vivint-selenium-docker

Extending Selenium with drop in replacements for Chrome and Firefox webdrivers that run in Docker containers. Additional goodies like automatic proxies, live video recording and driver-pools are included!

Getting Started

Install the module:

Latest stable version from pypi,

$ pip install selenium-docker

Development version from source,

$ pip install git+ssh://git@github.com:vivint/selenium-council.git

Download docker for your operating system and ensure it's running.

$ docker version
Client:
 Version:      17.10.0-ce
 API version:  1.33

Server:
 Version:      17.10.0-ce
 API version:  1.33 (minimum version 1.12)

You should know...

Calling getLogger('selenium_docker').setLevel(logging.DEBUG) during Logging setup will turn on lots of debug statements involved with with spawning and managing the underlying containers and driver instances.
You can use the script below to stop and remove all running containers created by this library:
```
from selenium_docker.base import ContainerFactory

factory = ContainerFactory.get_default_factory()
factory.scrub_containers()
```
This will do a search in the default Docker engine for all containers that use our browser and dynamic labels.
We use gevent for its concurrency idioms.
We call gevent.monkey.patch_socket to communicate with Docker engine via REST. Other libraries may need to be patched contingent on what your project is trying to accomplish.

Read about monkey patching on the gevent website.

Examples

Basic

Creates a single container with a running Chrome Driver instance inside. Connecting and managing the container is all done automatically. This should function as a drop in replacement for using the desktop version of Chrome and Firefox drivers.

import sys
import logging

from selenium_docker import ChromeDriver

logging.basicConfig(stream=sys.stdout, level=logging.DEBUG)
logging.getLogger('selenium_docker').setLevel(logging.DEBUG)

driver = ChromeDriver()

driver.get('https://google.com')

print(driver.title)

driver.quit()

Blocking driver pool

Used for performing a single task on multiple sites/items in parallel.

The blocking driver pool will create all the necessary containers in advance in order to distribute the work as resources become available. Drivers will be reused until the .execute() call is complete. If the driver throws an Exception then that driver will be removed from the pool.

from selenium_docker.pool import DriverPool


def get_title(driver, url):
    driver.get(url)
    return driver.title

urls = [
    'https://google.com',
    'https://reddit.com',
    'https://yahoo.com',
    'http://ksl.com',
    'http://cnn.com'
]

pool = DriverPool(size=3)

for result in pool.execute(get_title, urls):
    print(result)

Asynchronous driver pool,

from selenium_docker.pool import DriverPool


def get_title(driver, url):
    driver.get(url)
    return driver.title

def print_fn(s):
    print(s)

urls = [
    'https://google.com',
    'https://reddit.com',
    'https://yahoo.com',
    'http://ksl.com',
    'http://cnn.com'
]


pool = DriverPool(size=2)
pool.execute_async(get_title, urls, print_fn)
pool.add_async(['https://facebook.com',
                'https://mail.com',
                'https://outlook.com'])

for x in pool.results():
    print('result - ', x)

    if '.com' in x:
        pool.add_async(['https://wikipedia.org'])

    if x == 'Wikipedia':
        pool.stop_async()
        pool.factory.scrub_containers()

License

Apache V2 -- See LICENSE for full statement.

vivint / selenium-docker