URLExtract is python class for collecting URLs from given text based on TLD.
Clone or download
lipoja Better parsing URL surrounded with parenthesis
* devel:
  Splitting enclosure character in better way
  updated parsing of URL surrounded with parenthesis (issue #23)
Latest commit b2e1626 Oct 10, 2018

README.rst

URLExtract

URLExtract is python class for collecting (extracting) URLs from given text based on locating TLD.

Build Status Git tag Python Version Compatibility

How does it work

It tries to find any occurrence of TLD in given text. If TLD is found it starts from that position to expand boundaries to both sides searching for "stop character" (usually whitespace, comma, single or double quote).

NOTE: List of TLDs is downloaded from iana.org to keep you up to date with new TLDs.

Installation

Package is available on PyPI - you can install it via pip.

pip install urlextract

Documentation

Online documentation is published at http://urlextract.readthedocs.io/

Requirements

  • IDNA for converting links to IDNA format

  • uritools for domain name validation

    pip install idna
    pip install uritools
    

Example

You can look at command line program at the end of urlextract.py. But everything you need to know is this:

from urlextract import URLExtract

extractor = URLExtract()
urls = extractor.find_urls("Text with URLs. Let's have URL janlipovsky.cz as an example.")
print(urls) # prints: ['janlipovsky.cz']

Or you can get generator over URLs in text by:

from urlextract import URLExtract

extractor = URLExtract()
example_text = "Text with URLs. Let's have URL janlipovsky.cz as an example."

for url in extractor.gen_urls(example_text):
    print(url) # prints: ['janlipovsky.cz']

Or you if you want to just check if there is at least one URL you can do:

from urlextract import URLExtract

extractor = URLExtract()
example_text = "Text with URLs. Let's have URL janlipovsky.cz as an example."

if extractor.has_urls(example_text):
    print("Given text contains some URL")

Known issues

Since TLD can be not only shortcut but also some meaningful word we might see "false matches" when we are searching for URL in some HTML pages. The false match can occur for example in css or JS when you are referring to HTML item using its classes.

Example HTML code:

<p class="bold name">Jan</p>
<style>
  p.bold.name {
    font-weight: bold;
  }
</style>

If this HTML snippet is on the input of urlextract.find_urls() it will return p.bold.name as an URL. Behavior of urlextract is correct, because .name is valid TLD and urlextract just see that there is bold.name valid domain name and p is valid sub-domain.

License

This piece of code is licensed under The MIT License.