Tempo

This project is a scheduling library, written in Python, that provides a generic way to compose and query schedules of continuous recurrent events, such as working time of organizations, meetings, movie shows, etc.

It contains a Python implementation and bindings for PostgreSQL, Django and Django REST Framework.

Links

PyPI:	https://pypi.python.org/pypi/python-tempo
Documentation:	https://python-tempo.readthedocs.org/
Issues:	https://github.com/AndrewPashkin/python-tempo/issues/
Code:	https://github.com/AndrewPashkin/python-tempo/

Features

• Flexible schedule model, that can express schedules, that other libraries can't.
• Queries: containment of a single timestamp, future occurrences.
• Bindings:
  • PostgreSQL
    • Domain type for storing schedules
    • Procedures for performing tests on them (timestamp containment, future occurrences).
  • Django
    • Model field
    • Django-Admin widget
    • Custom lookups (timestamp containment, intersection with interval between two timestamps, test if scheduled event occurs within given interval between two timestamps).
  • Django-REST-Framework
    • Serializer field for serializing and deserializing schedules.

Quick example

Just a short example, that shows base capabilities of the library.

Let's create a schedule, that means 10:00 - 19:00 from Monday to Thursday and 10:00 - 16:00 in Friday:

>>> import datetime as dt
>>> from tempo.recurrenteventset import RecurrentEventSet
>>> recurrenteventset = RecurrentEventSet.from_json(
...     ('OR',
...         ('AND', [1, 5, 'day', 'week'], [10, 19, 'hour', 'day']),
...         ('AND', [5, 6, 'day', 'week'], [10, 16, 'hour', 'day']))
... )

Then, test if Thursday 18:00 is included in the schedule:

>>> d1 = dt.datetime(year=2000, month=10, day=5, hour=18)
>>> d1.weekday()  # Thursday
3
>>> d1 in recurrenteventset
True

Then, do the same thing, but for Friday 18:00:

>>> d2 = dt.datetime(year=2000, month=10, day=6, hour=18)
>>> d2.weekday()  # Friday
4
>>> d2 in recurrenteventset
False

It's not included. Because our schedule includes only time from 10:00 to 16:00 in Fridays.

We also can query for future time intervals, defined by the schedule, starting from certain point of time:

>>> from itertools import islice
>>> d = dt.datetime(year=2000, month=1, day=1)
>>> list(islice(recurrenteventset.forward(start=d), 3))
[(datetime.datetime(2000, 1, 3, 10, 0),
  datetime.datetime(2000, 1, 3, 19, 0)),
 (datetime.datetime(2000, 1, 4, 10, 0),
  datetime.datetime(2000, 1, 4, 19, 0)),
 (datetime.datetime(2000, 1, 5, 10, 0),
  datetime.datetime(2000, 1, 5, 19, 0))]

Schedule model

Example

Here is an example of how Tempo represents schedules:

('OR',
        ('AND', [1, 5, 'day', 'week'], [10, 19, 'hour', 'day']),
        ('AND', [5, 6, 'day', 'week'], [10, 16, 'hour', 'day'])))

It means "from Monday to Thursday between 10am and 7pm and in Friday between 10am and 4pm".

Informal definition

Basic building block of schedule is a recurrent event, which is defined is such way:

[<start time>, <end time>, <time unit>, <recurrence unit>]

<start time> and <end time> are numbers, that defines interval in which event takes it`s place. <time unit> defines a unit of measurement of time for values of the interval. And <recurrence unit> defines how often the interval repeats. <time unit> and <recurrence unit> values are time measurement units, such as 'second', 'hour', 'day', 'week', 'year', etc. <recurrence unit> also can be 'null', which means, that the interval doesn't repeats in time, it just defines two points in time, that corresponds to start and end points of the event.

Recurrent events can be composed, using operators: union - or, intersection - and and negation - not.

Alternatives

• python-dateutil
• croniter

TODO

1. More tests for RecurrentEventSet.
2. Implement negative indexing for schedules - indexing from an end of a day or month, etc. It will make library able to model schedules like "last Friday of the month".