public DataFile Compress(DataFile toCompress)
{
var ppmTables = new PPMTables(_maxOrder);
var ac = new ArithmeticCoder();
for (int i = 0; i < toCompress.Length; i++)
{
if (i % _cleanUpLimit == 0)
{
ppmTables.CleanUp();
}
var entry = new Entry(toCompress.GetByte(i), GetContextFromFile(toCompress, i));
ContextTable.ToEncode toEncode;
EncodeInfo encodeInfo;
while ((toEncode = ppmTables.LookUpAndUpdate(entry, out encodeInfo)) !=
ContextTable.ToEncode.EncodeSymbol)
{
if (toEncode == ContextTable.ToEncode.EncodeEscape)
{
ac.Encode(encodeInfo.Count, encodeInfo.CumulativeCount, encodeInfo.TotalCount);
}
entry.NextContext();
}
ac.Encode(encodeInfo.Count, encodeInfo.CumulativeCount, encodeInfo.TotalCount);
}
ac.FinalizeInterval();
var output = ac.GetEncodedBitString().ToArray();
return(new DataFile(output));
}
/// <summary>
/// Encodes <paramref name="symbol"/> using the <paramref name="coder"/> with
/// the provided <paramref name="model"/>.
/// </summary>
public static void Encode <TSymbolType>(this ArithmeticCoder coder, TSymbolType symbol, IModel <TSymbolType> model,
WriteBitDelegate bitWriter)
where TSymbolType : struct
{
// cumulate frequencies
Range count = model.GetRange(symbol);
// encode symbol
coder.Encode(count, model.TotalFrequencies, bitWriter);
// update model
model.Update(symbol);
}
public void TestBasicArithmeticCodingRoundtrip2()
{
const string testString = "ユチエフ-8は素晴らしいです";
var data = Encoding.UTF8.GetBytes(testString);
var frequencies = ArithmeticCoder.CalculateFrequencies(data);
var encodedData = ArithmeticCoder.Encode(data, frequencies);
var decodedData = ArithmeticCoder.Decode(encodedData, frequencies, data.Length);
var resultString = Encoding.UTF8.GetString(decodedData);
Console.WriteLine("{0} {1} {2}", data.Length, encodedData.Length, decodedData.Length);
Console.WriteLine(encodedData.Aggregate("", (s, b) => s + ", " + b));
Console.WriteLine(testString);
Console.WriteLine(resultString);
Assert.LessOrEqual(encodedData.Length, data.Length);
Assert.AreEqual(testString, resultString);
}