Difference between revisions of "DAT (Ever17)"

From Game Research Wiki
Jump to navigation Jump to search
 
(48 intermediate revisions by the same user not shown)
Line 1: Line 1:
[[Category:Archives]]
[[Category:Archives]]
 
Used in the following game(s):
Seen/used in the follow game(s):
* Ever17 (PC)
* Ever17


== Structure ==
== Structure ==
Line 14: Line 13:
| 4Bytes|| File Count ||
| 4Bytes|| File Count ||
|-
|-
| 8Bytes|| 0x00 Filler ||
| 8Bytes|| Padding || 0x00
|-
|-
!colspan="15"|Index
!colspan="15"|Index
Line 22: Line 21:
| 4Bytes || Offset || Starts from zero.
| 4Bytes || Offset || Starts from zero.
|-
|-
| 4Bytes|| File Size || Stored value is actual size doubled
| 4Bytes|| File Size || Stored value is double the actual size
|-
|-
| 24Bytes|| File name ||
| 24Bytes|| File name || Unused space will be padding (0x00).
|}
|}


Notes for 'wallpaper.dat'
== Notes ==
Extracting images straight from the container file will produce broken images. There is nothing wrong with the extraction process but rather the files themselves. Starting from offset 4352 and ending at 4608 in every single JPG, this 256 byte has been modified in some way.  
The files inside saver.dat, sysvoice.dat, and wallpaper.dat for Ever17 have had a small chunk data obfuscated. From offset 0x1100 to 0x1200 (256 bytes), data has been modified which breaks/corrupts the file. To undo this modification requires generating numbers to subtract from each byte in the affected area.
 
To generate these numbers, start off by adding each character in the file name of the file in question together, byte-by-byte. For example: 'test' 0x74 0x65 0x73 0x74 = 0x1C0 and then truncate the number down to 1 byte so 0xC0. This would be the first difference number out of the 256. The rest of the numbers are generated by using the previous number in a series of calculations.
* 1.) temp = previousNum + (previousNum * 2)
* 2.) temp = temp + (temp * 8)
* 3.) nextNum = previousNum + (temp * 4) + 1243
Note: This can probably be optimized to one statement to make it more readable.
 
y = x + (((x + (x * 2)) + (x + (x * 2) * 8))* 4) + 1243
 
y = 81x + 1243
 
nextNum = (81 * previousNum) + 1243;
 
A C++ representation.
<syntaxhighlight lang="cpp">
unsigned char diff_array[256];
 
void diff_gen(std::string &incoming_filename)
{
char firstValue = 0;
for (int i = 0; i < incoming_filename.size(); i++) //Better be 14 bytes... Though adding 0x00 shouldn't affect output.
{
firstValue += incoming_filename[i];
}
 
diff_array[0] = firstValue;
 
for (int i = 1; i < 255; i++)
{
                diff_array[i] = diff_array[i - 1] * 81 + 1243;
}
}
</syntaxhighlight>


2014/07/10 Currently looking into this with a dissembler and debugger. A small local subroutine for a function at 0x0000D286(input file) 0x0040D286 appears to be the annoying function. Seems the whole list of numbers used as the number that is subtracted from the incoming number from the extracted file, is in memory and not generated durring this operation. It may be generated sometimes else or a static set from the EXE image that may be used with other functions so there may be other files, such as the scripts, that may have obstructed spots in the file.
== Tools ==
-n/a-

Latest revision as of 23:19, 5 October 2016

Used in the following game(s):

  • Ever17 (PC)

Structure

Header
Size Content Description
4Bytes Magic/ID
4Bytes File Count
8Bytes Padding 0x00
Index
Size Content Description
4Bytes Offset Starts from zero.
4Bytes File Size Stored value is double the actual size
24Bytes File name Unused space will be padding (0x00).

Notes

The files inside saver.dat, sysvoice.dat, and wallpaper.dat for Ever17 have had a small chunk data obfuscated. From offset 0x1100 to 0x1200 (256 bytes), data has been modified which breaks/corrupts the file. To undo this modification requires generating numbers to subtract from each byte in the affected area.

To generate these numbers, start off by adding each character in the file name of the file in question together, byte-by-byte. For example: 'test' 0x74 0x65 0x73 0x74 = 0x1C0 and then truncate the number down to 1 byte so 0xC0. This would be the first difference number out of the 256. The rest of the numbers are generated by using the previous number in a series of calculations.

  • 1.) temp = previousNum + (previousNum * 2)
  • 2.) temp = temp + (temp * 8)
  • 3.) nextNum = previousNum + (temp * 4) + 1243

Note: This can probably be optimized to one statement to make it more readable.

y = x + (((x + (x * 2)) + (x + (x * 2) * 8))* 4) + 1243

y = 81x + 1243

nextNum = (81 * previousNum) + 1243;

A C++ representation.

unsigned char diff_array[256];

void diff_gen(std::string &incoming_filename)
{
	char firstValue = 0;
	for (int i = 0; i < incoming_filename.size(); i++) //Better be 14 bytes... Though adding 0x00 shouldn't affect output.
	{
		firstValue += incoming_filename[i];
	}

	diff_array[0] = firstValue;

	for (int i = 1; i < 255; i++)
	{
                diff_array[i] = diff_array[i - 1] * 81 + 1243;
	}
}

Tools

-n/a-