Plants vs Zombies main.pak unpacker

unpack_mainpak.py

#!/usr/bin/env python3

# Plants vs Zombies main.pak unpacker

import os
from calendar import timegm
from datetime import datetime, timezone

INPUT_FILE="main.pak"
OUTPUT_DIR="main_unpacked"

HEADER_MAGICK=0xBAC04AC0
HEADER_VERSION=0x00000000

EPOCH_AS_FILETIME = 116444736000000000  # January 1, 1970 as filetime
HUNDREDS_OF_NS = 10000000

def windows_filetime_to_datetime(filetime: int) -> datetime:
	"""
	Converts a Windows filetime number to a Python datetime. The new
	datetime object is timezone-naive but is equivalent to tzinfo=utc.
	"""

	# Get seconds and remainder in terms of Unix epoch
	s, ns100 = divmod(filetime - EPOCH_AS_FILETIME, HUNDREDS_OF_NS)
	# Convert to datetime object, with remainder as microseconds.
	return datetime.fromtimestamp(s, timezone.utc).replace(microsecond=(ns100 // 10))


if not os.path.exists(OUTPUT_DIR):
    os.makedirs(OUTPUT_DIR)

in_data = open(INPUT_FILE, 'rb').read()

in_data = list(map(lambda byte: byte^0xF7, in_data)) # decode xor "cypher"

# open(OUTPUT_DIR + "/main.pak.decoded", 'wb').write(bytes(in_data)) # save decoded file

data_header = bytes(in_data[:8])
in_data = in_data[8:]

data_magick = int.from_bytes(data_header[0:4], byteorder='little')
data_version = int.from_bytes(data_header[4:8], byteorder='little')

if data_magick != HEADER_MAGICK:
    print("Invalid header. Magick is {:x}, should be {:x}".format(data_magick, HEADER_MAGICK))
    exit(-1)

if data_version != HEADER_VERSION:
    print("Invalid header. Version is {:x}, should be {:x}".format(data_version, HEADER_VERSION))
    exit(-1)


class DataItemRecord:
    name = ""
    file_size = 0
    date = 0

    def __init__(self, name: str, file_size: int, date: datetime):
        self.name = name
        self.file_size = file_size
        self.date = date


item_records = []
catalog_size = 0
while True:
    record_type = in_data[catalog_size]

    if record_type != 0:
        catalog_size += 1
        break

    name_length = in_data[catalog_size+1]

    name = ''.join([chr(x) for x in in_data[catalog_size+2:catalog_size+2+name_length]])
    file_size = int.from_bytes(bytes(in_data[catalog_size+2+name_length+0:catalog_size+2+name_length+4]), 'little')
    file_date = int.from_bytes(bytes(in_data[catalog_size+2+name_length+4:catalog_size+2+name_length+4+8]), 'little')

    item_records.append(DataItemRecord(name, file_size, windows_filetime_to_datetime(file_date)))

    catalog_size+=2+name_length+4+8

content_offset = 0
for record in item_records:
    print("{:08x} {} {}".format(record.file_size, record.date, record.name))

    output_name = OUTPUT_DIR + '/' + record.name.replace('\\', '/')

    dirname = os.path.dirname(os.path.abspath(output_name))
    if not os.path.exists(dirname):
        os.makedirs(dirname)

    res_file = open(output_name, 'wb')
    res_file.write(bytes(in_data[catalog_size+content_offset:catalog_size+content_offset+record.file_size]))
    res_file.close()
    os.utime(output_name, (record.date.timestamp(), record.date.timestamp()))
    content_offset += record.file_size