PyChromeDevTools is a python module that allows one to interact with Google Chrome using Chrome DevTools Protocol within a Python script.
To use this tool, you must run an instance of Google Chrome with the remote-debugging
option, like in the following example.
google-chrome --remote-debugging-port=9222
You may want to enable further Chrome benchmarking capabilities using the --enable-benchmarking
and --enable-net-benchmarking
options. You can run Chrome in headless mode using the option --headless
.
For information about this Readme file and this tool please write to [email protected]
Very few dependencies must be satisfied: an updated Google-Chrome version and the python packages requests
and websocket
.
You can install them using the pip
tool.
You can install PyChromeDevTools
issuing the git command:
git clone https://github.com/marty90/PyChromeDevTools
Or, better, you install it and its dependencies by using pip
:
sudo pip3 install PyChromeDevTools
In your python script, as first, you must create a ChromeInterface object, like in the following:
chrome = PyChromeDevTools.ChromeInterface()
You can specify the host and the port of Chrome manually writing:
chrome = PyChromeDevTools.ChromeInterface(host="1.1.1.1",port=1234)
By default it uses localhost:9222
.
You can connect directly to a targetID
using the method connect_targetID
.
You must pass the targetID
as parameter to this function.
In this case, you avoid querying the json and finding the tab id
.
To send a command to Chrome, just invoke the corresponding method on the ChromeInterface object, and pass the desired parameters. For example, to visit a page write:
return_value, messages = chrome.Page.navigate(url="http://example.com/")
The return value of the command is passed as return value of the function, already interpreted as JSON. In messages, there are all the messages received before.
Chrome sends back messages for particular events in the browser. You can get them in three ways; they are returned already interpreted as JSON. All unread events are erased before any new command is run.
a) You can pop one message from the queue of received ones writing:
message=chrome.wait_message()
The method accepts an optional parameter timeout
which is the value in seconds after which it gives up and returns None
.
Default is 1.
b) You can wait for a specific event writing:
matching_event,all_events=chrome.wait_event(event_name)
It waits until an event with the name event_name
arrives, or a timeout elapses.
matching_event
contains the first found event that has event_name
, while all_events
contains all events arrived before.
Timeout value can be configured as in the previous method.
c) You can get all already received messages writing:
messages=chrome.pop_messages()
This method is not blocking, and, thus, no timeout is used.
import PyChromeDevTools
import time
chrome = PyChromeDevTools.ChromeInterface()
chrome.Network.enable()
chrome.Page.enable()
start_time=time.time()
chrome.Page.navigate(url="http://www.google.com/")
chrome.wait_event("Page.loadEventFired", timeout=60)
end_time=time.time()
print ("Page Loading Time:", end_time-start_time)
import PyChromeDevTools
import time
chrome = PyChromeDevTools.ChromeInterface()
chrome.Network.enable()
chrome.Page.enable()
chrome.Page.navigate(url="http://www.nytimes.com/")
chrome.wait_event("Page.frameStoppedLoading", timeout=60)
#Wait last objects to load
time.sleep(5)
cookies,messages = chrome.Network.getCookies()
for cookie in cookies["result"]["cookies"]:
print ("Cookie:")
print ("\tDomain:", cookie["domain"])
print ("\tKey:", cookie["name"])
print ("\tValue:", cookie["value"])
print ("\n")
import PyChromeDevTools
chrome = PyChromeDevTools.ChromeInterface()
chrome.Network.enable()
chrome.Page.enable()
chrome.Page.navigate(url="http://www.facebook.com")
event,messages=chrome.wait_event("Page.frameStoppedLoading", timeout=60)
for m in messages:
if "method" in m and m["method"] == "Network.responseReceived":
try:
url=m["params"]["response"]["url"]
print (url)
except:
pass