Initial commit

This commit is contained in:
2026-02-02 04:50:13 +01:00
commit 5b11698731
22592 changed files with 7677434 additions and 0 deletions

View File

@@ -0,0 +1,83 @@
/* ScummVM - Graphic Adventure Engine
*
* ScummVM is the legal property of its developers, whose names
* are too numerous to list here. Please refer to the COPYRIGHT
* file distributed with this source distribution.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/
#include "ultima/ultima4/core/lzw/hash.h"
namespace Ultima {
namespace Ultima4 {
namespace LZW {
int probe1(byte root, int codeword) {
int newHashCode = ((root << 4) ^ codeword) & 0xfff;
return (newHashCode);
}
/* The secondary probe uses some assembler instructions that aren't easily translated to C. */
int probe2(byte root, int codeword) {
/* registers[0] == AX, registers[1] == DX */
long registers[2], temp;
long carry, oldCarry;
int i, j;
/* the pre-mul part */
registers[1] = 0;
registers[0] = ((root << 1) + codeword) | 0x800;
/* the mul part (simulated mul instruction) */
/* DX:AX = AX * AX */
temp = (registers[0] & 0xff) * (registers[0] & 0xff);
temp += 2 * (registers[0] & 0xff) * (registers[0] >> 8) * 0x100;
registers[1] = (temp >> 16) + (registers[0] >> 8) * (registers[0] >> 8);
registers[0] = temp & 0xffff;
/* if DX != 0, the mul instruction sets the carry flag */
if (registers[1] == 00) {
carry = 0;
} else {
carry = 1;
}
/* the rcl part */
for (i = 0; i < 2; i++) { /* 2 rcl's */
for (j = 0; j < 2; j++) { /* rotate through 2 registers */
oldCarry = carry;
carry = (registers[j] >> 15) & 1;
registers[j] = (registers[j] << 1) | oldCarry;
registers[j] = registers[j] & 0xffff; /* make sure register stays 16 bit */
}
}
/* final touches */
registers[0] = ((registers[0] >> 8) | (registers[1] << 8)) & 0xfff;
return ((int)registers[0]);
}
int probe3(int hashCode) {
const long probeOffset = 0x1fd; /* I think 0x1fd is prime */
long newHashCode = (hashCode + probeOffset) & 0xfff;
return ((int)newHashCode);
}
} // End of namespace LZW
} // End of namespace Ultima4
} // End of namespace Ultima

View File

@@ -0,0 +1,39 @@
/* ScummVM - Graphic Adventure Engine
*
* ScummVM is the legal property of its developers, whose names
* are too numerous to list here. Please refer to the COPYRIGHT
* file distributed with this source distribution.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/
#ifndef ULTIMA4_CORE_LZW_HASH_H
#define ULTIMA4_CORE_LZW_HASH_H
#include "common/scummsys.h"
namespace Ultima {
namespace Ultima4 {
namespace LZW {
int probe1(byte root, int codeword);
int probe2(byte root, int codeword);
int probe3(int hashCode);
} // End of namespace LZW
} // End of namespace Ultima4
} // End of namespace Ultima
#endif

View File

@@ -0,0 +1,315 @@
/* ScummVM - Graphic Adventure Engine
*
* ScummVM is the legal property of its developers, whose names
* are too numerous to list here. Please refer to the COPYRIGHT
* file distributed with this source distribution.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/
/*
* A few files from Ultima 4 (PC version) have been compressed with the LZW algorithm.
* There are two things that make the U4 implementation of the LZW decoding algorithm special:
* 1) It uses a fixed codeword length of 12 bits.
* The advantages over variable-length codewords are faster decompression and simpler code.
* 2) The dictionary is implemented as a hash table.
* While the dictionary is supposed to implemented as a hash table in the LZW *en*coder (to speed up
* string searches), there is no reason not to implement it as a simple array in the decoder.
* But since U4 uses a hash table in the decoder, this C version must do the same (or it won't be
* able to decode the U4 files).
*
* An explanation on LZW data (de)compression can be found here:
* https://web.archive.org/web/20191231131544/https://marknelson.us/posts/1989/10/01/lzw-data-compression.html
* https://web.archive.org/web/20191231131532/https://marknelson.us/posts/2011/11/08/lzw-revisited.html
*/
#include "ultima/ultima4/core/lzw/lzw.h"
#include "ultima/ultima4/core/lzw/hash.h"
namespace Ultima {
namespace Ultima4 {
namespace LZW {
typedef void (*WRITE_DECOMP)(byte root, byte *destination, long *position);
struct lzwDictionaryEntry {
byte root;
int codeword;
byte occupied;
};
long generalizedDecompress(WRITE_DECOMP outFunc, byte *compressedMem, byte *decompressedMem, long compressedSize);
int getNextCodeword(long *bitsRead, byte *compressedMem);
void discardRoot(byte root, byte *destination, long *position);
void outputRoot(byte root, byte *destination, long *position);
void getString(int codeword, lzwDictionaryEntry *lzwDictionary, byte *stack, int *elementsInStack);
int getNewHashCode(byte root, int codeword, lzwDictionaryEntry *dictionary);
byte hashPosFound(int hashCode, byte root, int codeword, lzwDictionaryEntry *dictionary);
/*
* This function returns the decompressed size of a block of compressed data.
* It doesn't decompress the data.
* Use this function if you want to decompress a block of data, but don't know the decompressed size
* in advance.
*
* There is some error checking to detect if the compressed data is corrupt, but it's only rudimentary.
* Returns:
* No errors: (long) decompressed size
* Error: (long) -1
*/
long lzwGetDecompressedSize(byte *compressedMem, long compressedSize) {
return (generalizedDecompress(&discardRoot, compressedMem, nullptr, compressedSize));
}
/*
* Decompresses a block of compressed data from memory to memory.
* Use this function if you already know the decompressed size.
*
* This function assumes that *decompressed_mem is already allocated, and that the decompressed data
* will fit into *decompressed_mem.
* There is some error checking to detect if the compressed data is corrupt, but it's only rudimentary.
* Returns:
* No errors: (long) decompressed size
* Error: (long) -1
*/
long lzwDecompress(byte *compressedMem, byte *decompressedMem, long compressedSize) {
return (generalizedDecompress(&outputRoot, compressedMem, decompressedMem, compressedSize));
}
/* --------------------------------------------------------------------------------------
Functions used only inside lzw.c
-------------------------------------------------------------------------------------- */
/*
* This function does the actual decompression work.
* Parameters:
* perform_decompression: FALSE ==> return decompressed size, but discard decompressed data
* compressed_mem: compressed data
* decompressed_mem: this is where the compressed data will be decompressed to
* compressed_size: size of the compressed data (in bytes)
*/
long generalizedDecompress(WRITE_DECOMP outFunc, byte *compressedMem, byte *decompressedMem, long compressedSize) {
int i;
/* re-initialize the dictionary when there are more than 0xccc entries */
const int maxDictEntries = 0xccc;
const int lzwStackSize = 0x8000;
const int lzwDictionarySize = 0x1000;
int old_code;
int new_code;
byte character;
long bitsRead = 0;
long bytesWritten = 0;
/* newpos: position in the dictionary where new codeword was added */
/* must be equal to current codeword (if it isn't, the compressed data must be corrupt) */
/* unknownCodeword: is the current codeword in the dictionary? */
int newpos;
byte unknownCodeword;
/* initialize the dictionary and the stack */
lzwDictionaryEntry *lzwDictionary = (lzwDictionaryEntry *) malloc(sizeof(lzwDictionaryEntry) * lzwDictionarySize);
int codewordsInDictionary = 0;
byte *lzwStack = (byte *) malloc(sizeof(byte) * lzwStackSize);
int elementsInStack = 0;
/* clear the dictionary */
memset(lzwDictionary, 0, sizeof(lzwDictionaryEntry) * lzwDictionarySize);
for (i = 0; i < 0x100; i++) {
lzwDictionary[i].occupied = 1;
}
if (bitsRead + 12 <= compressedSize * 8) {
/* read OLD_CODE */
old_code = getNextCodeword(&bitsRead, compressedMem);
/* CHARACTER = OLD_CODE */
character = (byte)old_code;
/* output OLD_CODE */
outFunc(character, decompressedMem, &bytesWritten);
while (bitsRead + 12 <= compressedSize * 8) { /* WHILE there are still input characters DO */
/* read NEW_CODE */
new_code = getNextCodeword(&bitsRead, compressedMem);
if (lzwDictionary[new_code].occupied) { /* is the codeword in the dictionary? */
/* codeword is present in the dictionary */
/* it must either be a root or a non-root that has already been added to the dicionary */
unknownCodeword = 0;
/* STRING = get translation of NEW_CODE */
getString(new_code, lzwDictionary, lzwStack, &elementsInStack);
} else {
/* codeword is yet to be defined */
unknownCodeword = 1;
/* STRING = get translation of OLD_CODE */
/* STRING = STRING+CHARACTER */
lzwStack[elementsInStack] = character; /* push character on the stack */
elementsInStack++;
getString(old_code, lzwDictionary, lzwStack, &elementsInStack);
}
/* CHARACTER = first character in STRING */
character = lzwStack[elementsInStack - 1]; /* element at top of stack */
/* output STRING */
while (elementsInStack > 0) {
outFunc(lzwStack[elementsInStack - 1], decompressedMem, &bytesWritten);
elementsInStack--;
}
/* add OLD_CODE + CHARACTER to the translation table */
newpos = getNewHashCode(character, old_code, lzwDictionary);
lzwDictionary[newpos].root = character;
lzwDictionary[newpos].codeword = old_code;
lzwDictionary[newpos].occupied = 1;
codewordsInDictionary++;
/* check for errors */
if (unknownCodeword && (newpos != new_code)) {
/* clean up */
free(lzwStack);
free(lzwDictionary);
return (-1);
}
if (codewordsInDictionary > maxDictEntries) {
/* wipe dictionary */
codewordsInDictionary = 0;
memset(lzwDictionary, 0, sizeof(lzwDictionaryEntry) * lzwDictionarySize);
for (i = 0; i < 0x100; i++) {
lzwDictionary[i].occupied = 1;
}
if (bitsRead + 12 <= compressedSize * 8) {
new_code = getNextCodeword(&bitsRead, compressedMem);
character = (byte)new_code;
outFunc(character, decompressedMem, &bytesWritten);
} else {
/* clean up */
free(lzwStack);
free(lzwDictionary);
return (bytesWritten);
}
}
/* OLD_CODE = NEW_CODE */
old_code = new_code;
}
}
/* clean up */
free(lzwStack);
free(lzwDictionary);
return (bytesWritten);
}
/* read the next 12-bit codeword from the compressed data */
int getNextCodeword(long *bitsRead, byte *compressedMem) {
int codeword = (compressedMem[(*bitsRead) / 8] << 8) + compressedMem[(*bitsRead) / 8 + 1];
codeword = codeword >> (4 - ((*bitsRead) % 8));
codeword = codeword & 0xfff;
(*bitsRead) += 12;
return (codeword);
}
/* increment position pointer, but do not write root to memory */
void discardRoot(byte root, byte *destination, long *position) {
(*position)++;
}
/* output a root to memory */
void outputRoot(byte root, byte *destination, long *position) {
destination[*position] = root;
(*position)++;
}
/* --------------------------------------------------------------------------------------
Dictionary-related functions
-------------------------------------------------------------------------------------- */
/* pushes the string associated with codeword onto the stack */
void getString(int codeword, lzwDictionaryEntry *dictionary, byte *stack, int *elementsInStack) {
byte root;
int currentCodeword = codeword;
while (currentCodeword > 0xff) {
root = dictionary[currentCodeword].root;
currentCodeword = dictionary[currentCodeword].codeword;
stack[*elementsInStack] = root;
(*elementsInStack)++;
}
/* push the root at the leaf */
stack[*elementsInStack] = (byte)currentCodeword;
(*elementsInStack)++;
}
int getNewHashCode(byte root, int codeword, lzwDictionaryEntry *dictionary) {
int hashCode;
/* probe 1 */
hashCode = probe1(root, codeword);
if (hashPosFound(hashCode, root, codeword, dictionary)) {
return (hashCode);
}
/* probe 2 */
hashCode = probe2(root, codeword);
if (hashPosFound(hashCode, root, codeword, dictionary)) {
return (hashCode);
}
/* probe 3 */
do {
hashCode = probe3(hashCode);
} while (! hashPosFound(hashCode, root, codeword, dictionary));
return (hashCode);
}
byte hashPosFound(int hashCode, byte root, int codeword, lzwDictionaryEntry *dictionary) {
if (hashCode > 0xff) {
// hash codes must not be roots
byte c1 = 0, c2 = 0, c3 = 0;
if (dictionary[hashCode].occupied) {
// hash table position is occupied
c1 = 1;
// is our (root,codeword) pair already in the hash table?
c2 = dictionary[hashCode].root == root;
c3 = dictionary[hashCode].codeword == codeword;
} else {
// hash table position is free
c1 = 0;
}
return (!c1) || (c1 && c2 && c3);
} else {
return 0;
}
}
} // End of namespace LZW
} // End of namespace Ultima4
} // End of namespace Ultima

View File

@@ -0,0 +1,38 @@
/* ScummVM - Graphic Adventure Engine
*
* ScummVM is the legal property of its developers, whose names
* are too numerous to list here. Please refer to the COPYRIGHT
* file distributed with this source distribution.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/
#ifndef ULTIMA4_CORE_LZW_LZW_H
#define ULTIMA4_CORE_LZW_LZW_H
#include "common/scummsys.h"
namespace Ultima {
namespace Ultima4 {
namespace LZW {
long lzwGetDecompressedSize(byte *compressedMem, long compressedSize);
long lzwDecompress(byte *compressedMem, byte *decompressedMem, long compressedSize);
} // End of namespace LZW
} // End of namespace Ultima4
} // End of namespace Ultima
#endif

View File

@@ -0,0 +1,155 @@
/* ScummVM - Graphic Adventure Engine
*
* ScummVM is the legal property of its developers, whose names
* are too numerous to list here. Please refer to the COPYRIGHT
* file distributed with this source distribution.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/
#include "ultima/ultima4/core/lzw/lzw.h"
#include "ultima/ultima4/core/lzw/u4decode.h"
namespace Ultima {
namespace Ultima4 {
namespace LZW {
/*
* Loads a file, decompresses it (from memory to memory), and writes the decompressed data to another file
* Returns:
* -1 if there was an error
* the decompressed file length, on success
*/
long decompress_u4_file(Common::SeekableReadStream *in, long filesize, void **out) {
byte *compressed_mem, *decompressed_mem;
long compressed_filesize, decompressed_filesize;
long errorCode;
/* size of the compressed input file */
compressed_filesize = filesize;
/* input file should be longer than 0 bytes */
if (compressed_filesize == 0)
return (-1);
/* check if the input file is _not_ a valid LZW-compressed file */
if (!mightBeValidCompressedFile(in))
return (-1);
/* load compressed file into compressed_mem[] */
compressed_mem = (byte *) malloc(compressed_filesize);
in->read(compressed_mem, compressed_filesize);
/*
* determine decompressed file size
* if lzw_get_decompressed_size() can't determine the decompressed size (i.e. the compressed
* data is corrupt), it returns -1
*/
decompressed_filesize = lzwGetDecompressedSize(compressed_mem, compressed_filesize);
if (decompressed_filesize <= 0) {
return (-1);
}
/* decompress file from compressed_mem[] into decompressed_mem[] */
decompressed_mem = (byte *) malloc(decompressed_filesize);
/* testing: clear destination mem */
memset(decompressed_mem, 0, decompressed_filesize);
errorCode = lzwDecompress(compressed_mem, decompressed_mem, compressed_filesize);
free(compressed_mem);
*out = decompressed_mem;
return (errorCode);
}
long decompress_u4_memory(void *in, long inlen, void **out) {
byte *compressed_mem, *decompressed_mem;
long compressed_filesize, decompressed_filesize;
long errorCode;
/* size of the compressed input */
compressed_filesize = inlen;
/* input file should be longer than 0 bytes */
if (compressed_filesize == 0)
return (-1);
compressed_mem = (byte *) in;
/*
* determine decompressed data size
* if lzw_get_decompressed_size() can't determine the decompressed size (i.e. the compressed
* data is corrupt), it returns -1
*/
decompressed_filesize = lzwGetDecompressedSize(compressed_mem, compressed_filesize);
if (decompressed_filesize <= 0) {
return (-1);
}
/* decompress file from compressed_mem[] into decompressed_mem[] */
decompressed_mem = (byte *) malloc(decompressed_filesize);
/* testing: clear destination mem */
memset(decompressed_mem, 0, decompressed_filesize);
errorCode = lzwDecompress(compressed_mem, decompressed_mem, compressed_filesize);
*out = decompressed_mem;
return (errorCode);
}
/*
* Returns the size of a file, and moves the file pointer to the beginning.
* The file must already be open when this function is called.
*/
long getFilesize(Common::SeekableReadStream *input_file) {
return input_file->size();
}
/*
* If the input file is a valid LZW-compressed file, the upper 4 bits of
* the first byte must be 0, because the first codeword is always a root.
*/
byte mightBeValidCompressedFile(Common::SeekableReadStream *input_file) {
byte firstByte;
byte c1, c2, c3; /* booleans */
long input_filesize;
/* check if the input file has a valid size */
/* the compressed file is made up of 12-bit codewords, */
/* so there are either 0 or 4 bits of wasted space */
input_filesize = getFilesize(input_file);
c1 = (input_filesize * 8) % 12 == 0;
c2 = (input_filesize * 8 - 4) % 12 == 0;
// read first byte, and then reset back file pointer
input_file->seek(0);
firstByte = input_file->readByte();
input_file->seek(0);
c3 = (firstByte >> 4) == 0;
// check if upper 4 bits are 0
return ((c1 || c2) && c3);
}
} // End of namespace LZW
} // End of namespace Ultima4
} // End of namespace Ultima

View File

@@ -0,0 +1,40 @@
/* ScummVM - Graphic Adventure Engine
*
* ScummVM is the legal property of its developers, whose names
* are too numerous to list here. Please refer to the COPYRIGHT
* file distributed with this source distribution.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/
#ifndef ULTIMA4_CORE_LZW_U4DECODE_H
#define ULTIMA4_CORE_LZW_U4DECODE_H
#include "common/stream.h"
namespace Ultima {
namespace Ultima4 {
namespace LZW {
long decompress_u4_file(Common::SeekableReadStream *in, long filesize, void **out);
long getFilesize(Common::SeekableReadStream *input_file);
byte mightBeValidCompressedFile(Common::SeekableReadStream *compressed_file);
long decompress_u4_memory(void *in, long inlen, void **out);
} // End of namespace LZW
} // End of namespace Ultima4
} // End of namespace Ultima
#endif