Class LzmaCodec

Namespace

CHDSharp.Encoder

Assembly

CHDSharpLib.dll

LZMA compression matching MAME's chd_lzma_compressor: raw headerless LZMA with no end marker, properties lc=3/lp=0/pb=2, dictionary size = hunk bytes and numFastBytes=64 (the LZMA "level 8" profile that chdman configures via LzmaEncProps; the match-finder cycle count 16 + fb/2 = 48 follows automatically). Backed by Igor Pavlov's official LZMA SDK C# encoder (public domain, ported into this project); the encoder never writes the 5-byte property header, so the stream is already in CHD's raw format; the decoder synthesizes the properties (see CHDSharpLib's CHDReaders.Lzma). The port matches the C encoder's byte emission: the price table uses the C's kNumMoveReducingBits/kNumBitPriceShiftBits = 4/4 (the SDK C# line still uses 2/6, which flips near-tie optimal-parser decisions) and the BT4 match finder walks the btree with maxLen = 3 like MAME's Bt4_MatchFinder_GetMatches. Output is byte-identical to chdman on the battle-test corpus (verified per-hunk on 1,664 raw hunks and the cdlz CD hunks).

public sealed class LzmaCodec : IChdCodec

Inheritance

object

LzmaCodec

Implements

IChdCodec

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.ReferenceEquals(object, object)

object.ToString()

Constructors

LzmaCodec(uint)

Creates an LZMA codec for the given hunk size.

public LzmaCodec(uint hunkBytes)

Parameters

hunkBytes uint

Hunk size in bytes (becomes the LZMA dictionary size).

Properties

Tag

The four-character codec tag (see CodecTags).

public uint Tag { get; }

Property Value

uint

Methods

Compress(byte[])

Compresses a full hunk. Returns null when the codec does not reduce the size.

public byte[]? Compress(byte[] data)

Parameters

data byte[]

Returns

byte[]