Lock-file routines for Python. This code is a cleaned up version from some internal tummy.com projects, called "jlock.py", but never released separately. That code has been used in many projects over the years.

This code is meant to be easily embedded within some other code, just by reading it into the other source file. This is 50% of how I've used it.

Examples

Just select a lock-file name, and create a "lock" instance using that name any any parameters you want to override. Then either let the object fall out of scope or (preferred) call the "release()" method:

import pythonlock
lock = pythonlock.lock('/dev/shm/mylockfile', retries=100, do_raise=True)
if not lock.have_lock:
    print "Didn't get lock"
    print "This shouldn't happen because do_raise is True"
    sys.exit(0)
[do stuff]
lock.release()

The lock sleeps for one second, so the above will wait for the lock for 100 seconds. You can tune both of those settings.

"retries" is the number of attempts that will be made before the locking code gives up trying to get the lock.

"do_raise" if raise an exception if the lock cannot be aquired.

"try_break" if set will cause the code to read the lock-file and if the process ID listed in it is no longer running, the lock-file will be deleted.

"sleep_time" is the number of seconds between retries. This can be floating-point, as provided to "time.sleep()".