extract-social-media 0.4.0

Extract social media links from websites

Extract Social Media

Extract social media links from websites.

Many websites reference their facebook, twitter, linkedin, youtube accounts and these can be invaluable to gather 360 degree information about a company.

This library allows to extract links or handles for the most commonly used international social media networks.

• Free software: MIT license

• Python versions: 2.7, 3.4+

Features

• Extract social media links/handles from html content

• Attempts to extract links/handles also from widgets, scripts, etc.

• Supports most widely used social networks

  • facebook

  • linkedin

  • twitter

  • youtube

  • github

  • google plus

  • pinterest

  • instagram

  • snapchat

  • flipboard

  • flickr

  • weibo

  • periscope

  • telegram

  • soundcloud

  • feedburner

  • vimeo

  • slideshare

  • vkontakte

  • xing

Quickstart

import requests
from html_to_etree import parse_html_bytes
res = requests.get('https://techcrunch.com/contact/')
tree = parse_html_bytes(res.content, res.headers.get('content-type'))

set(find_links_tree(tree))

{'http://pinterest.com/techcrunch/',
 'http://www.youtube.com/user/techcrunch',
 'http://www.linkedin.com/company/techcrunch',
 'https://www.facebook.com/techcrunch',
 'https://flipboard.com/@techcrunch',
 'http://instagram.com/techcrunch',
 'https://plus.google.com/+TechCrunch',
 'https://instagram.com/techcrunch',
 'https://twitter.com/techcrunch'}

Caveats

• currently finds all social media links on a page

  • need to look into finding most relevant links based on link location, link context, company name, etc.

Credits

This package was created with Cookiecutter and the fluquid/cookiecutter-pypackage project template.

History

0.4.0 (2017-08-18)

• naive blacklisting for photos, videos, search, tweets, etc.

0.3.0 (2017-08-18)

• fixed exception when “href” is empty or non-string

0.2.0 (2017-06-08)

• better test coverage

• accepting data-href

0.1.0 (unreleased)

• First release on PyPI.