draci_string_extract.py

#!/usr/bin/python
# 
# Strings Extractor for Draci Historie
#
# usage:
#		run this script from the same directory as the
#		RETEZCE.DFW file. It will create a RETEZCE.TXT file
#		that can be edited in any text editor
#
from struct import *
from os import path
from sys import exit

# Target archive with the Draci Historie text strings
BASE_FILENAME = "RETEZCE.DFW"
DUMP_FILENAME = "RETEZCE.TXT"

# Read a byte off the filestream if no target specified. If a target
# is given, interpret it as a numeric value
def read(target = None):
	value = target if target != None else infile.read(1)
	return unpack("B", value)[0]

# Read 2-bytes (little-endian) off the filestream
def read16():
	value = infile.read(2)
	return unpack("<H", value)[0]

# Read 4-bytes (little-endian) off the filestream
def read32():
	value = infile.read(4)
	return unpack("<L", value)[0]	


# Ensure that the file is in the same location as the
# text extraction script
if not path.exists(BASE_FILENAME):
		print BASE_FILENAME + " not found"
		exit()

print "Found file [" + BASE_FILENAME + "]"
infile = open(BASE_FILENAME, "r")

# Read in the archive and record information
header        = infile.read(4)
entries       = read16()
footer_offset = read32()

infile.seek(footer_offset)
offsets = []

for i in range(0,entries):
	offsets.append(read32())

outfile = open(DUMP_FILENAME, "wb")

for j in range(0, entries):
	infile.seek(offsets[j])
	comp_len   = read16() # string length + header size (6)
	length     = read16() # string length
	comp_type  = read()   # always zero
	data_crc   = read()
	
	# XXX The final byte seems to indicate string length, but isn't
	# used directly (though it could be)
	#string_len = read()
	
	# Read the string in, but ignore the final character, as it's
	# always a pipe
	data = infile.read(length)		
	crc = 0
	
	for k in range(0, length):
		crc ^= read(data[k])
	
	if crc != data_crc:
		print "CRC Mismatch for entry #" + j
		infile.close()
		exit
	
	# Note that our output needs to make the following adjustments
	#  (a) start from start + 1 as that byte, which contains the length 
	#      of the string, is unused when extracting the string, but is 
	#      used for CRC calculation
	#  (b) drop the last character, which is always a pipe (|)
	#  (c) add a line break to the end of each record
	outfile.write(data[1:len(data)-1] + "\n")

print "Dumped strings to [" + DUMP_FILENAME + "]"

outfile.close()
infile.close()

draci_strings_build.py

#!/usr/bin/python
# 
# Strings Builder for Draci Historie
#
# usage:
#		run this script from the same directory as the
#		RETEZCE.TXT file (which was created with draci_strings_extract.py)
#		It will build a RETEZCE.DFW.NEW file that can then be used
#		with Draci Historie in ScummVM. 
#		This script appends a .NEW extension to the generated file just in
#		case it has been run from the Draci Historie location, and doen't
#		assume that you want to overwrite the original.
#
from struct import pack, unpack
from os     import path
from sys    import exit

# Target archive with the Draci Historie text strings
BASE_FILENAME = "RETEZCE.TXT"
DUMP_FILENAME = "RETEZCE.DFW.NEW"

# Return the given value as numeric
def read(target):	
	return unpack("B", target)[0]

# Read a byte off the filestream if no target specified. If a target
# is given, interpret it as a numeric value
def w(target):	
	return pack("B", target)[0]

# Read 2-bytes (little-endian) off the filestream
def w16(target):	
	return pack("<H", target)

# Read 4-bytes (little-endian) off the filestream
def w32(target):	
	return pack("<L", target)

# Generate a simple CRC value
def generate_crc(data):
	crc = 0
	for i in range(0, len(data)):
		crc ^= read(data[i])
	return crc

# Ensure that the file is in the same location as the
# strings builder script
if not path.exists(BASE_FILENAME):
		print BASE_FILENAME + " not found"
		exit()

print "Found file [" + BASE_FILENAME + "]"

infile  = open(BASE_FILENAME, "r")
entries = []

# Read each line into an array
for line in infile:
	entries.append(line)
infile.close()

outfile = open(DUMP_FILENAME, "wb")
outfile.write("BAR!")
outfile.write(w16(len(entries)))
outfile.write(w32(0)) # Placeholder for footer location

offsets = []

for entry in entries:	
	entry = entry[0:len(entry) - 1] # Remove the line break
	entry += "|"                    # Each entry is terminated by a pipe
	length = len(entry)
	entry = w(length) + entry
	comp_len = length + 7
	outfile.write(w16(comp_len))   # Compressed Length
	outfile.write(w16(length + 1)) # Entry Length
	outfile.write(w(0))            # Compression Type (always zero)
	crc = generate_crc(entry)
	outfile.write(w(crc))	
	outfile.write(entry)
	
	offsets.append(comp_len)

pos = 10
for offset in offsets:	
	outfile.write(w32(pos))
	pos += offset	

outfile.write(w32(pos)) # Final entry is footer location
outfile.seek(6)
outfile.write(w32(pos)) # Write foother location in header

print "Built strings file [" + DUMP_FILENAME + "]"
	
outfile.close()

MD5Check

Just so you know, running an md5 sum on the two files (with no alteration to the generated text file) should yield identical results ;)

alex@home:~$ md5sum RETEZCE.DFW RETEZCE.DFW.NEW
6127b1f6e0cfc23bb1a42959063034d8  RETEZCE.DFW
6127b1f6e0cfc23bb1a42959063034d8  RETEZCE.DFW.NEW