This is a python3 fork of Richard Penman's Browser Cookie
- What does it do? Loads cookies used by your web browser into a cookiejar object.
- Why is it useful? This means you can use python to download and get the same content you see in the web browser without needing to login.
- Which browsers are supported? Currently Chrome, Firefox, Opera, Edge, Chromium, and Brave.
- How are the cookies stored? All currently-supported browsers store cookies in a sqlite database in your home directory.
#!bash
pip3 install browser-cookie3
Here is a dangerous hack to extract the title from a webpage:
#!python
>>> import re
>>> get_title = lambda html: re.findall('<title>(.*?)</title>', html, flags=re.DOTALL)[0].strip()
And here is the webpage title when downloaded normally:
#!python
>>> import urllib2
>>> url = 'https://bitbucket.org/'
>>> public_html = urllib2.urlopen(url).read()
>>> get_title(public_html)
'Git and Mercurial code management for teams'
Now let's try with browser_cookie3 - make sure you are logged into Bitbucket in Firefox before trying this example:
#!python
>>> import browser_cookie3
>>> cj = browser_cookie3.firefox()
>>> opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj))
>>> login_html = opener.open(url).read()
>>> get_title(login_html)
'richardpenman / home — Bitbucket'
You should see your own username here, meaning the module successfully loaded the cookies from Firefox.
Here is an alternative example with requests, this time loading the Chrome cookies. Again make sure you are logged into Bitbucket in Chrome before running this:
#!python
>>> import browser_cookie3
>>> import requests
>>> cj = browser_cookie3.chrome()
>>> r = requests.get(url, cookies=cj)
>>> get_title(r.content)
'richardpenman / home — Bitbucket'
Alternatively if you don't know/care which browser has the cookies you want then all available browser cookies can be loaded:
#!python
>>> import browser_cookie3
>>> import requests
>>> cj = browser_cookie3.load()
>>> r = requests.get(url, cookies=cj)
>>> get_title(r.content)
'richardpenman / home — Bitbucket'
Alternatively if you are only interested in cookies from a specific domain, you can specify a domain filter.
#!python
>>> import browser_cookie3
>>> import requests
>>> cj = browser_cookie3.chrome(domain_name='www.bitbucket.com')
>>> r = requests.get(url, cookies=cj)
>>> get_title(r.content)
'richardpenman / home — Bitbucket'
Creating and testing a fresh cookie file can help eliminate some possible user specific issues. It also allows you to upload a cookie file you are having issus with, since you should never upload your main cookie file!
For linux and assumably mac:
Run google-chrome-stable --user-data-dir=browser_cookie3 #replace google-chrome-stable with your command to start chrome/chromium
and when you close the browser you will have a new cookie file at browser_cookie3/Default/Cookies
If you want to share a cookie file then visit some site that will generate cookie (without logging in!), example https://www.theverge.com/ will save cookies after you accept the GDPR notice.
If browser_cookie3
is unable to decrypt cookies, you can start your browser with the --password-store=gnome-keyring
parameter. For example:
/usr/bin/brave-browser-stable --password-store=gnome-keyring %U
This is confirmed to work with the Brave browser but may also work with other Chromium based browsers as well.
- more sensible cookie file checking order, like first using the default defined in profiles.ini for firefox
So far the following platforms are supported:
- Chrome: Linux, OSX, Windows
- Firefox: Linux, OSX, Windows
- Opera: Linux, OSX, Windows
- Edge: Linux, OSX, Windows
- Chromium: Linux, OSX, Windows
- Brave Linux (using Gnome keyring), OSX, Windows
OS | Chrome | Firefox | Opera | Edge | Chromium | Brave |
---|---|---|---|---|---|---|
Mac | 09/12/20 | 09/12/20 | 09/12/20 | 09/12/20 | 09/12/20 | |
Linux | 09/12/20 | 09/12/20 | 09/12/20 | 09/12/20 | 09/12/20 | 07/24/21 |
Windows | 09/12/20 | 09/12/20 | 09/12/20 | 09/12/20 | 09/12/20 |
However I only tested on a single version of each browser and so am not sure if the cookie sqlite format changes location or format in earlier/later versions. If you experience a problem please open an issue which includes details of the browser version and operating system. Also patches to support other browsers are very welcome, particularly for Chrome and Internet Explorer on Windows.
Special thanks to Nathan Henrie for his example of how to decode the Chrome cookies.