+                                                   151,  9, 407,  9,  87,  9, 343,  9, 215,  9,
+                                                   471,  9,  55,  9, 311,  9, 183,  9, 439,  9,
+                                                   119,  9, 375,  9, 247,  9, 503,  9,  15,  9,
+                                                   271,  9, 143,  9, 399,  9,  79,  9, 335,  9,
+                                                   207,  9, 463,  9,  47,  9, 303,  9, 175,  9,
+                                                   431,  9, 111,  9, 367,  9, 239,  9, 495,  9,
+                                                   31,  9, 287,  9, 159,  9, 415,  9,  95,  9,
+                                                   351,  9, 223,  9, 479,  9,  63,  9, 319,  9,
+                                                   191,  9, 447,  9, 127,  9, 383,  9, 255,  9,
+                                                   511,  9,   0,  7,  64,  7,  32,  7,  96,  7,
+                                                   16,  7,  80,  7,  48,  7, 112,  7,   8,  7,
+                                                   72,  7,  40,  7, 104,  7,  24,  7,  88,  7,
+                                                   56,  7, 120,  7,   4,  7,  68,  7,  36,  7,
+                                                   100,  7,  20,  7,  84,  7,  52,  7, 116,  7,
+                                                   3,  8, 131,  8,  67,  8, 195,  8,  35,  8,
+                                                   163,  8,  99,  8, 227,  8
+                                               };
+
+        internal static readonly short[] static_dtree = {
+                                                   0, 5, 16, 5,  8, 5, 24, 5,  4, 5,
+                                                   20, 5, 12, 5, 28, 5,  2, 5, 18, 5,
+                                                   10, 5, 26, 5,  6, 5, 22, 5, 14, 5,
+                                                   30, 5,  1, 5, 17, 5,  9, 5, 25, 5,
+                                                   5, 5, 21, 5, 13, 5, 29, 5,  3, 5,
+                                                   19, 5, 11, 5, 27, 5,  7, 5, 23, 5
+                                               };
+
+        internal static readonly StaticTree static_l_desc =
+            new StaticTree(static_ltree, Tree.extra_lbits,
+            LITERALS+1, L_CODES, MAX_BITS);
+
+        internal static readonly StaticTree static_d_desc =
+            new StaticTree(static_dtree, Tree.extra_dbits,
+            0,  D_CODES, MAX_BITS);
+
+        internal static readonly StaticTree static_bl_desc =
+            new StaticTree(null, Tree.extra_blbits,
+            0, BL_CODES, MAX_BL_BITS);
+
+        internal short[] static_tree;     // static tree or null
+        internal int[] extra_bits;        // extra bits for each code or null
+        internal int extra_base;          // base index for extra_bits
+        internal int elems;               // max number of elements in the tree
+        internal int max_length;          // max bit length for the codes
+
+        internal StaticTree(short[] static_tree,
+            int[] extra_bits,
+            int extra_base,
+            int elems,
+            int max_length
+            ){
+            this.static_tree=static_tree;
+            this.extra_bits=extra_bits;
+            this.extra_base=extra_base;
+            this.elems=elems;
+            this.max_length=max_length;
+        }
+    }
+}
\ No newline at end of file
diff --git a/Crypto/src/util/zlib/Tree.cs b/Crypto/src/util/zlib/Tree.cs
new file mode 100644
index 000000000..29d0a30d4
--- /dev/null
+++ b/Crypto/src/util/zlib/Tree.cs
@@ -0,0 +1,367 @@
+using System;
+/*
+ * $Id: Tree.cs,v 1.2 2008-05-10 09:35:40 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    internal sealed class Tree{
+        private const int MAX_BITS=15;
+        private const int BL_CODES=19;
+        private const int D_CODES=30;
+        private const int LITERALS=256;
+        private const int LENGTH_CODES=29;
+        private const int L_CODES=(LITERALS+1+LENGTH_CODES);
+        private const int HEAP_SIZE=(2*L_CODES+1);
+
+        // Bit length codes must not exceed MAX_BL_BITS bits
+        internal const int MAX_BL_BITS=7; 
+
+        // end of block literal code
+        internal const int END_BLOCK=256; 
+
+        // repeat previous bit length 3-6 times (2 bits of repeat count)
+        internal const int REP_3_6=16; 
+
+        // repeat a zero length 3-10 times  (3 bits of repeat count)
+        internal const int REPZ_3_10=17; 
+
+        // repeat a zero length 11-138 times  (7 bits of repeat count)
+        internal const int REPZ_11_138=18; 
+
+        // extra bits for each length code
+        internal static readonly int[] extra_lbits={
+                                             0,0,0,0,0,0,0,0,1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4,5,5,5,5,0
+                                         };
+
+        // extra bits for each distance code
+        internal static readonly int[] extra_dbits={
+                                             0,0,0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,12,12,13,13
+                                         };
+
+        // extra bits for each bit length code
+        internal static readonly int[] extra_blbits={
+                                              0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,2,3,7
+                                          };
+
+        internal static readonly byte[] bl_order={
+                                            16,17,18,0,8,7,9,6,10,5,11,4,12,3,13,2,14,1,15};
+
+
+        // The lengths of the bit length codes are sent in order of decreasing
+        // probability, to avoid transmitting the lengths for unused bit
+        // length codes.
+
+        internal const int Buf_size=8*2;
+
+        // see definition of array dist_code below
+        internal const int DIST_CODE_LEN=512;
+
+        internal static readonly byte[] _dist_code = {
+                                                0,  1,  2,  3,  4,  4,  5,  5,  6,  6,  6,  6,  7,  7,  7,  7,  8,  8,  8,  8,
+                                                8,  8,  8,  8,  9,  9,  9,  9,  9,  9,  9,  9, 10, 10, 10, 10, 10, 10, 10, 10,
+                                                10, 10, 10, 10, 10, 10, 10, 10, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,
+                                                11, 11, 11, 11, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+                                                12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 13, 13, 13, 13,
+                                                13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+                                                13, 13, 13, 13, 13, 13, 13, 13, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+                                                14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+                                                14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+                                                14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 15, 15, 15, 15, 15, 15, 15, 15,
+                                                15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+                                                15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+                                                15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,  0,  0, 16, 17,
+                                                18, 18, 19, 19, 20, 20, 20, 20, 21, 21, 21, 21, 22, 22, 22, 22, 22, 22, 22, 22,
+                                                23, 23, 23, 23, 23, 23, 23, 23, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+                                                24, 24, 24, 24, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+                                                26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+                                                26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 27, 27, 27, 27, 27, 27, 27, 27,
+                                                27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,
+                                                27, 27, 27, 27, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28,
+                                                28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28,
+                                                28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28,
+                                                28, 28, 28, 28, 28, 28, 28, 28, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29,
+                                                29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29,
+                                                29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29,
+                                                29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29
+                                            };
+
+        internal static readonly byte[] _length_code={
+                                                0,  1,  2,  3,  4,  5,  6,  7,  8,  8,  9,  9, 10, 10, 11, 11, 12, 12, 12, 12,
+                                                13, 13, 13, 13, 14, 14, 14, 14, 15, 15, 15, 15, 16, 16, 16, 16, 16, 16, 16, 16,
+                                                17, 17, 17, 17, 17, 17, 17, 17, 18, 18, 18, 18, 18, 18, 18, 18, 19, 19, 19, 19,
+                                                19, 19, 19, 19, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20,
+                                                21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 22, 22, 22, 22,
+                                                22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 23, 23, 23, 23, 23, 23, 23, 23,
+                                                23, 23, 23, 23, 23, 23, 23, 23, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+                                                24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+                                                25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+                                                25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 26, 26, 26, 26, 26, 26, 26, 26,
+                                                26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+                                                26, 26, 26, 26, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,
+                                                27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 28
+                                            };
+
+        internal static readonly int[] base_length = {
+                                               0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14, 16, 20, 24, 28, 32, 40, 48, 56,
+                                               64, 80, 96, 112, 128, 160, 192, 224, 0
+                                           };
+
+        internal static readonly int[] base_dist = {
+                                             0,   1,      2,     3,     4,    6,     8,    12,    16,     24,
+                                             32,  48,     64,    96,   128,  192,   256,   384,   512,    768,
+                                             1024, 1536,  2048,  3072,  4096,  6144,  8192, 12288, 16384, 24576
+                                         };
+
+        // Mapping from a distance to a distance code. dist is the distance - 1 and
+        // must not have side effects. _dist_code[256] and _dist_code[257] are never
+        // used.
+        internal static int d_code(int dist){
+            return ((dist) < 256 ? _dist_code[dist] : _dist_code[256+((dist)>>7)]);
+        }
+
+        internal short[] dyn_tree;      // the dynamic tree
+        internal int     max_code;      // largest code with non zero frequency
+        internal StaticTree stat_desc;  // the corresponding static tree
+
+        // Compute the optimal bit lengths for a tree and update the total bit length
+        // for the current block.
+        // IN assertion: the fields freq and dad are set, heap[heap_max] and
+        //    above are the tree nodes sorted by increasing frequency.
+        // OUT assertions: the field len is set to the optimal bit length, the
+        //     array bl_count contains the frequencies for each bit length.
+        //     The length opt_len is updated; static_len is also updated if stree is
+        //     not null.
+        internal void gen_bitlen(Deflate s){
+            short[] tree = dyn_tree;
+            short[] stree = stat_desc.static_tree;
+            int[] extra = stat_desc.extra_bits;
+            int based = stat_desc.extra_base;
+            int max_length = stat_desc.max_length;
+            int h;              // heap index
+            int n, m;           // iterate over the tree elements
+            int bits;           // bit length
+            int xbits;          // extra bits
+            short f;            // frequency
+            int overflow = 0;   // number of elements with bit length too large
+
+            for (bits = 0; bits <= MAX_BITS; bits++) s.bl_count[bits] = 0;
+
+            // In a first pass, compute the optimal bit lengths (which may
+            // overflow in the case of the bit length tree).
+            tree[s.heap[s.heap_max]*2+1] = 0; // root of the heap
+
+            for(h=s.heap_max+1; h<HEAP_SIZE; h++){
+                n = s.heap[h];
+                bits = tree[tree[n*2+1]*2+1] + 1;
+                if (bits > max_length){ bits = max_length; overflow++; }
+                tree[n*2+1] = (short)bits;
+                // We overwrite tree[n*2+1] which is no longer needed
+
+                if (n > max_code) continue;  // not a leaf node
+
+                s.bl_count[bits]++;
+                xbits = 0;
+                if (n >= based) xbits = extra[n-based];
+                f = tree[n*2];
+                s.opt_len += f * (bits + xbits);
+                if (stree!=null) s.static_len += f * (stree[n*2+1] + xbits);
+            }
+            if (overflow == 0) return;
+
+            // This happens for example on obj2 and pic of the Calgary corpus
+            // Find the first bit length which could increase:
+            do {
+                bits = max_length-1;
+            while(s.bl_count[bits]==0) bits--;
+                s.bl_count[bits]--;      // move one leaf down the tree
+                s.bl_count[bits+1]+=2;   // move one overflow item as its brother
+                s.bl_count[max_length]--;
+                // The brother of the overflow item also moves one step up,
+                // but this does not affect bl_count[max_length]
+                overflow -= 2;
+            }
+            while (overflow > 0);
+
+            for (bits = max_length; bits != 0; bits--) {
+                n = s.bl_count[bits];
+                while (n != 0) {
+                    m = s.heap[--h];
+                    if (m > max_code) continue;
+                    if (tree[m*2+1] != bits) {
+                        s.opt_len += (int)(((long)bits - (long)tree[m*2+1])*(long)tree[m*2]);
+                        tree[m*2+1] = (short)bits;
+                    }
+                    n--;
+                }
+            }
+        }
+
+        // Construct one Huffman tree and assigns the code bit strings and lengths.
+        // Update the total bit length for the current block.
+        // IN assertion: the field freq is set for all tree elements.
+        // OUT assertions: the fields len and code are set to the optimal bit length
+        //     and corresponding code. The length opt_len is updated; static_len is
+        //     also updated if stree is not null. The field max_code is set.
+        internal void build_tree(Deflate s){
+            short[] tree=dyn_tree;
+            short[] stree=stat_desc.static_tree;
+            int elems=stat_desc.elems;
+            int n, m;          // iterate over heap elements
+            int max_code=-1;   // largest code with non zero frequency
+            int node;          // new node being created
+
+            // Construct the initial heap, with least frequent element in
+            // heap[1]. The sons of heap[n] are heap[2*n] and heap[2*n+1].
+            // heap[0] is not used.
+            s.heap_len = 0;
+            s.heap_max = HEAP_SIZE;
+
+            for(n=0; n<elems; n++) {
+                if(tree[n*2] != 0) {
+                    s.heap[++s.heap_len] = max_code = n;
+                    s.depth[n] = 0;
+                }
+                else{
+                    tree[n*2+1] = 0;
+                }
+            }
+
+            // The pkzip format requires that at least one distance code exists,
+            // and that at least one bit should be sent even if there is only one
+            // possible code. So to avoid special checks later on we force at least
+            // two codes of non zero frequency.
+            while (s.heap_len < 2) {
+                node = s.heap[++s.heap_len] = (max_code < 2 ? ++max_code : 0);
+                tree[node*2] = 1;
+                s.depth[node] = 0;
+                s.opt_len--; if (stree!=null) s.static_len -= stree[node*2+1];
+                // node is 0 or 1 so it does not have extra bits
+            }
+            this.max_code = max_code;
+
+            // The elements heap[heap_len/2+1 .. heap_len] are leaves of the tree,
+            // establish sub-heaps of increasing lengths:
+
+            for(n=s.heap_len/2;n>=1; n--)
+                s.pqdownheap(tree, n);
+
+            // Construct the Huffman tree by repeatedly combining the least two
+            // frequent nodes.
+
+            node=elems;                 // next internal node of the tree
+            do{
+                // n = node of least frequency
+                n=s.heap[1];
+                s.heap[1]=s.heap[s.heap_len--];
+                s.pqdownheap(tree, 1);
+                m=s.heap[1];                // m = node of next least frequency
+
+                s.heap[--s.heap_max] = n; // keep the nodes sorted by frequency
+                s.heap[--s.heap_max] = m;
+
+                // Create a new node father of n and m
+                tree[node*2] = (short)(tree[n*2] + tree[m*2]);
+                s.depth[node] = (byte)(System.Math.Max(s.depth[n],s.depth[m])+1);
+                tree[n*2+1] = tree[m*2+1] = (short)node;
+
+                // and insert the new node in the heap
+                s.heap[1] = node++;
+                s.pqdownheap(tree, 1);
+            }
+            while(s.heap_len>=2);
+
+            s.heap[--s.heap_max] = s.heap[1];
+
+            // At this point, the fields freq and dad are set. We can now
+            // generate the bit lengths.
+
+            gen_bitlen(s);
+
+            // The field len is now set, we can generate the bit codes
+            gen_codes(tree, max_code, s.bl_count);
+        }
+
+        // Generate the codes for a given tree and bit counts (which need not be
+        // optimal).
+        // IN assertion: the array bl_count contains the bit length statistics for
+        // the given tree and the field len is set for all tree elements.
+        // OUT assertion: the field code is set for all tree elements of non
+        //     zero code length.
+        internal static void gen_codes(short[] tree, // the tree to decorate
+            int max_code, // largest code with non zero frequency
+            short[] bl_count // number of codes at each bit length
+            ){
+            short[] next_code=new short[MAX_BITS+1]; // next code value for each bit length
+            short code = 0;            // running code value
+            int bits;                  // bit index
+            int n;                     // code index
+
+            // The distribution counts are first used to generate the code values
+            // without bit reversal.
+            for (bits = 1; bits <= MAX_BITS; bits++) {
+                next_code[bits] = code = (short)((code + bl_count[bits-1]) << 1);
+            }
+
+            // Check that the bit counts in bl_count are consistent. The last code
+            // must be all ones.
+            //Assert (code + bl_count[MAX_BITS]-1 == (1<<MAX_BITS)-1,
+            //        "inconsistent bit counts");
+            //Tracev((stderr,"\ngen_codes: max_code %d ", max_code));
+
+            for (n = 0;  n <= max_code; n++) {
+                int len = tree[n*2+1];
+                if (len == 0) continue;
+                // Now reverse the bits
+                tree[n*2] = (short)(bi_reverse(next_code[len]++, len));
+            }
+        }
+
+        // Reverse the first len bits of a code, using straightforward code (a faster
+        // method would use a table)
+        // IN assertion: 1 <= len <= 15
+        internal static int bi_reverse(int code, // the value to invert
+            int len   // its bit length
+            ){
+            int res = 0;
+            do{
+                res|=code&1;
+                code>>=1;
+                res<<=1;
+            } 
+            while(--len>0);
+            return res>>1;
+        }
+    }
+}
\ No newline at end of file
diff --git a/Crypto/src/util/zlib/ZDeflaterOutputStream.cs b/Crypto/src/util/zlib/ZDeflaterOutputStream.cs
new file mode 100644
index 000000000..6dea74223
--- /dev/null
+++ b/Crypto/src/util/zlib/ZDeflaterOutputStream.cs
@@ -0,0 +1,155 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+    /// <summary>
+    /// Summary description for DeflaterOutputStream.
+    /// </summary>
+    [Obsolete("Use 'ZOutputStream' instead")]
+    public class ZDeflaterOutputStream : Stream {
+        protected ZStream z=new ZStream();
+        protected int flushLevel=JZlib.Z_NO_FLUSH;
+        private const int BUFSIZE = 4192;
+        protected byte[] buf=new byte[BUFSIZE];
+        private byte[] buf1=new byte[1];
+
+        protected Stream outp;
+
+        public ZDeflaterOutputStream(Stream outp) : this(outp, 6, false) {
+        }
+    
+        public ZDeflaterOutputStream(Stream outp, int level) : this(outp, level, false) {
+        }
+    
+        public ZDeflaterOutputStream(Stream outp, int level, bool nowrap) {
+            this.outp=outp;
+            z.deflateInit(level, nowrap);
+        }
+    
+    
+        public override bool CanRead {
+            get {
+                // TODO:  Add DeflaterOutputStream.CanRead getter implementation
+                return false;
+            }
+        }
+    
+        public override bool CanSeek {
+            get {
+                // TODO:  Add DeflaterOutputStream.CanSeek getter implementation
+                return false;
+            }
+        }
+    
+        public override bool CanWrite {
+            get {
+                // TODO:  Add DeflaterOutputStream.CanWrite getter implementation
+                return true;
+            }
+        }
+    
+        public override long Length {
+            get {
+                // TODO:  Add DeflaterOutputStream.Length getter implementation
+                return 0;
+            }
+        }
+    
+        public override long Position {
+            get {
+                // TODO:  Add DeflaterOutputStream.Position getter implementation
+                return 0;
+            }
+            set {
+                // TODO:  Add DeflaterOutputStream.Position setter implementation
+            }
+        }
+    
+        public override void Write(byte[] b, int off, int len) {
+            if(len==0)
+                return;
+            int err;
+            z.next_in=b;
+            z.next_in_index=off;
+            z.avail_in=len;
+            do{
+                z.next_out=buf;
+                z.next_out_index=0;
+                z.avail_out=BUFSIZE;
+                err=z.deflate(flushLevel);
+                if(err!=JZlib.Z_OK)
+                    throw new IOException("deflating: "+z.msg);
+				if (z.avail_out < BUFSIZE)
+				{
+					outp.Write(buf, 0, BUFSIZE-z.avail_out);
+				}
+            }
+            while(z.avail_in>0 || z.avail_out==0);
+        }
+    
+        public override long Seek(long offset, SeekOrigin origin) {
+            // TODO:  Add DeflaterOutputStream.Seek implementation
+            return 0;
+        }
+    
+        public override void SetLength(long value) {
+            // TODO:  Add DeflaterOutputStream.SetLength implementation
+
+        }
+    
+        public override int Read(byte[] buffer, int offset, int count) {
+            // TODO:  Add DeflaterOutputStream.Read implementation
+            return 0;
+        }
+    
+        public override void Flush() {
+            outp.Flush();
+        }
+    
+        public override void WriteByte(byte b) {
+            buf1[0]=(byte)b;
+            Write(buf1, 0, 1);
+        }
+
+        public void Finish() {
+            int err;
+            do{
+                z.next_out=buf;
+                z.next_out_index=0;
+                z.avail_out=BUFSIZE;
+                err=z.deflate(JZlib.Z_FINISH);
+                if(err!=JZlib.Z_STREAM_END && err != JZlib.Z_OK)
+                    throw new IOException("deflating: "+z.msg);
+                if(BUFSIZE-z.avail_out>0){
+                    outp.Write(buf, 0, BUFSIZE-z.avail_out);
+                }
+            }
+            while(z.avail_in>0 || z.avail_out==0);
+            Flush();
+        }
+
+        public void End() {
+            if(z==null)
+                return;
+            z.deflateEnd();
+            z.free();
+            z=null;
+        }
+
+        protected override void Dispose(bool disposing)
+        {
+            if (disposing)
+            {
+                try{
+                    try { Finish(); }
+                    catch (IOException) { }
+                }
+                finally{
+                    End();
+                    outp.Dispose();
+                    outp = null;
+                }
+            }
+        }
+    }
+}
diff --git a/Crypto/src/util/zlib/ZInflaterInputStream.cs b/Crypto/src/util/zlib/ZInflaterInputStream.cs
new file mode 100644
index 000000000..d2d70ebd1
--- /dev/null
+++ b/Crypto/src/util/zlib/ZInflaterInputStream.cs
@@ -0,0 +1,131 @@
+using System;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+    /// <summary>
+    /// Summary description for DeflaterOutputStream.
+    /// </summary>
+    [Obsolete("Use 'ZInputStream' instead")]
+    public class ZInflaterInputStream : Stream {
+        protected ZStream z=new ZStream();
+        protected int flushLevel=JZlib.Z_NO_FLUSH;
+        private const int BUFSIZE = 4192;
+        protected byte[] buf=new byte[BUFSIZE];
+        private byte[] buf1=new byte[1];
+
+        protected Stream inp=null;
+        private bool nomoreinput=false;
+
+        public ZInflaterInputStream(Stream inp) : this(inp, false) {
+        }
+    
+        public ZInflaterInputStream(Stream inp, bool nowrap) {
+            this.inp=inp;
+            z.inflateInit(nowrap);
+            z.next_in=buf;
+            z.next_in_index=0;
+            z.avail_in=0;
+        }
+    
+        public override bool CanRead {
+            get {
+                // TODO:  Add DeflaterOutputStream.CanRead getter implementation
+                return true;
+            }
+        }
+    
+        public override bool CanSeek {
+            get {
+                // TODO:  Add DeflaterOutputStream.CanSeek getter implementation
+                return false;
+            }
+        }
+    
+        public override bool CanWrite {
+            get {
+                // TODO:  Add DeflaterOutputStream.CanWrite getter implementation
+                return false;
+            }
+        }
+    
+        public override long Length {
+            get {
+                // TODO:  Add DeflaterOutputStream.Length getter implementation
+                return 0;
+            }
+        }
+    
+        public override long Position {
+            get {
+                // TODO:  Add DeflaterOutputStream.Position getter implementation
+                return 0;
+            }
+            set {
+                // TODO:  Add DeflaterOutputStream.Position setter implementation
+            }
+        }
+    
+        public override void Write(byte[] b, int off, int len) {
+        }
+    
+        public override long Seek(long offset, SeekOrigin origin) {
+            // TODO:  Add DeflaterOutputStream.Seek implementation
+            return 0;
+        }
+    
+        public override void SetLength(long value) {
+            // TODO:  Add DeflaterOutputStream.SetLength implementation
+
+        }
+    
+        public override int Read(byte[] b, int off, int len) {
+            if(len==0)
+                return(0);
+            int err;
+            z.next_out=b;
+            z.next_out_index=off;
+            z.avail_out=len;
+            do {
+                if((z.avail_in==0)&&(!nomoreinput)) { // if buffer is empty and more input is avaiable, refill it
+                    z.next_in_index=0;
+                    z.avail_in=inp.Read(buf, 0, BUFSIZE);//(BUFSIZE<z.avail_out ? BUFSIZE : z.avail_out));
+                    if(z.avail_in<=0) {
+                        z.avail_in=0;
+                        nomoreinput=true;
+                    }
+                }
+                err=z.inflate(flushLevel);
+                if(nomoreinput&&(err==JZlib.Z_BUF_ERROR))
+                    return(0);
+                if(err!=JZlib.Z_OK && err!=JZlib.Z_STREAM_END)
+                    throw new IOException("inflating: "+z.msg);
+                if((nomoreinput||err==JZlib.Z_STREAM_END)&&(z.avail_out==len))
+                    return(0);
+            } 
+            while(z.avail_out==len&&err==JZlib.Z_OK);
+            //System.err.print("("+(len-z.avail_out)+")");
+            return(len-z.avail_out);
+        }
+    
+        public override void Flush() {
+            inp.Flush();
+        }
+    
+        public override void WriteByte(byte b) {
+        }
+
+        protected override void Dispose(bool disposing)
+        {
+            if (disposing)
+            {
+                inp.Dispose();
+            }
+        }
+    
+        public override int ReadByte() {
+            if(Read(buf1, 0, 1)<=0)
+                return -1;
+            return(buf1[0]&0xFF);
+        }
+    }
+}
diff --git a/Crypto/src/util/zlib/ZInputStream.cs b/Crypto/src/util/zlib/ZInputStream.cs
new file mode 100644
index 000000000..179c133bf
--- /dev/null
+++ b/Crypto/src/util/zlib/ZInputStream.cs
@@ -0,0 +1,196 @@
+/*
+Copyright (c) 2001 Lapo Luchini.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHORS
+OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+/* This file is a port of jzlib v1.0.7, com.jcraft.jzlib.ZInputStream.java
+ */
+
+using System;
+using System.Diagnostics;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.Zlib
+{
+	public class ZInputStream
+		: Stream
+	{
+		private const int BufferSize = 512;
+
+		protected ZStream z = new ZStream();
+		protected int flushLevel = JZlib.Z_NO_FLUSH;
+		// TODO Allow custom buf
+		protected byte[] buf = new byte[BufferSize];
+		protected byte[] buf1 = new byte[1];
+		protected bool compress;
+
+		protected Stream input;
+		protected bool closed;
+
+		private bool nomoreinput = false;
+
+		public ZInputStream(Stream input)
+			: this(input, false)
+		{
+		}
+
+		public ZInputStream(Stream input, bool nowrap)
+		{
+			Debug.Assert(input.CanRead);
+
+			this.input = input;
+			this.z.inflateInit(nowrap);
+			this.compress = false;
+			this.z.next_in = buf;
+			this.z.next_in_index = 0;
+			this.z.avail_in = 0;
+		}
+
+		public ZInputStream(Stream input, int level)
+		{
+			Debug.Assert(input.CanRead);
+			
+			this.input = input;
+			this.z.deflateInit(level);
+			this.compress = true;
+			this.z.next_in = buf;
+			this.z.next_in_index = 0;
+			this.z.avail_in = 0;
+		}
+
+		/*public int available() throws IOException {
+		return inf.finished() ? 0 : 1;
+		}*/
+
+		public sealed override bool CanRead { get { return !closed; } }
+		public sealed override bool CanSeek { get { return false; } }
+		public sealed override bool CanWrite { get { return false; } }
+
+        protected override void Dispose(bool disposing)
+        {
+            if (disposing)
+            {
+                if (!closed)
+                {
+                    closed = true;
+                    input.Dispose();
+                }
+            }
+		}
+
+		public sealed override void Flush() {}
+
+		public virtual int FlushMode
+		{
+			get { return flushLevel; }
+			set { this.flushLevel = value; }
+		}
+
+		public sealed override long Length { get { throw new NotSupportedException(); } }
+		public sealed override long Position
+		{
+			get { throw new NotSupportedException(); }
+			set { throw new NotSupportedException(); }
+		}
+
+		public override int Read(byte[]	b, int off, int len)
+		{
+			if (len==0)
+				return 0;
+
+			z.next_out = b;
+			z.next_out_index = off;
+			z.avail_out = len;
+
+			int err;
+			do
+			{
+				if (z.avail_in == 0 && !nomoreinput)
+				{
+					// if buffer is empty and more input is available, refill it
+					z.next_in_index = 0;
+					z.avail_in = input.Read(buf, 0, buf.Length); //(bufsize<z.avail_out ? bufsize : z.avail_out));
+
+					if (z.avail_in <= 0)
+					{
+						z.avail_in = 0;
+						nomoreinput = true;
+					}
+				}
+
+				err = compress
+					?	z.deflate(flushLevel)
+					:	z.inflate(flushLevel);
+
+				if (nomoreinput && err == JZlib.Z_BUF_ERROR)
+					return 0;
+				if (err != JZlib.Z_OK && err != JZlib.Z_STREAM_END)
+					// TODO
+//					throw new ZStreamException((compress ? "de" : "in") + "flating: " + z.msg);
+					throw new IOException((compress ? "de" : "in") + "flating: " + z.msg);
+				if ((nomoreinput || err == JZlib.Z_STREAM_END) && z.avail_out == len)
+					return 0;
+			} 
+			while(z.avail_out == len && err == JZlib.Z_OK);
+			//Console.Error.WriteLine("("+(len-z.avail_out)+")");
+			return len - z.avail_out;
+		}
+
+		public override int ReadByte()
+		{
+			if (Read(buf1, 0, 1) <= 0)
+				return -1;
+			return buf1[0];
+		}
+
+//  public long skip(long n) throws IOException {
+//    int len=512;
+//    if(n<len)
+//      len=(int)n;
+//    byte[] tmp=new byte[len];
+//    return((long)read(tmp));
+//  }
+
+		public sealed override long Seek(long offset, SeekOrigin origin) { throw new NotSupportedException(); }
+		public sealed override void SetLength(long value) { throw new NotSupportedException(); }
+
+		public virtual long TotalIn
+		{
+			get { return z.total_in; }
+		}
+
+		public virtual long TotalOut
+		{
+			get { return z.total_out; }
+		}
+
+		public sealed override void Write(byte[] buffer, int offset, int count) { throw new NotSupportedException(); }
+	}
+}
diff --git a/Crypto/src/util/zlib/ZOutputStream.cs b/Crypto/src/util/zlib/ZOutputStream.cs
new file mode 100644
index 000000000..7ea7cbb33
--- /dev/null
+++ b/Crypto/src/util/zlib/ZOutputStream.cs
@@ -0,0 +1,220 @@
+/*
+Copyright (c) 2001 Lapo Luchini.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHORS
+OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+/* This file is a port of jzlib v1.0.7, com.jcraft.jzlib.ZOutputStream.java
+ */
+
+using System;
+using System.Diagnostics;
+using System.IO;
+
+namespace Org.BouncyCastle.Utilities.Zlib
+{
+	public class ZOutputStream
+		: Stream
+	{
+		private const int BufferSize = 512;
+
+		protected ZStream z = new ZStream();
+		protected int flushLevel = JZlib.Z_NO_FLUSH;
+		// TODO Allow custom buf
+		protected byte[] buf = new byte[BufferSize];
+		protected byte[] buf1 = new byte[1];
+		protected bool compress;
+
+		protected Stream output;
+		protected bool closed;
+
+		public ZOutputStream(Stream output)
+			: base()
+		{
+			Debug.Assert(output.CanWrite);
+
+			this.output = output;
+			this.z.inflateInit();
+			this.compress = false;
+		}
+
+		public ZOutputStream(Stream output, int level)
+			: this(output, level, false)
+		{
+		}
+
+		public ZOutputStream(Stream output, int level, bool nowrap)
+			: base()
+		{
+			Debug.Assert(output.CanWrite);
+
+			this.output = output;
+			this.z.deflateInit(level, nowrap);
+			this.compress = true;
+		}
+
+		public sealed override bool CanRead { get { return false; } }
+        public sealed override bool CanSeek { get { return false; } }
+        public sealed override bool CanWrite { get { return !closed; } }
+
+        protected override void Dispose(bool disposing)
+        {
+            if (disposing)
+            {
+                if (this.closed)
+                    return;
+
+                try
+                {
+                    try
+                    {
+                        Finish();
+                    }
+                    catch (IOException)
+                    {
+                        // Ignore
+                    }
+                }
+                finally
+                {
+                    this.closed = true;
+                    End();
+                    output.Dispose();
+                    output = null;
+                }
+            }
+		}
+
+		public virtual void End()
+		{
+			if (z == null)
+				return;
+			if (compress)
+				z.deflateEnd();
+			else
+				z.inflateEnd();
+			z.free();
+			z = null;
+		}
+
+		public virtual void Finish()
+		{
+			do
+			{
+				z.next_out = buf;
+				z.next_out_index = 0;
+				z.avail_out = buf.Length;
+
+				int err = compress
+					?	z.deflate(JZlib.Z_FINISH)
+					:	z.inflate(JZlib.Z_FINISH);
+
+				if (err != JZlib.Z_STREAM_END && err != JZlib.Z_OK)
+					// TODO
+//					throw new ZStreamException((compress?"de":"in")+"flating: "+z.msg);
+					throw new IOException((compress ? "de" : "in") + "flating: " + z.msg);
+
+				int count = buf.Length - z.avail_out;
+				if (count > 0)
+				{
+					output.Write(buf, 0, count);
+				}
+			}
+			while (z.avail_in > 0 || z.avail_out == 0);
+
+			Flush();
+		}
+
+		public override void Flush()
+		{
+			output.Flush();
+		}
+
+		public virtual int FlushMode
+		{
+			get { return flushLevel; }
+			set { this.flushLevel = value; }
+		}
+
+        public sealed override long Length { get { throw new NotSupportedException(); } }
+        public sealed override long Position
+        {
+            get { throw new NotSupportedException(); }
+            set { throw new NotSupportedException(); }
+        }
+        public sealed override int Read(byte[] buffer, int offset, int count) { throw new NotSupportedException(); }
+        public sealed override long Seek(long offset, SeekOrigin origin) { throw new NotSupportedException(); }
+        public sealed override void SetLength(long value) { throw new NotSupportedException(); }
+
+		public virtual long TotalIn
+		{
+			get { return z.total_in; }
+		}
+
+		public virtual long TotalOut
+		{
+			get { return z.total_out; }
+		}
+
+		public override void Write(byte[] b, int off, int len)
+		{
+			if (len == 0)
+				return;
+
+			z.next_in = b;
+			z.next_in_index = off;
+			z.avail_in = len;
+
+			do
+			{
+				z.next_out = buf;
+				z.next_out_index = 0;
+				z.avail_out = buf.Length;
+
+				int err = compress
+					?	z.deflate(flushLevel)
+					:	z.inflate(flushLevel);
+
+				if (err != JZlib.Z_OK)
+					// TODO
+//					throw new ZStreamException((compress ? "de" : "in") + "flating: " + z.msg);
+					throw new IOException((compress ? "de" : "in") + "flating: " + z.msg);
+
+				output.Write(buf, 0, buf.Length - z.avail_out);
+			}
+			while (z.avail_in > 0 || z.avail_out == 0);
+		}
+
+		public override void WriteByte(byte b)
+		{
+			buf1[0] = b;
+			Write(buf1, 0, 1);
+		}
+	}
+}
diff --git a/Crypto/src/util/zlib/ZStream.cs b/Crypto/src/util/zlib/ZStream.cs
new file mode 100644
index 000000000..7ff961462
--- /dev/null
+++ b/Crypto/src/util/zlib/ZStream.cs
@@ -0,0 +1,214 @@
+using System;
+/*
+ * $Id: ZStream.cs,v 1.1 2006-07-31 13:59:26 bouncy Exp $
+ *
+Copyright (c) 2000,2001,2002,2003 ymnk, JCraft,Inc. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+  1. Redistributions of source code must retain the above copyright notice,
+     this list of conditions and the following disclaimer.
+
+  2. Redistributions in binary form must reproduce the above copyright 
+     notice, this list of conditions and the following disclaimer in 
+     the documentation and/or other materials provided with the distribution.
+
+  3. The names of the authors may not be used to endorse or promote products
+     derived from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL JCRAFT,
+INC. OR ANY CONTRIBUTORS TO THIS SOFTWARE BE LIABLE FOR ANY DIRECT, INDIRECT,
+INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA,
+OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * This program is based on zlib-1.1.3, so all credit should go authors
+ * Jean-loup Gailly(jloup@gzip.org) and Mark Adler(madler@alumni.caltech.edu)
+ * and contributors of zlib.
+ */
+
+namespace Org.BouncyCastle.Utilities.Zlib {
+
+    public sealed class ZStream{
+
+        private const int MAX_WBITS=15;        // 32K LZ77 window
+        private const int DEF_WBITS=MAX_WBITS;
+
+        private const int Z_NO_FLUSH=0;
+        private const int Z_PARTIAL_FLUSH=1;
+        private const int Z_SYNC_FLUSH=2;
+        private const int Z_FULL_FLUSH=3;
+        private const int Z_FINISH=4;
+
+        private const int MAX_MEM_LEVEL=9;
+
+        private const int Z_OK=0;
+        private const int Z_STREAM_END=1;
+        private const int Z_NEED_DICT=2;
+        private const int Z_ERRNO=-1;
+        private const int Z_STREAM_ERROR=-2;
+        private const int Z_DATA_ERROR=-3;
+        private const int Z_MEM_ERROR=-4;
+        private const int Z_BUF_ERROR=-5;
+        private const int Z_VERSION_ERROR=-6;
+
+        public byte[] next_in;     // next input byte
+        public int next_in_index;
+        public int avail_in;       // number of bytes available at next_in
+        public long total_in;      // total nb of input bytes read so far
+
+        public byte[] next_out;    // next output byte should be put there
+        public int next_out_index;
+        public int avail_out;      // remaining free space at next_out
+        public long total_out;     // total nb of bytes output so far
+
+        public String msg;
+
+        internal Deflate dstate; 
+        internal Inflate istate; 
+
+        internal int data_type; // best guess about the data type: ascii or binary
+
+        public long adler;
+        internal Adler32 _adler=new Adler32();
+
+        public int inflateInit(){
+            return inflateInit(DEF_WBITS);
+        }
+        public int inflateInit(bool nowrap){
+            return inflateInit(DEF_WBITS, nowrap);
+        }
+        public int inflateInit(int w){
+            return inflateInit(w, false);
+        }
+
+        public int inflateInit(int w, bool nowrap){
+            istate=new Inflate();
+            return istate.inflateInit(this, nowrap?-w:w);
+        }
+
+        public int inflate(int f){
+            if(istate==null) return Z_STREAM_ERROR;
+            return istate.inflate(this, f);
+        }
+        public int inflateEnd(){
+            if(istate==null) return Z_STREAM_ERROR;
+            int ret=istate.inflateEnd(this);
+            istate = null;
+            return ret;
+        }
+        public int inflateSync(){
+            if(istate == null)
+                return Z_STREAM_ERROR;
+            return istate.inflateSync(this);
+        }
+        public int inflateSetDictionary(byte[] dictionary, int dictLength){
+            if(istate == null)
+                return Z_STREAM_ERROR;
+            return istate.inflateSetDictionary(this, dictionary, dictLength);
+        }
+
+        public int deflateInit(int level){
+            return deflateInit(level, MAX_WBITS);
+        }
+        public int deflateInit(int level, bool nowrap){
+            return deflateInit(level, MAX_WBITS, nowrap);
+        }
+        public int deflateInit(int level, int bits){
+            return deflateInit(level, bits, false);
+        }
+        public int deflateInit(int level, int bits, bool nowrap){
+            dstate=new Deflate();
+            return dstate.deflateInit(this, level, nowrap?-bits:bits);
+        }
+        public int deflate(int flush){
+            if(dstate==null){
+                return Z_STREAM_ERROR;
+            }
+            return dstate.deflate(this, flush);
+        }
+        public int deflateEnd(){
+            if(dstate==null) return Z_STREAM_ERROR;
+            int ret=dstate.deflateEnd();
+            dstate=null;
+            return ret;
+        }
+        public int deflateParams(int level, int strategy){
+            if(dstate==null) return Z_STREAM_ERROR;
+            return dstate.deflateParams(this, level, strategy);
+        }
+        public int deflateSetDictionary (byte[] dictionary, int dictLength){
+            if(dstate == null)
+                return Z_STREAM_ERROR;
+            return dstate.deflateSetDictionary(this, dictionary, dictLength);
+        }
+
+        // Flush as much pending output as possible. All deflate() output goes
+        // through this function so some applications may wish to modify it
+        // to avoid allocating a large strm->next_out buffer and copying into it.
+        // (See also read_buf()).
+        internal void flush_pending(){
+            int len=dstate.pending;
+
+            if(len>avail_out) len=avail_out;
+            if(len==0) return;
+
+            if(dstate.pending_buf.Length<=dstate.pending_out ||
+                next_out.Length<=next_out_index ||
+                dstate.pending_buf.Length<(dstate.pending_out+len) ||
+                next_out.Length<(next_out_index+len)){
+                //      System.out.println(dstate.pending_buf.length+", "+dstate.pending_out+
+                //			 ", "+next_out.length+", "+next_out_index+", "+len);
+                //      System.out.println("avail_out="+avail_out);
+            }
+
+            System.Array.Copy(dstate.pending_buf, dstate.pending_out,
+                next_out, next_out_index, len);
+
+            next_out_index+=len;
+            dstate.pending_out+=len;
+            total_out+=len;
+            avail_out-=len;
+            dstate.pending-=len;
+            if(dstate.pending==0){
+                dstate.pending_out=0;
+            }
+        }
+
+        // Read a new buffer from the current input stream, update the adler32
+        // and total number of bytes read.  All deflate() input goes through
+        // this function so some applications may wish to modify it to avoid
+        // allocating a large strm->next_in buffer and copying from it.
+        // (See also flush_pending()).
+        internal int read_buf(byte[] buf, int start, int size) {
+            int len=avail_in;
+
+            if(len>size) len=size;
+            if(len==0) return 0;
+
+            avail_in-=len;
+
+            if(dstate.noheader==0) {
+                adler=_adler.adler32(adler, next_in, next_in_index, len);
+            }
+            System.Array.Copy(next_in, next_in_index, buf, start, len);
+            next_in_index  += len;
+            total_in += len;
+            return len;
+        }
+
+        public void free(){
+            next_in=null;
+            next_out=null;
+            msg=null;
+            _adler=null;
+        }
+    }
+}
\ No newline at end of file