Cleaned up code a bit

proxytools/cleansocks.py

@@ -2,39 +2,35 @@
 # CleanSocks - Developed by acidvegas in Python (https://git.acid.vegas/proxytools)
 
 '''
-Requirements:
-	PySocks (https://pypi.python.org/pypi/pysocks)
 
-This script will clean a list of proxies by removing duplicates, checking for valid formats (IP:PORT), and testing if the proxies are working
+This script will clean a list of proxies by removing duplicates, checking for valid formats (IP:PORT), & testing if the proxies are working
+
 '''
 
 import argparse
 import concurrent.futures
 import os
 import re
-import sys
 
-sys.dont_write_bytecode = True
+# Globals
+good      = list()
+print_bad = True
 
-def is_proxy(proxy):
-	return re.match('^(([0-9]|[1-9][0-9]|1[0-9]{2}|2[0-4][0-9]|25[0-5])\.){3}([0-9]|[1-9][0-9]|1[0-9]{2}|2[0-4][0-9]|25[0-5]):(?:6553[0-5]|655[0-2][0-9]|65[0-4][0-9]{2}|6[0-4][0-9]{3}|[1-5][0-9]{4}|[1-9][0-9]{1,3}|[0-9])$', proxy)
-
-def test_proxy(proxy):
-	global good
+def check(proxy):
 	ip, port = proxy.split(':')
 	try:
 		sock = socks.socksocket()
 		sock.set_proxy(socks.SOCKS5, ip, int(port))
 		sock.settimeout(args.timeout)
-		sock.connect(('www.google.com', 80))
+		sock.connect('www.google.com', 80)) # NOTE: use your own host instead in-case connections are blocked
 	except:
-		print('BAD  | ' + proxy)
+		if print_bad:
+			print('\033[1;31mBAD\033[0m  \033[30m|\033[0m ' + proxy)
 	else:
-		print('GOOD | ' + proxy)
+		print('\033[1;32mGOOD\033[0m \033[30m|\033[0m ' + proxy)
 		good.append(proxy)
-	finally:
-		sock.close()
 
+# Main
 parser = argparse.ArgumentParser(usage='%(prog)s <input> <output> [options]')
 parser.add_argument('input',           help='file to scan')
 parser.add_argument('output',          help='file to output')
@@ -47,24 +43,18 @@ except ImportError:
 	raise SystemExit('missing pysocks module (https://pypi.python.org/pypi/pysocks)')
 if not os.path.isfile(args.input):
 	raise SystemExit('no such input file')
-proxies = set([line.strip() for line in open(args.input).readlines() if is_proxy(line)])
+initial = len(open(args.input).readlines())
+proxies = set([proxy for proxy in re.findall('[0-9]+\.[0-9]+\.[0-9]+\.[0-9]+:[0-9]+', open(args.input).read(), re.MULTILINE)])
 if not proxies:
 	raise SystemExit('no proxies found from input file')
-deduped, ips = list(), list()
-for proxy in proxies:
-	ip = proxy.split(':')[0]
-	if ip not in ips:
-		ips.append(ip)
-		deduped.append(proxy)
-deduped.sort()
-good = list()
 with concurrent.futures.ThreadPoolExecutor(max_workers=args.threads) as executor:
-	checks = {executor.submit(test_proxy, proxy): proxy for proxy in deduped}
+	checks = {executor.submit(check, proxy): proxy for proxy in proxies}
 	for future in concurrent.futures.as_completed(checks):
 		checks[future]
 good.sort()
 with open(args.output, 'w') as output_file:
 	output_file.write('\n'.join(good))
-print('Total : ' + format(len(proxies),           ',d'))
-print('Good  : ' + format(len(good),              ',d'))
-print('Bad   : ' + format(len(proxies)-len(good), ',d'))
+print('\033[34mTotal\033[0m : ' + format(len(proxies),           ',d'))
+print('\033[34mGood\033[0m  : ' + format(len(good),              ',d'))
+print('\033[34mBad\033[0m   : ' + format(len(proxies)-len(good), ',d'))
+print('\033[34mDupe\033[0m  : ' + format(initial-len(proxies),   ',d'))