Editing CC's script?

[Serge50]

Hello
I've been trying to find a way to edit Chrono Cross's dialogue text since a while but always counter dead-ends.
All I've reached is dump the game files using Yazoo's cctools. Problem is the files extracted were all in .out format which I have no idea how to decompress/convert.The rest of Yazoo's c++ files didn't want to compile due to errors in  the code(or I'm doing it wrong?) .

I just want to edit the text for fun purpose    I hope I can get some help.

[Danetta]

Well, editing dialogs can be easily done with Terminus Tradiction Tools. It's not neccesary to use them for this purpose, but it's what it's designed for, so it's the easiest way.

[Serge50]

My thanks for your reply.
I tried the terminus traduction tools on a windows xp laptop and it worked fine.But on another one with win7 it didnt work and i get unusual errors when running the rein execs.I ran them on windows xp sp3 compatibility,gives fewer errors but still doesn't work when applying the patch on the iso.
It's not a big deal but I'm just curious what the problem is  .
Thanks again!

[Schala Zeal]

Just out of sheer interest, I've been trying to make my own Square LZSS de/compressor library in C11, but I've been stumped on the compression function as to what the 4 byte value at offset 8 is in the file header. My best guess is that it is a checksum of some sort. I've been trying to reverse-CRC it in hopes of finding a magic value that would confirm this, but re-CRCing with the magic value does not yield the same value as is in the file header. My gut still says that 4 byte value is used as an integrity checksum though.

I tried the CRC validation with some of FaustWolf's test files: Script.lzs (at offset 12 til EOF) and the extracted Script.txt as well as 2540.out (at offs 12 til EOF) on CD1

[Schala Zeal]

Think I may be doing something wrong here, but... give it a try

Code: [Select]

#include <cstring>
#include <fstream>
#include <iostream>
#include <numeric>
#include <string>
#include <vector>

int main(int argc, char **argv)
{
std::ifstream inf;
std::ofstream outf;
char *buf = new char[4096];
char c;
unsigned inlen = 0, outlen = 0;
std::vector<bool> data;
auto p = data.begin();

inf.open(argv[1], std::ifstream::binary);
inf.seekg(0, inf.end);
inlen = inf.tellg();
inf.seekg(0, inf.beg);
inf.read(buf, 4);
if (std::strncmp(buf, "sszl", 4) != 0) return 1;
std::memset(buf, 0, 4);
inf.read((char*)&outlen, 4);
std::cout << inlen << " / " << outlen << '\n';
inf.seekg(4, inf.cur);

while (inf.get(c))
for (int i = 0; i < 8; i++)
data.push_back((bool)((c >> i) & 1));
inf.close();
unsigned j = 0;
unsigned o = 0;
outf.open(std::string(argv[1])+".uncompressed", std::ofstream::binary);
for (int i = 0; i < data.size();)
{
if ((bool)data[i++])
{
char k = std::accumulate(data.begin()+i, data.begin()+(i+8), 0,
[](char x, char y)
{
return (x<<1)+y;
});
std::memmove(buf+(j%4096), &k, 1);
outf.put(k);
i+=8;
j++;
}
else
{
o = std::accumulate(data.begin()+i, data.begin()+(i+12), 0,
[](unsigned x, unsigned y)
{
return (x<<1)+y;
})-1;
unsigned k = std::accumulate(data.begin()+i, data.begin()+(i+4), 0,
[](unsigned x, unsigned y)
{
return (x<<1)+y;
})+2;
o %= 4096;

for (int m = 0; m < k; m++)
{
buf[(j+m)%4096]=buf[o];
outf.put(buf[o]);
o=(o+1)%4096;
i+=16;
j+=o;
}
}
}
outf.close();

/*for (auto i: data) std::cout << (i ? 1 : 0);
std::cout << '\n';*/
return 0;
}


[EgyLynx]

It is little surpised that iostream think need if something (only hello world) text are put screen... or ...

Well... i dont begin comments what include what...
If there are someone who are more exp these c family langugues...

[Schala Zeal]

I used iostream for verbose purposes.

[Lynxos]

Think I may be doing something wrong here, but... give it a try

Code: [Select]

#include <cstring>
#include <fstream>
#include <iostream>
#include <numeric>
#include <string>
#include <vector>

int main(int argc, char **argv)
{
std::ifstream inf;
std::ofstream outf;
char *buf = new char[4096];
char c;
unsigned inlen = 0, outlen = 0;
std::vector<bool> data;
auto p = data.begin();

inf.open(argv[1], std::ifstream::binary);
inf.seekg(0, inf.end);
inlen = inf.tellg();
inf.seekg(0, inf.beg);
inf.read(buf, 4);
if (std::strncmp(buf, "sszl", 4) != 0) return 1;
std::memset(buf, 0, 4);
inf.read((char*)&outlen, 4);
std::cout << inlen << " / " << outlen << '\n';
inf.seekg(4, inf.cur);

while (inf.get(c))
for (int i = 0; i < 8; i++)
data.push_back((bool)((c >> i) & 1));
inf.close();
unsigned j = 0;
unsigned o = 0;
outf.open(std::string(argv[1])+".uncompressed", std::ofstream::binary);
for (int i = 0; i < data.size();)
{
if ((bool)data[i++])
{
char k = std::accumulate(data.begin()+i, data.begin()+(i+8), 0,
[](char x, char y)
{
return (x<<1)+y;
});
std::memmove(buf+(j%4096), &k, 1);
outf.put(k);
i+=8;
j++;
}
else
{
o = std::accumulate(data.begin()+i, data.begin()+(i+12), 0,
[](unsigned x, unsigned y)
{
return (x<<1)+y;
})-1;
unsigned k = std::accumulate(data.begin()+i, data.begin()+(i+4), 0,
[](unsigned x, unsigned y)
{
return (x<<1)+y;
})+2;
o %= 4096;

for (int m = 0; m < k; m++)
{
buf[(j+m)%4096]=buf[o];
outf.put(buf[o]);
o=(o+1)%4096;
i+=16;
j+=o;
}
}
}
outf.close();

/*for (auto i: data) std::cout << (i ? 1 : 0);
std::cout << '\n';*/
return 0;
}


Care to share info on what this code does?

[alfadorredux]

Slight necro here, but the decompression algorithm goes approximately like this (stripping my comments from an implementation that at least somewhat works):

#create a 4K buffer filled with zeroes

#read 4 bytes ("sszl")

#read the decompressed size (4-byte little-endian)

#read the four mystery bytes

#while the input file isn't empty, read 512 bytes, or whatever's left

#unpack into a bit string, split, and reverse to produce an array of "bits" (that corresponds to the line "my @bits = reverse(split("", unpack("B4096", $buf)));" in Perl, you might need to read "man perlpacktut" if you can't figure out what's going on)

#(if there's slopover from an old read, push it onto the end)

#while there are at least 17 bits left (or we're processing the last few bits in the file), do
    #if the next bit is a 1
        #output an 8-bit literal byte to the output file and the buffer
    #if the next bit is a 0
        #create a pointer into the buffer from the next 12 bits
        #create a length from the next 4 bits
        #starting from the indicated position in the buffer, read length bytes and 
         write them into the buffer and the output file

That's about it.  Good luck.

[froodo]

A good thing for Schala (she's making a open-source reimplementation of CC engine)

[Schala Zeal]

Indeed, and an Assimp extension to read CC mdl files (and .mesh files maybe) as part of it.