daggerhashimoto/bot.py

## readme.md

      
    Raw
  

              readme.md
            
          
    Cagin Donmez (cd0x23) | 24/05/2016
Viewer bot #2342 online Kappa

So I got bored last night and played around with faking Twitch viewers. I've failed to actually find a practical way to do this but, switching on to Eric Zhang's method, I've been able to get "acceptable (aka "meh") results.
The issue is that Twitch uses HLS for streaming and limits 10 connections per IP. Any more than 10 connections are not counted as viewers on stream. So what I did was, I added proxies, but this also was no good because good luck finding free proxies capable of sending requests every 5 seconds. However premium proxies is worth giving a shot and might actually work.
There's also another issue with CPU usage. I've tested the script with 3 builder threads on my low-end Digitalocean server.


As you can see the CPU usage spikes around 90% while there's no noticable change in bandwidth usage, which is odd because one would think that bandwidth would be the bottleneck in this case. I should also note that I personally don't think this method is practical as well (unless you have awesome hardware) since this test costs such processing power only for 10 viewers. If you want more viewers you're gonna have to use premium speed proxies and much more threads running this code.
The main reason I'm writing this is actually seeing a lot of people having a hard time running the code mainly because it's outdated. The code I'm posting is updated and tested on both Windows and Ubuntu 14.04. I've also added a proxy option.
You need to meet these requirements to be able to run this code properly.

livestreamer
requests
pyopenssl
urllib3

###Troubleshooting
If you're having parsing errors, you might want to check if livestreamer module is working properly. Also switching to "simplejson" to see a better error message might help too.
You need to modify the number of threads according to your hardware and needs or you might experience threading errors.

  
## bot.py
import requests
import subprocess
import json
import sys
import threading
import time
from Queue import Queue
import urllib3
import urllib3.contrib.pyopenssl

urllib3.contrib.pyopenssl.inject_into_urllib3()
#urllib3.disable_warnings() unverified HTTPS requests
numberOfViewers = int(sys.argv[1])
builderThreads = int(sys.argv[2])
startTime = time.time()
numberOfSockets = 0
concurrent = 25
urls = []
urlsUsed = []
proxies = {
    'http': 'http://173.201.183.172:8000',
    'http': 'http://94.181.34.64:81',
}


def getURL(): # Get tokens
  output = subprocess.Popen(["livestreamer", "twitch.tv/swagvyper", "-j"], stdout=subprocess.PIPE).communicate()[0]
  return json.loads(output)['streams']['worst']['url'] # Parse json and return the URL parameter

def build(): # Builds a set of tokens, aka viewers
	global numberOfSockets
	global numberOfViewers
	while True:
		if numberOfSockets < numberOfViewers:
			numberOfSockets += 1
			print "Building viewers " + str(numberOfSockets) + "/" + str(numberOfViewers)
			urls.append(getURL())

def view(): # Opens connections to send views
	global numberOfSockets
	while True:
		url=q.get()
        # requests.head(url, proxies=proxies)
		requests.head(url) # Sending a HEAD request is enough to count as a view
		if (url in urlsUsed):
			urls.remove(url)
			urlsUsed.remove(url)
			numberOfSockets -= 1
		else:
			urlsUsed.append(url)
		q.task_done()

if __name__ == '__main__':
	for i in range(0, builderThreads):
		threading.Thread(target = build).start()

	while True:
		while (numberOfViewers != numberOfSockets): # Wait until sockets are built
			time.sleep(1)

		q=Queue(concurrent*2)
		for i in range(concurrent):
			try:
				t=threading.Thread(target=view)
				t.daemon=True
				t.start()
			except:
				print 'thread error'
		try:
			for url in urls:
				print url
				q.put(url.strip())
				q.join()
		except KeyboardInterrupt:
			sys.exit(1)
	import requests
	import subprocess
	import json
	import sys
	import threading
	import time
	from Queue import Queue
	import urllib3
	import urllib3.contrib.pyopenssl

	urllib3.contrib.pyopenssl.inject_into_urllib3()
	#urllib3.disable_warnings() unverified HTTPS requests
	numberOfViewers = int(sys.argv[1])
	builderThreads = int(sys.argv[2])
	startTime = time.time()
	numberOfSockets = 0
	concurrent = 25
	urls = []
	urlsUsed = []
	proxies = {
	'http': 'http://173.201.183.172:8000',
	'http': 'http://94.181.34.64:81',
	}


	def getURL(): # Get tokens
	output = subprocess.Popen(["livestreamer", "twitch.tv/swagvyper", "-j"], stdout=subprocess.PIPE).communicate()[0]
	return json.loads(output)['streams']['worst']['url'] # Parse json and return the URL parameter

	def build(): # Builds a set of tokens, aka viewers
	global numberOfSockets
	global numberOfViewers
	while True:
	if numberOfSockets < numberOfViewers:
	numberOfSockets += 1
	print "Building viewers " + str(numberOfSockets) + "/" + str(numberOfViewers)
	urls.append(getURL())

	def view(): # Opens connections to send views
	global numberOfSockets
	while True:
	url=q.get()
	# requests.head(url, proxies=proxies)
	requests.head(url) # Sending a HEAD request is enough to count as a view
	if (url in urlsUsed):
	urls.remove(url)
	urlsUsed.remove(url)
	numberOfSockets -= 1
	else:
	urlsUsed.append(url)
	q.task_done()

	if __name__ == '__main__':
	for i in range(0, builderThreads):
	threading.Thread(target = build).start()

	while True:
	while (numberOfViewers != numberOfSockets): # Wait until sockets are built
	time.sleep(1)

	q=Queue(concurrent*2)
	for i in range(concurrent):
	try:
	t=threading.Thread(target=view)
	t.daemon=True
	t.start()
	except:
	print 'thread error'
	try:
	for url in urls:
	print url
	q.put(url.strip())
	q.join()
	except KeyboardInterrupt:
	sys.exit(1)