schedule

• Added @repeat() decorator. See #148. Thanks @rhagenaars!

• Added execute .until(). See #195. Thanks @fredthomsen!

• Added job retrieval filtered by tags using get_jobs('tag'). See #419. Thanks @skenvy!

• Added type annotations. See #427. Thanks @martinthoma!

• Bugfix: str() of job when there is no __name__. See #430. Thanks @biggerfisch!

• Improved error messages. See #280, #439. Thanks @connorskees and @sosolidkk!

• Improved logging. See #193. Thanks @zcking!

• Documentation improvements and fix typos. See #424, #435, #436, #453, #437, #448. Thanks @ebllg!

Depending on your configuration, the following bugfixes might change schedule's behaviour:

• Fix: idle_seconds crashes when no jobs are scheduled. See #401. Thanks @yoonghm!

• Fix: day.at('HH:MM:SS') where HMS=now+10s doesn't run today. See #331. Thanks @qmorek!

• Fix: hour.at('MM:SS'), the seconds are set to 00. See #290. Thanks @eladbi!

• Fix: Long-running jobs skip a day when they finish in the next day #404. Thanks @4379711!

Other changes:

• Dropped Python 2.7 and 3.5 support, added 3.8 and 3.9 support. See #409

• Fix RecursionError when the job is passed to the do function as an arg. See #190. Thanks @connorskees!

• Fix DeprecationWarning of 'collections'. See #296. Thanks @gaguirregabiria!

• Replaced Travis with Github Actions for automated testing

• Revamp and extend documentation. See #395

• Improved tests. Thanks @connorskees and @Jamim!

• Changed log messages to DEBUG level. Thanks @aisk!

• Make at() accept timestamps with 1 second precision (#267). Thanks @NathanWailes!

• Introduce proper exception hierarchy (#271). Thanks @ConnorSkees!

• Keep partially scheduled jobs from breaking the scheduler (#125)

• Add support for random intervals (Thanks @grampajoe and @gilbsgilbs)

• Improve docs & clean up docstrings

• Publish to PyPI as a universal (py2/py3) wheel

• Add proper HTML (Sphinx) docs available at https://schedule.readthedocs.io/

• CI builds now run against Python 2.7 and 3.5 (3.3 and 3.4 should work fine but are untested)

• Fixed an issue with run_all() and having more than one job that deletes itself in the same iteration. Thanks @alaingilbert.

• Add ability to tag jobs and to cancel jobs by tag. Thanks @Zerrossetto.

• Improve schedule docs. Thanks @Zerrossetto.

• Additional docs fixes by @fkromer and @yetingsky.

• Fixed issues where scheduling a job with a functools.partial as the job function fails. Thanks @dylwhich.

• Fixed an issue where scheduling a job to run every >= 2 days would cause the initial execution to happen one day early. Thanks @WoLfulus for identifying this and providing a fix.

• Added a FAQ item to describe how to schedule a job that runs only once.

• Fixed an issue with unicode handling in setup.py that was causing trouble on Python 3 and Debian (https://github.com/dbader/schedule/issues/27). Thanks to @waghanza for reporting it.

• Added an FAQ item to describe how to deal with job functions that throw exceptions. Thanks @mplewis.

• Added support for scheduling jobs on specific weekdays. Example: schedule.every().tuesday.do(job) or schedule.every().wednesday.at("13:15").do(job) (Thanks @abultman.)

• Run tests against Python 2.7 and 3.4. Python 3.3 should continue to work but we're not actively testing it on CI anymore.

• Fixed history (no code changes).

• This release introduces two new features in a backwards compatible way:

• Allow jobs to cancel repeated execution: Jobs can be cancelled by calling schedule.cancel_job() or by returning schedule.CancelJob from the job function. (Thanks to @cfrco and @matrixise.)

• Updated at_time() to allow running jobs at a particular time every hour. Example: every().hour.at(':15').do(job) will run job 15 minutes after every full hour. (Thanks @mattss.)

• Refactored unit tests to mock datetime in a cleaner way. (Thanks @matts.)

• Fixed an issue with next_run() throwing a ValueError exception when the job queue is empty. Thanks to @dpagano for pointing this out and thanks to @mrhwick for quickly providing a fix.

• Fixed issue with at_time jobs not running on the same day the job is created (Thanks to @mattss)

• Added schedule.next_run()

• Added schedule.idle_seconds()

• Args passed into do() are forwarded to the job function at call time

• Increased test coverage to 100%

• Changed default delay_seconds for schedule.run_all() to 0 (from 60)

• Increased test coverage

• API change: renamed schedule.run_all_jobs() to schedule.run_all()

• API change: renamed schedule.run_pending_jobs() to schedule.run_pending()

• API change: renamed schedule.clear_all_jobs() to schedule.clear()

• Added schedule.jobs

• Fix packaging

• README fixes

• API change: renamed schedule.tick() to schedule.run_pending_jobs()

• Updated README and setup.py packaging