Python Crontab API
Bug Reports and Development
Note: If you get the error TypeError: __init__() takes exactly 2 arguments when using CronTab, you have the wrong module installed. You need to install python-crontab and not crontab from pypi or your local package manager and try again.
Crontab module for reading and writing crontab files and accessing the system cron automatically and simply using a direct API.
Comparing the below chart you will note that W, L, # and ? symbols are not supported as they are not standard Linux or SystemV crontab format.
|Field Name||Mandatory||Allowed Values||Special Characters||Extra Values|
|Minutes||Yes||0-59||* / , -||< >|
|Hours||Yes||0-23||* / , -||< >|
|Day of month||Yes||1-31||* / , -||< >|
|Month||Yes||1-12 or JAN-DEC||* / , -||< >|
|Day of week||Yes||0-6 or SUN-SAT||* / , -||< >|
Extra Values are ‘<’ for minimum value, such as 0 for minutes or 1 for months. And ‘>’ for maximum value, such as 23 for hours or 12 for months.
Supported special cases allow crontab lines to not use fields. These are the supported aliases which are not available in SystemV mode:
|@hourly||0 * * * *|
|@daily||0 0 * * *|
|@weekly||0 0 * * 0|
|@monthly||0 0 1 * *|
|@yearly||0 0 1 1 *|
|@annually||0 0 1 1 *|
|@midnight||0 0 * * *|
How to Use the Module
Here is a simple example of how python-crontab is typically used. First the CronTab class is used to instantiate a cron object, then the cron object is used to declaratively manipulate the cron (spawning a new job in this case). Lastly, declared changes get written to the crontab by calling write on the object:
from crontab import CronTab cron = CronTab(user='root') job = cron.new(command='echo hello_world') job.minute.every(1) cron.write()
Alternatively, you can use the with context manager which will automatically call write on the cron object upon exit:
with CronTab(user='root') as cron: job = cron.new(command='echo hello_world') job.minute.every(1) print('cron.write() was just executed')
Note: Several users have reported their new crontabs not saving automatically or that the module doesn’t do anything. You MUST use write() if you want your edits to be saved out. See below for full details on the use of the write function.
Getting access to a crontab can happen in five ways, three system methods that will work only on Unix and require you to have the right permissions:
from crontab import CronTab empty_cron = CronTab() my_user_cron = CronTab(user=True) users_cron = CronTab(user='username')
And two ways from non-system sources that will work on Windows too:
file_cron = CronTab(tabfile='filename.tab') mem_cron = CronTab(tab=""" * * * * * command """)
Special per-command user flag for vixie cron format (new in 1.9):
system_cron = CronTab(tabfile='/etc/crontab', user=False) job = system_cron job.user != None system_cron.new(command='new_command', user='root')
Creating a new job is as simple as:
job = cron.new(command='/usr/bin/echo')
And setting the job’s time restrictions:
job.minute.during(5,50).every(5) job.hour.every(4) job.day.on(4, 5, 6) job.dow.on('SUN') job.dow.on('SUN', 'FRI') job.month.during('APR', 'NOV')
Each time restriction will clear the previous restriction:
job.hour.every(10) # Set to * */10 * * * job.hour.on(2) # Set to * 2 * * *
Appending restrictions is explicit:
job.hour.every(10) # Set to * */10 * * * job.hour.also.on(2) # Set to * 2,*/10 * * *
Setting all time slices at once:
job.setall(2, 10, '2-4', '*/2', None) job.setall('2 10 * * *')
Setting the slice to a python date object:
job.setall(time(10, 2)) job.setall(date(2000, 4, 2)) job.setall(datetime(2000, 4, 2, 10, 2))
Run a jobs command. Running the job here will not effect it’s existing schedule with another crontab process:
job_standard_output = job.run()
Creating a job with a comment:
job = cron.new(command='/foo/bar', comment='SomeID')
Get the comment or command for a job:
command = job.command comment = job.comment
Modify the comment or command on a job:
job.set_command("new_script.sh") job.set_comment("New ID or comment here")
Disabled or Enable Job:
job.enable() job.enable(False) False is job.is_enabled()
True is job.is_valid()
Use a special syntax:
Find an existing job by command sub-match or regular expression:
iter = cron.find_command('bar') # matches foobar1 iter = cron.find_command(re.compile(r'b[ab]r$'))
Find an existing job by comment exact match or regular expression:
iter = cron.find_comment('ID or some text') iter = cron.find_comment(re.compile(' or \w'))
Find an existing job by schedule:
iter = cron.find_time(2, 10, '2-4', '*/2', None) iter = cron.find_time("*/2 * * * *")
Clean a job of all rules:
Iterate through all jobs, this includes disabled (commented out) cron jobs:
for job in cron: print(job)
Iterate through all lines, this includes all comments and empty lines:
for line in cron.lines: print(line)
cron.remove( job ) cron.remove_all('echo') cron.remove_all(comment='foo') cron.remove_all(time='*/2')
Clear entire cron of all jobs:
Write CronTab back to system or filename:
Write CronTab to new filename:
cron.write( 'output.tab' )
Write to this user’s crontab (unix only):
cron.write_to_user( user=True )
Write to some other user’s crontab:
cron.write_to_user( user='bob' )
Validate a cron time string:
from crontab import CronSlices bool = CronSlices.is_valid('0/2 * * * *')
Compare list of cron objects against another and return the difference:
difference = set([CronItem1, CronItem2, CronItem3]) - set([CronItem2, CronItem3])
Compare two CronItems for equality:
CronItem1 = CronTab(tab="* * * * * COMMAND # Example Job") CronItem2 = CronTab(tab="10 * * * * COMMAND # Example Job 2") if CronItem1 != CronItem2: print("Cronjobs do not match")
Some versions of vixie cron support variables outside of the command line. Sometimes just update the envronment when commands are run, the Cronie fork of vixie cron also supports CRON_TZ which looks like a regular variable but actually changes the times the jobs are run at.
Very old vixie crons don’t support per-job variables, but most do.
Iterate through cron level environment variables:
for (name, value) in cron.env.items(): print(name) print(value)
Create new or update cron level environment variables:
print(cron.env['SHELL']) cron.env['SHELL'] = '/bin/bash' print(cron.env)
Each job can also have a list of environment variables:
for job in cron: job.env['NEW_VAR'] = 'A' print(job.env)
Proceeding Unit Confusion
It is sometimes logical to think that job.hour.every(2) will set all proceeding units to ‘0’ and thus result in “0 */2 * * *”. Instead you are controlling only the hours units and the minute column is unaffected. The real result would be “* */2 * * *” and maybe unexpected to those unfamiliar with crontabs.
There is a special ‘every’ method on a job to clear the job’s existing schedule and replace it with a simple single unit:
job.every(4).hours() == '0 */4 * * *' job.every().dom() == '0 0 * * *' job.every().month() == '0 0 0 * *' job.every(2).dows() == '0 0 * * */2'
This is a convenience method only, it does normal things with the existing api.
Running the Scheduler
The module is able to run a cron tab as a daemon as long as the optional croniter module is installed; each process will block and errors will be logged (new in 2.0).
(note this functionality is new and not perfect, if you find bugs report them!)
Running the scheduler:
tab = CronTab(tabfile='MyScripts.tab') for result in tab.run_scheduler(): print("This was printed to stdout by the process.")
Do not do this, it won’t work because it returns generator function:
Timeout and cadence can be changed for testing or error management:
for result in tab.run_scheduler(timeout=600): print("Will run jobs every 1 minutes for ten minutes from now()") for result in tab.run_scheduler(cadence=1, warp=True): print("Will run jobs every 1 second, counting each second as 1 minute")
Every job’s schedule has a frequency. We can attempt to calculate the number of times a job would execute in a give amount of time. We have three simple methods:
job.setall("1,2 1,2 * * *") job.frequency_per_day() == 4
The per year frequency method will tell you how many days a year the job would execute:
job.setall("* * 1,2 1,2 *") job.frequency_per_year(year=2010) == 4
These are combined to give the number of times a job will execute in any year:
job.setall("1,2 1,2 1,2 1,2 *") job.frequency(year=2010) == 16
Frequency can be quickly checked using python built-in operators:
job < "*/2 * * * *" job > job2 job.slices == "*/5"
The log functionality will read a cron log backwards to find you the last run instances of your crontab and cron jobs.
The crontab will limit the returned entries to the user the crontab is for:
cron = CronTab(user='root') for d in cron.log: print(d['pid'] + " - " + d['date'])
Each job can return a log iterator too, these are filtered so you can see when the last execution was:
for d in cron.find_command('echo').log: print(d['pid'] + " - " + d['date'])
All System CronTabs Functionality
The crontabs (note the plural) module can attempt to find all crontabs on the system. This works well for Linux systems with known locations for cron files and user spolls. It will even extract anacron jobs so you can get a picture of all the jobs running on your system:
from crontabs import CronTabs for cron in CronTabs(): print(repr(cron))
All jobs can be brought together to run various searches, all jobs are added to a CronTab object which can be used as documented above:
jobs = CronTabs().all.find_command('foo')
If you have the croniter python module installed, you will have access to a schedule on each job. For example if you want to know when a job will next run:
schedule = job.schedule(date_from=datetime.now())
This creates a schedule croniter based on the job from the time specified. The default date_from is the current date/time if not specified. Next we can get the datetime of the next job:
datetime = schedule.get_next()
Or the previous:
datetime = schedule.get_prev()
The get methods work in the same way as the default croniter, except that they will return datetime objects by default instead of floats. If you want the original functionality, pass float into the method when calling:
datetime = schedule.get_current(float)
If you don’t have the croniter module installed, you’ll get an ImportError when you first try using the schedule function on your cron job object.
If you have the cron-descriptor module installed, you will be able to ask for a translated string which describes the frequency of the job in the current locale language. This should be mostly human readable.
See cron-descriptor for details of the supported languages and options.
- Support for vixie cron with username addition with user flag
- Support for SunOS, AIX & HP with compatibility ‘SystemV’ mode.
- Python 3 (3.5, 3.6, 3.7) and Python 2.7 tested, python 2.6 removed from support.
- Windows support works for non-system crontabs only. ( see mem_cron and file_cron examples above for usage )
Release history Release notifications | RSS feed
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size python-crontab-2.5.1.tar.gz (52.6 kB)||File type Source||Python version None||Upload date||Hashes View|