raember/spoofbot

For browser spoofing when web crawling

Spoofbot

Web bot for spoofing browser behaviour when using python requests. Supports Firefox and Chrome browser on most generic Windows, MacOS and Linux spoofing.

Example usage

from spoofbot.browser import Chrome
from spoofbot.adapter.file import FileCache

browser = Chrome()
adapter = FileCache()
adapter.prepare_session(browser)

browser.navigate('https://httpbin.org/')
spec = browser.get('https://httpbin.org/spec.json').json()
print(spec['info']['description'])
# A simple HTTP Request & Response Service.<br/> <br/> <b>Run locally: </b> <code>$ docker run -p 80:80 kennethreitz/httpbin</code>
headers = browser.navigate(
    'https://httpbin.org/headers',
    headers={'Accept': 'mime/type'}
)[0].json()
print(headers['headers']['Accept'])
# mime/type

Browsers

spoofbot allows for useragents to be generate using information like the platform and browser version. Currently, only Firefox and Chrome are supported, but through inheritance one can add more browsers. The browser classes are derived from the requests.Session class and extend it by additional features.

Brotli encoding

Firefox indicates that brotli encoding (br) is acceptable, but that might lead to issues when parsing the responses. It is possible to change that default header:

from spoofbot.browser import Firefox

ff = Firefox()
ff._accept_encoding = ['deflate', 'gzip']  # brotli (br) is cumbersome

Request timeout for lower server load

The browsers have an automatic request delay built in which can be controlled with the request_timeout and honor_timeout. After requests have been made, did_wait and waiting_period provide further information.

Cache adapters

For more info refer to the adapter module.

File Cache

Using FileCache, one can store responses (without metadata such as headers or cookies) in the filesystem. The cache indicates whether the last made request got a cache hit. If there is a request that should be cached that cannot be adequately stored with only hostname and path, one can specify and alternative url to use instead of the request's prior to the request using the adapter's next_request_cache_url property. This is also supported when deleting the last request from the cache. By using the backup method, the cache will backup the subsequent requests' original cached responses inside a new Backup object. If it is then determined that the backup should be restored, the restore_all/restore methods can be used. The backup process can be stopped explicitly with stop_backup or by using a with block on the backup object.

HAR Cache

Using HarCache, one is able to load .har and MITMProxy flow files to use as cache. This cache does not make actual HTTP requests to the net, but fails if no matching request could be found. It can be specified whether matching a request should be strict (must match all headers) or not. When matching for requests, one can toggle rules to use (such as matching headers, header order or post data) when looking for a match using the adapter's properties.

Example usage

Please take a look at the tests. Take note that the loggers provide helpful data when testing from matches in the cache on the DEBUG level.

Tips

Turn off logging of other libraries:

import logging

logging.getLogger('urllib3.connectionpool').setLevel(logging.INFO)
logging.getLogger('chardet.charsetprober').setLevel(logging.INFO)
logging.getLogger('chardet.universaldetector').setLevel(logging.INFO)