croniters 0.2.1

A port of python's `croniter` package in Rust and exposed on PyPI via `pyo3` as `croniters`.

croniters

[!WARNING] This is a side project, and in some cases this may be marginally slower than the python version until the core logic is ported to rust, avoiding superfluous boundary crossing. However, it does pass the entire original test suite on all platforms and is a drop-in replacement in prefect.

a port of python's croniter package in Rust and exposed on PyPI via pyo3 as croniters.

original usage docs (not my work, but adapted imports)

Usage

A simple example::

>>> from croniters import croniter
>>> from datetime import datetime
>>> base = datetime(2010, 1, 25, 4, 46)
>>> iter = croniter('*/5 * * * *', base)  # every 5 minutes
>>> print(iter.get_next(datetime))   # 2010-01-25 04:50:00
>>> print(iter.get_next(datetime))   # 2010-01-25 04:55:00
>>> print(iter.get_next(datetime))   # 2010-01-25 05:00:00
>>>
>>> iter = croniter('2 4 * * mon,fri', base)  # 04:02 on every Monday and Friday
>>> print(iter.get_next(datetime))   # 2010-01-26 04:02:00
>>> print(iter.get_next(datetime))   # 2010-01-30 04:02:00
>>> print(iter.get_next(datetime))   # 2010-02-02 04:02:00
>>>
>>> iter = croniter('2 4 1 * wed', base)  # 04:02 on every Wednesday OR on 1st day of month
>>> print(iter.get_next(datetime))   # 2010-01-27 04:02:00
>>> print(iter.get_next(datetime))   # 2010-02-01 04:02:00
>>> print(iter.get_next(datetime))   # 2010-02-03 04:02:00
>>>
>>> iter = croniter('2 4 1 * wed', base, day_or=False)  # 04:02 on every 1st day of the month if it is a Wednesday
>>> print(iter.get_next(datetime))   # 2010-09-01 04:02:00
>>> print(iter.get_next(datetime))   # 2010-12-01 04:02:00
>>> print(iter.get_next(datetime))   # 2011-06-01 04:02:00
>>>
>>> iter = croniter('0 0 * * sat#1,sun#2', base)  # 1st Saturday, and 2nd Sunday of the month
>>> print(iter.get_next(datetime))   # 2010-02-06 00:00:00
>>>
>>> iter = croniter('0 0 * * 5#3,L5', base)  # 3rd and last Friday of the month
>>> print(iter.get_next(datetime))   # 2010-01-29 00:00:00
>>> print(iter.get_next(datetime))   # 2010-02-19 00:00:00

All you need to know is how to use the constructor and the get_next method, the signature of these methods are listed below::

>>> def __init__(self, cron_format, start_time=time.time(), day_or=True)

croniter iterates along with cron_format from start_time. cron_format is min hour day month day_of_week, you can refer to http://en.wikipedia.org/wiki/Cron for more details. The day_or switch is used to control how croniter handles day and day_of_week entries. Default option is the cron behaviour, which connects those values using OR. If the switch is set to False, the values are connected using AND. This behaves like fcron and enables you to e.g. define a job that executes each 2nd Friday of a month by setting the days of month and the weekday. ::

>>> def get_next(self, ret_type=float)

get_next calculates the next value according to the cron expression and returns an object of type ret_type. ret_type should be a float or a datetime object.

Supported added for get_prev method. (>= 0.2.0)::

>>> base = datetime(2010, 8, 25)
>>> itr = croniter('0 0 1 * *', base)
>>> print(itr.get_prev(datetime))  # 2010-08-01 00:00:00
>>> print(itr.get_prev(datetime))  # 2010-07-01 00:00:00
>>> print(itr.get_prev(datetime))  # 2010-06-01 00:00:00

You can validate your crons using is_valid class method. (>= 0.3.18)::

>>> croniter.is_valid('0 0 1 * *')  # True
>>> croniter.is_valid('0 wrong_value 1 * *')  # False

About DST

Be sure to init your croniter instance with a TZ aware datetime for this to work!

Example using pytz::

>>> import pytz
>>> tz = pytz.timezone("Europe/Paris")
>>> local_date = tz.localize(datetime(2017, 3, 26))
>>> val = croniter('0 0 * * *', local_date).get_next(datetime)

Example using python_dateutil::

>>> import dateutil.tz
>>> tz = dateutil.tz.gettz('Asia/Tokyo')
>>> local_date = datetime(2017, 3, 26, tzinfo=tz)
>>> val = croniter('0 0 * * *', local_date).get_next(datetime)

Example using python built in module::

>>> from datetime import datetime, timezone
>>> local_date = datetime(2017, 3, 26, tzinfo=timezone.utc)
>>> val = croniter('0 0 * * *', local_date).get_next(datetime)

About second repeats

Croniter is able to do second repetition crontabs form and by default seconds are the 6th field::

>>> base = datetime(2012, 4, 6, 13, 26, 10)
>>> itr = croniter('* * * * * 15,25', base)
>>> itr.get_next(datetime) # 4/6 13:26:15
>>> itr.get_next(datetime) # 4/6 13:26:25
>>> itr.get_next(datetime) # 4/6 13:27:15

You can also note that this expression will repeat every second from the start datetime.::

>>> croniter('* * * * * *', local_date).get_next(datetime)

You can also use seconds as first field::

>>> itr = croniter('15,25 * * * * *', base, second_at_beginning=True)

About year

Croniter also support year field. Year presents at the seventh field, which is after second repetition. The range of year field is from 1970 to 2099. To ignore second repetition, simply set second to 0 or any other const::

>>> base = datetime(2012, 4, 6, 2, 6, 59)
>>> itr = croniter('0 0 1 1 * 0 2020/2', base)
>>> itr.get_next(datetime) # 2020 1/1 0:0:0
>>> itr.get_next(datetime) # 2022 1/1 0:0:0
>>> itr.get_next(datetime) # 2024 1/1 0:0:0

Support for start_time shifts

See https://github.com/kiorky/croniter/pull/76, You can set start_time=, then expand_from_start_time=True for your generations to be computed from start_time instead of calendar days::

>>> from pprint import pprint
>>> iter = croniter('0 0 */7 * *', start_time=datetime(2024, 7, 11), expand_from_start_time=True);pprint([iter.get_next(datetime) for a in range(10)])
[datetime.datetime(2024, 7, 18, 0, 0),
 datetime.datetime(2024, 7, 25, 0, 0),
 datetime.datetime(2024, 8, 4, 0, 0),
 datetime.datetime(2024, 8, 11, 0, 0),
 datetime.datetime(2024, 8, 18, 0, 0),
 datetime.datetime(2024, 8, 25, 0, 0),
 datetime.datetime(2024, 9, 4, 0, 0),
 datetime.datetime(2024, 9, 11, 0, 0),
 datetime.datetime(2024, 9, 18, 0, 0),
 datetime.datetime(2024, 9, 25, 0, 0)]
>>> # INSTEAD OF THE DEFAULT BEHAVIOR:
>>> iter = croniter('0 0 */7 * *', start_time=datetime(2024, 7, 11), expand_from_start_time=False);pprint([iter.get_next(datetime) for a in range(10)])
[datetime.datetime(2024, 7, 15, 0, 0),
 datetime.datetime(2024, 7, 22, 0, 0),
 datetime.datetime(2024, 7, 29, 0, 0),
 datetime.datetime(2024, 8, 1, 0, 0),
 datetime.datetime(2024, 8, 8, 0, 0),
 datetime.datetime(2024, 8, 15, 0, 0),
 datetime.datetime(2024, 8, 22, 0, 0),
 datetime.datetime(2024, 8, 29, 0, 0),
 datetime.datetime(2024, 9, 1, 0, 0),
 datetime.datetime(2024, 9, 8, 0, 0)]

Testing if a date matches a crontab

Test for a match with (>=0.3.32)::

>>> croniter.match("0 0 * * *", datetime(2019, 1, 14, 0, 0, 0, 0))
True
>>> croniter.match("0 0 * * *", datetime(2019, 1, 14, 0, 2, 0, 0))
False
>>>
>>> croniter.match("2 4 1 * wed", datetime(2019, 1, 1, 4, 2, 0, 0)) # 04:02 on every Wednesday OR on 1st day of month
True
>>> croniter.match("2 4 1 * wed", datetime(2019, 1, 1, 4, 2, 0, 0), day_or=False) # 04:02 on every 1st day of the month if it is a Wednesday
False

Testing if a crontab matches in datetime range

Test for a match_range with (>=2.0.3)::

>>> croniter.match_range("0 0 * * *", datetime(2019, 1, 13, 0, 59, 0, 0), datetime(2019, 1, 14, 0, 1, 0, 0))
True
>>> croniter.match_range("0 0 * * *", datetime(2019, 1, 13, 0, 1, 0, 0), datetime(2019, 1, 13, 0, 59, 0, 0))
False
>>> croniter.match_range("2 4 1 * wed", datetime(2019, 1, 1, 3, 2, 0, 0), datetime(2019, 1, 1, 5, 1, 0, 0))
# 04:02 on every Wednesday OR on 1st day of month
True
>>> croniter.match_range("2 4 1 * wed", datetime(2019, 1, 1, 3, 2, 0, 0), datetime(2019, 1, 1, 5, 2, 0, 0), day_or=False)
# 04:02 on every 1st day of the month if it is a Wednesday
False

Gaps between date matches

For performance reasons, croniter limits the amount of CPU cycles spent attempting to find the next match. Starting in v0.3.35, this behavior is configurable via the max_years_between_matches parameter, and the default window has been increased from 1 year to 50 years.

The defaults should be fine for many use cases. Applications that evaluate multiple cron expressions or handle cron expressions from untrusted sources or end-users should use this parameter. Iterating over sparse cron expressions can result in increased CPU consumption or a raised CroniterBadDateError exception which indicates that croniter has given up attempting to find the next (or previous) match. Explicitly specifying max_years_between_matches provides a way to limit CPU utilization and simplifies the iterable interface by eliminating the need for CroniterBadDateError. The difference in the iterable interface is based on the reasoning that whenever max_years_between_matches is explicitly agreed upon, there is no need for croniter to signal that it has given up; simply stopping the iteration is preferable.

This example matches 4 AM Friday, January 1st. Since January 1st isn't often a Friday, there may be a few years between each occurrence. Setting the limit to 15 years ensures all matches::

>>> it = croniter("0 4 1 1 fri", datetime(2000,1,1), day_or=False, max_years_between_matches=15).all_next(datetime)
>>> for i in range(5):
...     print(next(it))
...
2010-01-01 04:00:00
2016-01-01 04:00:00
2021-01-01 04:00:00
2027-01-01 04:00:00
2038-01-01 04:00:00

However, when only concerned with dates within the next 5 years, simply set max_years_between_matches=5 in the above example. This will result in no matches found, but no additional cycles will be wasted on unwanted matches far in the future.

Iterating over a range using cron

Find matches within a range using the croniter_range() function. This is much like the builtin range(start,stop,step) function, but for dates. The step argument is a cron expression. Added in (>=0.3.34)

List the first Saturday of every month in 2019::

>>> from croniter import croniter_range
>>> for dt in croniter_range(datetime(2019, 1, 1), datetime(2019, 12, 31), "0 0 * * sat#1"):
>>>     print(dt)

Hashed expressions

croniter supports Jenkins-style hashed expressions, using the "H" definition keyword and the required hash_id keyword argument. Hashed expressions remain consistent, given the same hash_id, but different hash_ids will evaluate completely different to each other. This allows, for example, for an even distribution of differently-named jobs without needing to manually spread them out.

>>> itr = croniter("H H * * *", hash_id="hello")
>>> itr.get_next(datetime)
datetime.datetime(2021, 4, 10, 11, 10)
>>> itr.get_next(datetime)
datetime.datetime(2021, 4, 11, 11, 10)
>>> itr = croniter("H H * * *", hash_id="hello")
>>> itr.get_next(datetime)
datetime.datetime(2021, 4, 10, 11, 10)
>>> itr = croniter("H H * * *", hash_id="bonjour")
>>> itr.get_next(datetime)
datetime.datetime(2021, 4, 10, 20, 52)

Random expressions

Random "R" definition keywords are supported, and remain consistent only within their croniter() instance.

>>> itr = croniter("R R * * *")
>>> itr.get_next(datetime)
datetime.datetime(2021, 4, 10, 22, 56)
>>> itr.get_next(datetime)
datetime.datetime(2021, 4, 11, 22, 56)
>>> itr = croniter("R R * * *")
>>> itr.get_next(datetime)
datetime.datetime(2021, 4, 11, 4, 19)

Note about Ranges

Note that as a deviation from cron standard, croniter is somehow laxist with ranges and will allow ranges of Jan-Dec, & Sun-Sat in reverse way and interpret them as following examples:

- ``Apr-Jan``: from April to january
- ``Sat-Sun``: Saturday, Sunday
- ``Wed-Sun``: Wednesday to Saturday, Sunday

Please note that if a /step is given, it will be respected.

Note about Sunday

Note that as a deviation from cron standard, croniter like numerous cron implementations supports SUNDAY to be expressed as DAY7, allowing such expressions:

- ``0 0 * * 7``
- ``0 0 * * 6-7``
- ``0 0 * * 6,7``

Keyword expressions

Vixie cron-style "@" keyword expressions are supported. What they evaluate to depends on whether you supply hash_id: no hash_id corresponds to Vixie cron definitions (exact times, minute resolution), while with hash_id corresponds to Jenkins definitions (hashed within the period, second resolution).

============ ============ ================
Keyword      No hash_id   With hash_id
============ ============ ================
@midnight    0 0 * * *    H H(0-2) * * * H
@hourly      0 * * * *    H * * * * H
@daily       0 0 * * *    H H * * * H
@weekly      0 0 * * 0    H H * * H H
@monthly     0 0 1 * *    H H H * * H
@yearly      0 0 1 1 *    H H H H * H
@annually    0 0 1 1 *    H H H H * H
============ ============ ================

Overview

croniters intends to be a drop-in replacement for the python croniter package, implemented in Rust.

[!IMPORTANT] I say "intends to" because while the test suite should cover the vast majority of cases important to croniter users, it seems likely that subtle differences in behavior between rust and python implementations could eventually exist.

Why?

croniter is no longer maintained and everyone vendoring means more collective toil.

rust already has good datetime support and rust types should be good for this sort of pedantry.

Project Roadmap

• port everything to rust incrementally, maintaining the public croniter API
• minimize boundary crossings between rust and python, using deprecation where necessary
• make best use of existing rust datetime/cronstring parsing crates

Phase 0: Setup

• get pyo3 setup
• port test suite from croniter and it passes
• move constants and simple utils to rust
• basic benchmarking

Phase 1: Migrate the croniter class itself

• get a similarly shaped croniter class to compile
• run tests and weep at all the failures
• systematically identify disparities and fix them

[!NOTE] Think I'm doing something stupid?

Drop me a line! I probably am.

(Planned) Technical Stack

• Cron Parsing: cron crate
• Python Bindings: pyo3
• Build System: maturin

Contributing

Please do - I am only learning rust and pyo3! See the justfile for details on running locally.

License

Same as croniter - MIT License - see the LICENSE file for details.

Acknowledgements

• croniter - original implementation
• pyo3 - Python bindings
• maturin - build system