dateparser – python parser for human readable dates

travis build status pypi version Documentation Status Code Coverage Join the chat at https://gitter.im/scrapinghub/dateparser

dateparser provides modules to easily parse localized dates in almost any string formats commonly found on web pages.

Documentation

Documentation is built automatically and can be found on Read the Docs.

Features

  • Generic parsing of dates in English, Spanish, Dutch, Russian and over 20 other languages plus numerous formats in a language agnostic fashion.
  • Generic parsing of relative dates like: '1 min ago', '2 weeks ago', '3 months, 1 week and 1 day ago', 'in 2 days', 'tomorrow'.
  • Generic parsing of dates with time zones abbreviations or UTC offsets like: 'August 14, 2015 EST', 'July 4, 2013 PST', '21 July 2013 10:15 pm +0500'.
  • Support for non-Gregorian calendar systems. See Supported Calendars.
  • Extensive test coverage.

Usage

The most straightforward way is to use the dateparser.parse function, that wraps around most of the functionality in the module.

dateparser.parse(*args, **kwargs)[source]

Parse date and time from given date string.

Parameters:
  • date_string (str|unicode) – A string representing date and/or time in a recognizably valid format.
  • date_formats (list) – A list of format strings using directives as given here. The parser applies formats one by one, taking into account the detected languages.
  • languages (list) – A list of two letters language codes.e.g. [‘en’, ‘es’]. If languages are given, it will not attempt to detect the language.
  • settings (dict) – Configure customized behavior using settings defined in dateparser.conf.Settings.
Returns:

Returns datetime representing parsed date if successful, else returns None

Return type:

datetime.

Raises:

ValueError - Unknown Language

Relative Dates

>>> parse('1 hour ago')
datetime.datetime(2015, 5, 31, 23, 0)
>>> parse(u'Il ya 2 heures')  # French (2 hours ago)
datetime.datetime(2015, 5, 31, 22, 0)
>>> parse(u'1 anno 2 mesi')  # Italian (1 year 2 months)
datetime.datetime(2014, 4, 1, 0, 0)
>>> parse(u'yaklaşık 23 saat önce')  # Turkish (23 hours ago)
datetime.datetime(2015, 5, 31, 1, 0)
>>> parse(u'Hace una semana')  # Spanish (a week ago)
datetime.datetime(2015, 5, 25, 0, 0)
>>> parse(u'2小时前')  # Chinese (2 hours ago)
datetime.datetime(2015, 5, 31, 22, 0)

Note

Testing above code might return different values for you depending on your environment’s current date and time.

Note

Support for relative dates in future needs a lot of improvement, we look forward to community’s contribution to get better on that part. See `Contributing`_.

OOTB Language Based Date Order Preference

>>> # parsing ambiguous date
>>> parse('02-03-2016')  # assumes english language, uses MDY date order
datetime.datetime(2016, 3, 2, 0, 0)
>>> parse('le 02-03-2016')  # detects french, uses DMY date order
datetime.datetime(2016, 3, 2, 0, 0)

Note

Ordering is not locale based, that’s why do not expect DMY order for UK/Australia English. You can specify date order in that case as follows usings Settings:

>>> parse('18-12-15 06:00', settings={'DATE_ORDER': 'DMY'})
datetime.datetime(2015, 12, 18, 6, 0)

For more on date order, please look at Settings.

Timezone and UTC Offset

Dateparser assumes all dates to be in UTC if no timezone is specified in the string. To convert the resultant datetime object to required timezone. You can do the following:

>>> parse('January 12, 2012 10:00 PM', settings={'TIMEZONE': 'US/Eastern'})
datetime.datetime(2012, 1, 12, 17, 0)

Support for tzaware objects:

>>> parse('12 Feb 2015 10:56 PM EST', settings={'RETURN_AS_TIMEZONE_AWARE': True})
datetime.datetime(2015, 2, 13, 3, 56, tzinfo=<StaticTzInfo 'UTC'>)
>>> parse('12 Feb 2015 10:56 PM EST', settings={'RETURN_AS_TIMEZONE_AWARE': True, 'TIMEZONE': None})
datetime.datetime(2015, 2, 12, 22, 56, tzinfo=<StaticTzInfo 'EST'>)

For more on timezones, please look at Settings.

Incomplete Dates

>>> from dateparser import parse
>>> parse(u'December 2015')  # default behavior
datetime.datetime(2015, 12, 16, 0, 0)
>>> parse(u'December 2015', settings={'PREFER_DAY_OF_MONTH': 'last'})
datetime.datetime(2015, 12, 31, 0, 0)
>>> parse(u'December 2015', settings={'PREFER_DAY_OF_MONTH': 'first'})
datetime.datetime(2015, 12, 1, 0, 0)
>>> parse(u'March')
datetime.datetime(2015, 3, 16, 0, 0)
>>> parse(u'March', settings={'PREFER_DATES_FROM': 'future'})
datetime.datetime(2016, 3, 16, 0, 0)
>>> # parsing with preference set for 'past'
>>> parse('August', settings={'PREFER_DATES_FROM': 'past'})
datetime.datetime(2015, 8, 15, 0, 0)

You can also ignore parsing incomplete dates altogether by setting STRICT_PARSING flag as follows:

>>> parse(u'December 2015', settings={'STRICT_PARSING': True})
None

For more on handling incomplete dates, please look at Settings.

Dependencies

dateparser relies on following libraries in some ways:

  • dateutil‘s module relativedelta for its freshness parser.
  • ruamel.yaml for reading language and configuration files.
  • jdatetime to convert Jalali dates to Gregorian.
  • umalqurra to convert Hijri dates to Gregorian.

Supported languages

  • Arabic
  • Belarusian
  • Chinese
  • Czech
  • Danish
  • Dutch
  • English
  • Filipino/Tagalog
  • Finnish
  • French
  • Hungarian
  • German
  • Indonesian
  • Italian
  • Japanese
  • Persian
  • Polish
  • Portuguese
  • Romanian
  • Russian
  • Spanish
  • Thai
  • Turkish
  • Ukrainian
  • Vietnamese

Supported Calendars

  • Gregorian calendar.

  • Persian Jalali calendar. For more information, refer to Persian Jalali Calendar.

  • Hijri/Islamic Calendar. For more information, refer to Hijri Calendar.

    >>> from dateparser.calendars.jalali import JalaliCalendar
    >>> JalaliCalendar(u'جمعه سی ام اسفند ۱۳۸۷').get_date()
    {'date_obj': datetime.datetime(2009, 3, 20, 0, 0), 'period': 'day'}
    
    >>> from dateparser.calendars.hijri import HijriCalendar
    >>> HijriCalendar(u'17-01-1437 هـ 08:30 مساءً').get_date()
    {'date_obj': datetime.datetime(2015, 10, 30, 20, 30), 'period': 'day'}
    

Note

HijriCalendar has some limitations with Python 3.

Note

For Finnish language, please specify settings={‘SKIP_TOKENS’: []} to correctly parse freshness dates.

Using DateDataParser

dateparser.parse() uses a default parser which tries to detect language every time it is called and is not the most efficient way while parsing dates from the same source.

DateDataParser provides an alternate and efficient way to control language detection behavior.

The instance of DateDataParser reduces the number of applicable languages, until only one or no language is left. It assumes the previously detected language for all the subsequent dates supplied.

This class wraps around the core dateparser functionality, and by default assumes that all of the dates fed to it are in the same language.

class dateparser.date.DateDataParser(*args, **kwargs)[source]

Class which handles language detection, translation and subsequent generic parsing of string representing date and/or time.

Parameters:
  • languages (list) – A list of two letters language codes, e.g. [‘en’, ‘es’]. If languages are given, it will not attempt to detect the language.
  • allow_redetect_language (bool) – Enables/disables language re-detection.
  • settings (dict) – Configure customized behavior using settings defined in dateparser.conf.Settings.
Returns:

A parser instance

Raises:

ValueError - Unknown Language, TypeError - Languages argument must be a list

get_date_data(date_string, date_formats=None)[source]

Parse string representing date and/or time in recognizable localized formats. Supports parsing multiple languages and timezones.

Parameters:
  • date_string (str|unicode) – A string representing date and/or time in a recognizably valid format.
  • date_formats (list) –

    A list of format strings using directives as given here. The parser applies formats one by one, taking into account the detected languages.

Returns:

a dict mapping keys to datetime.datetime object and period. For example: {‘date_obj’: datetime.datetime(2015, 6, 1, 0, 0), ‘period’: u’day’}

Raises:

ValueError - Unknown Language

Note

Period values can be a ‘day’ (default), ‘week’, ‘month’, ‘year’.

Period represents the granularity of date parsed from the given string.

In the example below, since no day information is present, the day is assumed to be current day 16 from current date (which is June 16, 2015, at the moment of writing this). Hence, the level of precision is month:

>>> DateDataParser().get_date_data(u'March 2015')
{'date_obj': datetime.datetime(2015, 3, 16, 0, 0), 'period': u'month'}

Similarly, for date strings with no day and month information present, level of precision is year and day 16 and month 6 are from current_date.

>>> DateDataParser().get_date_data(u'2014')
{'date_obj': datetime.datetime(2014, 6, 16, 0, 0), 'period': u'year'}

Dates with time zone indications or UTC offsets are returned in UTC time unless specified using Settings.

>>> DateDataParser().get_date_data(u'23 March 2000, 1:21 PM CET')
{'date_obj': datetime.datetime(2000, 3, 23, 14, 21), 'period': 'day'}

Warning

It fails to parse English dates in the example below, because Spanish was detected and stored with the ddp instance:

>>> ddp.get_date_data('11 August 2012')
{'date_obj': None, 'period': 'day'}

dateparser.date.DateDataParser can also be initialized with known languages:

>>> ddp = DateDataParser(languages=['de', 'nl'])
>>> ddp.get_date_data(u'vr jan 24, 2014 12:49')
{'date_obj': datetime.datetime(2014, 1, 24, 12, 49), 'period': u'day'}
>>> ddp.get_date_data(u'18.10.14 um 22:56 Uhr')
{'date_obj': datetime.datetime(2014, 10, 18, 22, 56), 'period': u'day'}

Settings

dateparser‘s parsing behavior can be configured by supplying settings as a dictionary to settings argument in dateparser.parse or DateDataParser constructor.

All supported settings with their usage examples are given below:

DATE_ORDER specifies the order in which date components year, month and day are expected while parsing ambiguous dates. It defaults to MDY which translates to month first, day second and year last order. Characters M, D or Y can be shuffled to meet required order. For example, DMY specifies day first, month second and year last order:

>>> parse('15-12-18 06:00')  # assumes default order: MDY
datetime.datetime(2018, 12, 15, 6, 0)  # since 15 is not a valid value for Month, it is swapped with Day's
>>> parse('15-12-18 06:00', settings={'DATE_ORDER': 'YMD'})
datetime.datetime(2015, 12, 18, 6, 0)

PREFER_LANGUAGE_DATE_ORDER defaults to True. Most languages have a default DATE_ORDER specified for them. For example, for French it is DMY:

>>> # parsing ambiguous date
>>> parse('02-03-2016')  # assumes english language, uses MDY date order
datetime.datetime(2016, 3, 2, 0, 0)
>>> parse('le 02-03-2016')  # detects french, hence, uses DMY date order
datetime.datetime(2016, 3, 2, 0, 0)

Note

There’s no language level default DATE_ORDER associated with en language. That’s why it assumes MDY which is :obj:settings <dateparser.conf.settings> default. If the language has a default DATE_ORDER associated, supplying custom date order will not be applied unless we set PREFER_LANGUAGE_DATE_ORDER to False:

>>> parse('le 02-03-2016', settings={'DATE_ORDER': 'MDY'})
datetime.datetime(2016, 3, 2, 0, 0)  # MDY didn't apply
>>> parse('le 02-03-2016', settings={'DATE_ORDER': 'MDY', 'PREFER_LANGUAGE_DATE_ORDER': False})
datetime.datetime(2016, 2, 3, 0, 0)  # MDY worked!

TIMEZONE defaults to UTC. All dates, complete or relative, are assumed to be in UTC. When specified, resultant datetime converts according to the supplied timezone:

>>> parse('January 12, 2012 10:00 PM')
datetime.datetime(2012, 1, 12, 22, 0)
>>> parse('January 12, 2012 10:00 PM', settings={'TIMEZONE': 'US/Eastern'})
datetime.datetime(2012, 1, 12, 17, 0)

RETURN_AS_TIMEZONE_AWARE is a flag to turn on timezone aware dates if timezone is detected or specified.:

>>> parse('12 Feb 2015 10:56 PM EST', settings={'RETURN_AS_TIMEZONE_AWARE': True})
datetime.datetime(2015, 2, 13, 3, 56, tzinfo=<StaticTzInfo 'UTC'>)
>>> parse('12 Feb 2015 10:56 PM EST', settings={'RETURN_AS_TIMEZONE_AWARE': True, 'TIMEZONE': 'EST'})
datetime.datetime(2015, 2, 12, 22, 56, tzinfo=<StaticTzInfo 'EST'>)

if TIMEZONE is set to None and RETURN_AS_TIMEZONE_AWARE to True, a tz aware date will only be returned if timezone is detected in the string.

>>> parse('12 Feb 2015 10:56 PM', settings={'RETURN_AS_TIMEZONE_AWARE': True, 'TIMEZONE': None})
datetime.datetime(2015, 2, 12, 22, 56)
>>> parse('12 Feb 2015 10:56 PM EST', settings={'RETURN_AS_TIMEZONE_AWARE': True, 'TIMEZONE': None})
datetime.datetime(2015, 2, 12, 22, 56, tzinfo=<StaticTzInfo 'EST'>)

PREFER_DAY_OF_MONTH This option comes handy when the date string is missing the day part. It defaults to current and can have first and last denoting first and last day of months respectively as values:

>>> from dateparser import parse
>>> parse(u'December 2015')  # default behavior
datetime.datetime(2015, 12, 16, 0, 0)
>>> parse(u'December 2015', settings={'PREFER_DAY_OF_MONTH': 'last'})
datetime.datetime(2015, 12, 31, 0, 0)
>>> parse(u'December 2015', settings={'PREFER_DAY_OF_MONTH': 'first'})
datetime.datetime(2015, 12, 1, 0, 0)

PREFER_DATES_FROM defaults to current_period and can have past and future as values.

If date string is missing some part, this option ensures consistent results depending on the past or future preference, for example, assuming current date is June 16, 2015:

>>> from dateparser import parse
>>> parse(u'March')
datetime.datetime(2015, 3, 16, 0, 0)
>>> parse(u'March', settings={'PREFER_DATES_FROM': 'future'})
datetime.datetime(2016, 3, 16, 0, 0)
>>> # parsing with preference set for 'past'
>>> parse('August', settings={'PREFER_DATES_FROM': 'past'})
datetime.datetime(2015, 8, 15, 0, 0)

STRICT_PARSING defaults to False.

When set to True if missing any of day, month or year parts, it does not return any result altogether.:

>>> parse(u'March', settings={'STRICT_PARSING': True})
None

SKIP_TOKENS is a list of tokens to discard while detecting language. Defaults to ['t'] which skips T in iso format datetime string .e.g. 2015-05-02T10:20:19+0000.:

>>> from dateparser.date import DateDataParser
>>> DateDataParser(settings={'SKIP_TOKENS': ['de']}).get_date_data(u'27 Haziran 1981 de')  # Turkish (at 27 June 1981)
{'date_obj': datetime.datetime(1981, 6, 27, 0, 0), 'period': 'day'}

Indices and tables