Extract XML from NSP

extractnsp.py

#!/usr/bin/env python

import argparse
import os
import struct
import sys
import xml.etree.ElementTree as etree

parser = argparse.ArgumentParser()
parser.add_argument('file', help='NSP file')
parser.add_argument('--output', '-o', help='Output directory')
parser.add_argument('--verbose', '-v', help='Display verbose log', action='store_true')

args = parser.parse_args()

# https://switchbrew.org/index.php?title=NCA_Format#PFS0
with open(args.file, 'rb') as fin:
	magic = fin.read(4)
	if magic != b'PFS0'[:4]:
		print('Not a valid NSP file', file=sys.stderr)
		sys.exit(1)

	number_of_files = struct.unpack('i', fin.read(4))[0]
	size_of_the_string_table = struct.unpack('i', fin.read(4))[0]
	struct.unpack('i', fin.read(4))[0] # Zero/Reserved
	
	string_table_offset = 0x10 + (0x18 * number_of_files)
	data_offset = string_table_offset + size_of_the_string_table
	
	if args.verbose:
		print('Number of files: {}'.format(number_of_files), file=sys.stderr)
		print('String Table offset: {:08X}'.format(string_table_offset), file=sys.stderr)
		print('Data offset: {:08X}'.format(data_offset), file=sys.stderr)
		print('', file=sys.stderr)
	
	fin.seek(0x10)
	
	files = []
	
	for i in range(0, number_of_files):
		offset_of_file_in_data = struct.unpack('q', fin.read(8))[0]
		size_of_file_in_data = struct.unpack('q', fin.read(8))[0]
		offset_of_filename_in_string_table = struct.unpack('i', fin.read(4))[0]
		struct.unpack('i', fin.read(4))[0] # Zero/Reserved
		
		files.append((offset_of_file_in_data, size_of_file_in_data, offset_of_filename_in_string_table, ))
	
	for f in files:
		fin.seek(string_table_offset + f[2])
		
		filename = ''
		while True:
			c = fin.read(1)
			if c == b'\x00':
				break
			filename += c.decode('utf-8')
		
		if args.verbose:
			print('Filename: {}'.format(filename), file=sys.stderr)
			print('Offset of file in Data: {:08X}'.format(data_offset + f[0]), file=sys.stderr)
			print('Size of file in Data: {:08X}'.format(f[1]), file=sys.stderr)
			print('', file=sys.stderr)
		
		fin.seek(data_offset + f[0])
		xml = fin.read(f[1])
		
		if args.output:
			if not os.path.exists(args.output):
				os.makedirs(args.output)
			filename = args.output + '/' + filename
		
		with open(filename, 'wb') as fout:
			fout.write(xml)
		
		print(filename)