A simple Python module to bypass Cloudflare's anti-bot page (also known as "I'm Under Attack Mode", or IUAM), implemented with Requests. Cloudflare changes their techniques periodically, so I will update this repo frequently.
This can be useful if you wish to scrape or crawl a website protected with Cloudflare. Cloudflare's anti-bot page currently just checks if the client supports Javascript, though they may add additional techniques in the future.
Due to Cloudflare continually changing and hardening their protection page, cloudscraper requires a JavaScript interpreter to solve Javascript challenges. This allows the script to easily impersonate a regular web browser without explicitly deobfuscating and parsing Cloudflare's Javascript.
Note: This only works when regular Cloudflare anti-bots is enabled (the "Checking your browser before accessing..." loading page). If there is a reCAPTCHA challenge, you're out of luck (At this stage... however we will be adding in Anti-CAPTCHA 3rd party support). Thankfully, the Javascript check page is much more common.
For reference, this is the default message Cloudflare uses for these sorts of pages:
Checking your browser before accessing website.com.
This process is automatic. Your browser will redirect to your requested content shortly.
Please allow up to 5 seconds...
Any script using cloudscraper will sleep for ~5 seconds for the first visit to any site with Cloudflare anti-bots enabled, though no delay will occur after the first request.
Simply run pip install cloudscraper
. The PyPI package is at https://pypi.python.org/pypi/cloudscraper/
Alternatively, clone this repository and run python setup.py install
.
- Python 2.7 - 3.x
- Requests >= 2.9.2
- Brotli >= 1.0.7
- requests_toolbelt >= 0.9.1
Have the ability to choose between Javascript Interpreters.
- js2py >=0.60
- Node.js
- Your computer or server may already have it (check with
node -v
). If not, you can install it withapt-get install nodejs
on Ubuntu. Debian requiresnodejs-legacy
. Otherwise, please read Node's installation instructions.
- Your computer or server may already have it (check with
python setup.py install
will install the Python dependencies automatically. Node is the only application you need to install yourself.
Cloudflare modifies their anti-bot protection page occasionally, So far it has changed maybe once per year on average.
If you notice that the anti-bot page has changed, or if this module suddenly stops working, please create a GitHub issue so that I can update the code accordingly.
- Many issues are a result of users not updating to the latest release of this project. Before filing an issue, please run the following command:
pip show cloudscraper
If the value of the version field is not the latest release, please run the following to update your package:
pip install cloudscraper -U
If you are still encountering a problem, open an issue and please include:
- The full exception and stack trace.
- The URL of the Cloudflare-protected page which the script does not work on.
- A Pastebin or Gist containing the HTML source of the protected page.
- The version number from
pip show cloudscraper
.
The simplest way to use cloudscraper is by calling create_scraper()
.
import cloudscraper
scraper = cloudscraper.create_scraper() # returns a CloudScraper instance
# Or: scraper = cloudscraper.CloudScraper() # CloudScraper inherits from requests.Session
print scraper.get("http://somesite.com").content # => "<!DOCTYPE html><html><head>..."
That's it...
Any requests made from this session object to websites protected by Cloudflare anti-bot will be handled automatically. Websites not using Cloudflare will be treated normally. You don't need to configure or call anything further, and you can effectively treat all websites as if they're not protected with anything.
You use cloudscraper exactly the same way you use Requests. CloudScraper
works identically to a Requests Session
object, just instead of calling requests.get()
or requests.post()
, you call scraper.get()
or scraper.post()
.
Consult Requests' documentation for more information.
If you already have an existing Requests session, you can pass it to create_scraper()
to continue using that session.
session = requests.session()
scraper = cloudscraper.create_scraper(sess=session)
Unfortunately, not all of Requests' session attributes are easily transferable, so if you run into problems with this, you should replace your initial sess = requests.session()
call with sess = cloudscraper.create_scraper()
.
scraper = cloudscraper.create_scraper(debug=True)
Or
scraper = cloudscraper.create_scraper()
scraper.debug = True
Normally, when a browser is faced with a Cloudflare IUAM challenge page, Cloudflare requires the browser to wait ~5 seconds before submitting the challenge answer. If a website is under heavy load, sometimes this may fail. One solution is to increase the delay (perhaps to 10 or 15 seconds, depending on the website). If you would like to override this delay, pass the delay
keyword argument to create_scraper()
or CloudScraper()
.
There is no need to override this delay unless cloudscraper generates an error recommending you increase the delay.
scraper = cloudscraper.create_scraper(delay=10)
or
scraper = cloudscraper.create_scraper()
scraper.delay = 10
Cloudscraper currently supports two JavaScript Interpreters
The default interpreter is set to js2py
, you can set which to use by defining the interpreter
parameter.
scraper = cloudscraper.create_scraper(interpreter='nodejs')
or
scraper = cloudscraper.create_scraper()
scraper.interpreter = 'nodejs'
We have added in Brotli decompression support in, and it is enabled by default, the only way to disable it, is by passing the allow_brotli
parameter set toFalse
to create_scraper()
scraper = cloudscraper.create_scraper(allow_brotli=False)
It's easy to integrate cloudscraper with other applications and tools. Cloudflare uses two cookies as tokens: one to verify you made it past their challenge page and one to track your session. To bypass the challenge page, simply include both of these cookies (with the appropriate user-agent) in all HTTP requests you make.
To retrieve just the cookies (as a dictionary), use cloudscraper.get_tokens()
. To retrieve them as a full Cookie
HTTP header, use cloudscraper.get_cookie_string()
.
get_tokens
and get_cookie_string
both accept Requests' usual keyword arguments (like get_tokens(url, proxies={"http": "socks5://localhost:9050"})
).
Please read Requests' documentation on request arguments for more information.
The two integration functions return a tuple of (cookie, user_agent_string)
.
You must use the same user-agent string for obtaining tokens and for making requests with those tokens, otherwise Cloudflare will flag you as a bot.
That means you have to pass the returned user_agent_string
to whatever script, tool, or service you are passing the tokens to (e.g. curl, or a specialized scraping tool), and it must use that passed user-agent when it makes HTTP requests.
Remember, you must always use the same user-agent when retrieving or using these cookies. These functions all return a tuple of (cookie_dict, user_agent_string)
.
get_tokens
is a convenience function for returning a Python dict containing Cloudflare's session cookies. For demonstration, we will configure this request to use a proxy. (Please note that if you request Cloudflare clearance tokens through a proxy, you must always use the same proxy when those tokens are passed to the server. Cloudflare requires that the challenge-solving IP and the visitor IP stay the same.)
If you do not wish to use a proxy, just don't pass the proxies
keyword argument. These convenience functions support all of Requests' normal keyword arguments, like params
, data
, and headers
.
import cloudscraper
proxies = {"http": "http://localhost:8080", "https": "http://localhost:8080"}
tokens, user_agent = cloudscraper.get_tokens("http://somesite.com", proxies=proxies)
print tokens
# => {
'cf_clearance': 'c8f913c707b818b47aa328d81cab57c349b1eee5-1426733163-3600',
'__cfduid': 'dd8ec03dfdbcb8c2ea63e920f1335c1001426733158'
}
get_cookie_string
is a convenience function for returning the tokens as a string for use as a Cookie
HTTP header value.
This is useful when crafting an HTTP request manually, or working with an external application or library that passes on raw cookie headers.
import cloudscraper
cookie_value, user_agent = cloudscraper.get_cookie_string('http://somesite.com')
print 'GET / HTTP/1.1\r\nCookie: {}\r\nUser-Agent: {}\r\n'.format(cookie_value, user_agent)
# GET / HTTP/1.1\r\n
# Cookie: cf_clearance=c8f913c707b818b47aa328d81cab57c349b1eee5-1426733163-3600; __cfduid=dd8ec03dfdbcb8c2ea63e920f1335c1001426733158
# User-Agent: Some/User-Agent String
Here is an example of integrating cloudscraper with curl. As you can see, all you have to do is pass the cookies and user-agent to curl.
import subprocess
import cloudscraper
# With get_tokens() cookie dict:
# tokens, user_agent = cloudscraper.get_tokens("http://somesite.com")
# cookie_arg = 'cf_clearance={}; __cfduid={}'.format(tokens['cf_clearance'], tokens['__cfduid'])
# With get_cookie_string() cookie header; recommended for curl and similar external applications:
cookie_arg, user_agent = cloudscraper.get_cookie_string('http://somesite.com')
# With a custom user-agent string you can optionally provide:
# ua = "Scraping Bot"
# cookie_arg, user_agent = cloudscraper.get_cookie_string("http://somesite.com", user_agent=ua)
result = subprocess.check_output(
[
'curl',
'--cookie',
cookie_arg,
'-A',
user_agent,
'http://somesite.com'
]
)
Trimmed down version. Prints page contents of any site protected with Cloudflare, via curl.
Warning: shell=True
can be dangerous to use with subprocess
in real code.
url = "http://somesite.com"
cookie_arg, user_agent = cloudscraper.get_cookie_string(url)
cmd = "curl --cookie {cookie_arg} -A {user_agent} {url}"
print(
subprocess.check_output(
cmd.format(
cookie_arg=cookie_arg,
user_agent=user_agent,
url=url
),
shell=True
)
)