Skip to content
Find the creation date of web pages using common structural patterns, text-based heuristics and robust date extraction
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


htmldate: find the publication date of web pages

Issue tracker:

In a nutshell, with Python:

>>> from htmldate import find_date
>>> find_date('')
>>> find_date('', original_date=True)

On the command-line:

$ htmldate -u


htmldate finds original and updated publication dates of web pages using a combination of tree traversal, common structural patterns, text-based heuristics and robust date extraction. All the steps needed from web page download to HTML parsing, including scraping and textual analysis are handled. URLs, HTML files or HTML trees are given as input, the library outputs a date string in the desired format.

htmldate provides following ways to date a HTML document:

  1. Markup in header: common patterns are used to identify relevant elements (e.g. link and meta elements) including Open Graph protocol attributes and a large number of CMS idiosyncracies
  2. HTML code: The whole document is then searched for structural markers: abbr/time elements and a series of attributes (e.g. postmetadata)
  3. Bare HTML content: A series of heuristics is run on text and markup:
  1. in fast mode the HTML page is cleaned and precise patterns are targeted
  2. in extensive mode date expressions are collected and the best one is chosen based on a disambiguation algorithm

The module then returns a date if a valid cue could be found in the document, per default the updated date w.r.t. the original publishing statement. The output string defaults to ISO 8601 YMD format.

  • Should be compatible with all common versions of Python 3 (see tests and coverage)
  • Safety belt included, the output is thouroughly verified with respect to its plausibility and adequateness
  • Designed to be computationally efficient and used in production on millions of documents
  • Handles batch processing of a list of URLs
  • Switch between original and updated date

The library currently focuses on texts written in English or German.


Install from package repository: pip install htmldate

For the latest version (check build status): pip install git+

Version 0.5.3 is the last to support Python 3.4, later versions are 3.5+ compatible.

With Python

>>> from htmldate import find_date
>>> find_date('')

The module can resort to a guess based on a complete screning of the document (extensive_search parameter) which can be deactivated:

>>> find_date('')
'2017-08-11' # has been updated since
>>> find_date('', extensive_search=False)

Input format

The module expects strings as shown above, it is also possible to use already parsed HTML (i.e. a LXML tree object):

# simple HTML document as string
>>> htmldoc = '<html><body><span class="entry-date">July 12th, 2016</span></body></html>'
>>> find_date(mytree)
# parsed LXML tree
>>> from lxml import html
>>> mytree = html.fromstring('<html><body><span class="entry-date">July 12th, 2016</span></body></html>')
>>> find_date(mytree)

Date format

The output format of the dates found can be set in a format known to Python's datetime module, the default being %Y-%m-%d:

>>> find_date('', outputformat='%d %B %Y')
'18 November 2016' # may have changed since

Original date

Although the time delta between the original publication and the last modified statement is usually a matter of hours or days at most, it can be useful in some contexts to prioritize the original publication date during extraction:

>>> find_date('') # default setting
>>> find_date('', original_date=True) # modified behavior

On the command-line

A basic command-line interface is included:

$ htmldate -u

For usage instructions see htmldate -h:

The batch mode -i takes one URL per line as input and returns one result per line in tab-separated format:

$ htmldate -i list-of-urls.txt

Additional information

Going further

For more details check the online documentation:

If the date is nowhere to be found, it might be worth considering carbon dating the web page, however this is computationally expensive. In addition, datefinder features pattern-based date extraction for texts written in English.

Pull requests are welcome.


This module is part of methods to derive metadata from web documents in order to build text corpora for computational linguistic and NLP analysis, the original problem being that there are web pages for which neither the URL nor the server response provide a reliable way to date the document, i.e. find when it was first published and/or last modified. For more information:

Kudos to...


See my contact page for details.

You can’t perform that action at this time.