Skip to main content

Continuous set support for Python

Project description

Spans

Test status Test coverage Documentation status PyPI version status Python version MIT License

Spans is a pure Python implementation of PostgreSQL’s range types. Range types are convenient when working with intervals of any kind. Every time you’ve found yourself working with date_start and date_end, an interval may have been what you were actually looking for.

Spans has successfully been used in production since its first release 30th August, 2013.

Installation

Spans exists on PyPI.

$ pip install Spans

Documentation is hosted on Read the Docs.

Example

Imagine you are building a calendar and want to display all weeks that overlaps the current month. Normally you have to do some date trickery to achieve this, since the month’s bounds may be any day of the week. With Spans’ set-like operations and shortcuts the problem becomes a breeze.

We start by importing date and daterange

>>> from datetime import date
>>> from spans import daterange

Using daterange.from_month we can get range representing January in the year 2000

>>> month = daterange.from_month(2000, 1)
>>> month
daterange(datetime.date(2000, 1, 1), datetime.date(2000, 2, 1))

Now we can calculate the ranges for the weeks where the first and last day of month are

>>> start_week = daterange.from_date(month.lower, period="week")
>>> end_week = daterange.from_date(month.last, period="week")
>>> start_week
daterange(datetime.date(1999, 12, 27), datetime.date(2000, 1, 3))
>>> end_week
daterange(datetime.date(2000, 1, 31), datetime.date(2000, 2, 7))

Using a union we can express the calendar view.

>>> start_week.union(month).union(end_week)
daterange(datetime.date(1999, 12, 27), datetime.date(2000, 2, 7))

Do you want to know more? Head over to the documentation.

Use with Psycopg2

To use these range types with Psycopg2 the PsycoSpans.

Motivation

For a project of mine I started using PostgreSQL’s tsrange type and needed an equivalent in Python. These range types attempt to mimick PostgreSQL’s behavior in every way. Deviating from it is considered as a bug and should be reported.

Contribute

I appreciate all the help I can get! Some things to think about:

  • If it’s a simple fix, such as documentation or trivial bug fix, please file an issue or submit a pull request. Make sure to only touch lines relevant to the issue. I don’t accept pull requests that simply reformat the code to be PEP8-compliant. To me the history of the repository is more important.

  • If it’s a feature request or a non-trivial bug, always open an issue first to discuss the matter. It would be a shame if good work went to waste because a pull request doesn’t fit the scope of this project.

Pull requests are credited in the change log which is displayed on PyPI and the documentaion on Read the Docs.

Changelog

Version are structured like the following: <major>.<minor>.<bugfix>. The first 0.1 release does not properly adhere to this. Unless explicitly stated, changes are made by Andreas Runfalk.

Version 1.1.1

Released on 21st March, 2021

  • Normalize ranges to be empty when start and end is the same and either bound is exclusive (bug #18, lgharibashvili)

Version 1.1.0

Released on 2nd June, 2019

This release changes a lot of internal implementation details that should prevent methods from not handling unbounded ranges correctly in the future.

  • Added validation to ensure unbounded ranges are never inclusive

  • Changed __repr__ for ranges to be more similar to proper Python syntax. The old representation looked like mismatched parentheses. For instance floatrange((,10.0]) becomes floatrange(upper=10.0, upper_inc=True)

  • Dropped Python 3.3 support since it’s been EOL for almost two years. It probably still works but it is no longer tested

  • Fixed pickling of empty range sets not working (bug #14)

  • Fixed union() not working properly with unbounded ranges

  • Fixed lowerly unbounded ranges improperly being lower inclusive

  • Fixed startswith() and endsbefore() being not handling empty ranges

Version 1.0.2

Released on 22th February, 2019

Version 1.0.1

Released on 31st January, 2018

  • Fixed PartialOrderingMixin not using __slots__ (bug #10)

Version 1.0.0

Released on 8th June, 2017

  • Added NotImplemented for << and >> operators when there is a type mismatch

  • Added | operator for unions of Range and NotImplemented support for RangeSet

  • Added & operator for intersections of Range and NotImplemented support for RangeSet

  • Added - operator for differences of Range and NotImplemented support for RangeSet

  • Added reversed() iterator support for DiscreteRange

  • Fixed overlap with empty range incorrectly returns True (bug #7)

  • Fixed issue with contains() for scalars on unbounded ranges

  • Fixed type check for right_of()

  • Fixed type check for contains()

  • Fixed type check for union()

  • Fixed type check for intersection()

  • Fixed type check for difference()

  • Fixed infinite iterators not being supported for DiscreteRange

Version 0.5.0

Released on 16th April, 2017

This release is a preparation for a stable 1.0 release.

  • Fixed comparison operators when working with empty or unbounded ranges. They would previously raise exceptions. Ranges are now partially ordered instead of totally ordered

  • Added more unit tests

  • Renamed classes to match PEP8 conventions. This does not apply to classes that works on built-in that does not follow PEP8.

  • Refactored left_of()

  • Refactored overlap()

  • Refactored union()

Version 0.4.0

Released on 20th March, 2017

This release is called 0.4.1 on PyPI because I messed up the upload.

  • Added new argument to from_date() for working with different kinds of date intervals. The argument accepts a period of either "day" (default), "week" (ISO week), "american_week" (starts on sunday), "month", "quarter" or "year".

  • Added new methods to daterange for working with different kinds of date intervals: from_week(), from_month(), from_quarter() and from_year().

  • Added a new class PeriodRange for working with periods like weeks, months, quarters or years. It inherits all methods from daterange and is aware of its own period type. It allows things like getting the previous or next week.

  • Fixed daterange not accepting subclasses of date (bug #5)

  • Fixed some broken doctests

  • Moved unit tests to pytest

  • Removed Tox config

  • Minor documentation tweaks

Version 0.3.0

Released on 26th August, 2016

  • Added documentation for __iter__()

  • Fixed intersection of multiple range sets not working correctly (bug #3)

  • Fixed iteration of RangeSet returning an empty range when RangeSet is empty (bug #4)

Version 0.2.1

Released on 27th June, 2016

  • Fixed RangeSet not returning NotImplemented when comparing to classes that are not sub classes of RangeSet, pull request #2 (Michael Krahe)

  • Updated license in setup.py to follow recommendations by PyPA

Version 0.2.0

Released on 22nd December, 2015

  • Added __len__() to range sets (Michael Krahe)

  • Added contains() to range sets (Michael Krahe)

  • Added Sphinx style doc strings to all methods

  • Added proper Sphinx documentation

  • Added unit tests for uncovered parts, mostly error checking

  • Added wheel to PyPI along with source distribution

  • Fixed a potential bug where comparing ranges of different types would result in an infinite loop

  • Changed meta class implementation for range sets to allow more mixins for custom range sets

Version 0.1.4

Released on 15th May, 2015

  • Added .last property to DiscreteRange

  • Added from_date() helper to daterange

  • Added more unit tests

  • Improved pickle implementation

  • Made type checking more strict for date ranges to prevent datetime from being allowed in daterange

Version 0.1.3

Released on 27th February, 2015

  • Added offset() to some range types

  • Added offset() to some range set types

  • Added sanity checks to range boundaries

  • Fixed incorrect __slots__ usage, resulting in __slots__ not being used on most ranges

  • Fixed pickling of ranges and range sets

  • Simplified creation of new range sets, by the use of the meta class MetaRangeSet

Version 0.1.2

Released on 13th June, 2014

  • Fix for inproper version detection on Ubuntu’s bundled Python interpreter

Version 0.1.1

Released on 12th June, 2014

  • Readme fixes

  • Syntax highlighting for PyPI page

Version 0.1.0

Released on 30th August, 2013

  • Initial release

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

Spans-1.1.1.tar.gz (45.8 kB view details)

Uploaded Source

Built Distribution

Spans-1.1.1-py2.py3-none-any.whl (24.3 kB view details)

Uploaded Python 2 Python 3

File details

Details for the file Spans-1.1.1.tar.gz.

File metadata

  • Download URL: Spans-1.1.1.tar.gz
  • Upload date:
  • Size: 45.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.1 importlib_metadata/4.0.1 pkginfo/1.7.0 requests/2.24.0 requests-toolbelt/0.9.1 tqdm/4.60.0 CPython/3.9.2

File hashes

Hashes for Spans-1.1.1.tar.gz
Algorithm Hash digest
SHA256 7f3e94492ba131d5ff062a5d631424e613fbeb38bb7ea613d8fd11f6e755fd79
MD5 d4f202420cc099b30dd3a5851a9f3a5b
BLAKE2b-256 03ec0c9cc8cc1f533de3d1ffd4baad9d9ea0dccbb2c692d1425a18f4a518ffeb

See more details on using hashes here.

File details

Details for the file Spans-1.1.1-py2.py3-none-any.whl.

File metadata

  • Download URL: Spans-1.1.1-py2.py3-none-any.whl
  • Upload date:
  • Size: 24.3 kB
  • Tags: Python 2, Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.1 importlib_metadata/4.0.1 pkginfo/1.7.0 requests/2.24.0 requests-toolbelt/0.9.1 tqdm/4.60.0 CPython/3.9.2

File hashes

Hashes for Spans-1.1.1-py2.py3-none-any.whl
Algorithm Hash digest
SHA256 2d64ea863ceb2cce77f93fc78cf4f68759828bcef0cd82087cfa22f09f47652a
MD5 21b2d7ff541d5adf107fdc0bc4260f9e
BLAKE2b-256 a77572360570ca7ec368333e45eabffacd9123db6aa4af830336b7ed30c8a469

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page