wireshark mac manuf to json file

analysis.py

import json 

def main():
	contents = open('json_data_raw.json').read()
	response = json.loads(contents)

	items = {k.replace('-','').replace(':','').lower():v for k, v in response.iteritems()}

	short_macs = filter(lambda x: len(x[0]) == 6, items.iteritems())
	long_macs = filter(lambda x: len(x[0]) == 10, items.iteritems())

	data = {
		'mac_prefixes': dict(short_macs),
		'extended_mac_prefixes': dict(long_macs)
	}

	json_string = json.dumps(data)
	with open('indexed_mac_addresses_no_sort.json','w') as f:
		f.write(json_string)


if __name__ == '__main__':
	main()

manuf_to_json.py

import requests
import re
import json 

def main():

	response = dict()

	r = requests.get('https://code.wireshark.org/review/gitweb?p=wireshark.git;a=blob_plain;f=manuf', verify=False)
	lines = r.text.splitlines()

	for line in lines:

		try:

			line = line.encode('ascii', 'ignore')

			if line.startswith('#'):
				continue

			splitted = line.split('\t')
			mac_prefix = splitted[0]
			mac_prefix = mac_prefix.strip()

			is_valid_mac_prefix = re.match('^([0-9|A-F|a-f]{2}[:|-]){2}([0-9|A-F|a-f]{2})$',mac_prefix)
			is_valid_mac_mask = re.match('^([0-9|A-F|a-f]{2}[:|-]){5}(00\/36)$',mac_prefix)
			if not is_valid_mac_prefix:
				if is_valid_mac_mask:
					mac_prefix = mac_prefix.replace(':00/36', '')
				else:
					continue

			splitted = splitted[1].split('#')
			name = splitted[0]
			name = name.strip()

			description = name
			if len(splitted) > 1 :
				description = splitted[1]
				description = description.strip()

			if not name:
				continue

			response[mac_prefix] = {'name':name, 'description':description}
			
		except Exception as e:
			pass

	return response

if __name__ == '__main__':
	data = main()
	json_string = json.dumps(data)

	with open('json_data_raw.json','w') as f:
		f.write(json_string)

Scraping utility for wireshark's mac mapping file
https://code.wireshark.org/review/gitweb?p=wireshark.git;a=blob_plain;f=manuf