Thread Module
* Note: Thread is not recommended for multithreaded programming during actual use, this document is only for learning (or familiarity) multithreading.
The thread module provides basic synchronization data structures in addition to derived threads, called Lock Objects (also called Yuri, mutexes, mutexes, and binary semaphores).
Common thread functions and methods for LockType lock objects:
| Functions/Methods |
Description |
| Functions of the Thread module |
|
| Start_new_thread (function, args, kwargs=none) |
Derives a new thread, using the given args and optional Kwargs to perform the function |
| Allocate_lock () |
Assigning LockType Lock objects |
| Exit () |
Give thread exit instruction |
| LockType Lock Object method |
|
| Acquire (Wait=none) |
Attempt to get lock object |
| Locked () |
Returns true if the lock object is acquired, otherwise returns flase |
| Release () |
Release lock |
For a simple example using the thread module, the code is as follows (mtsleepa.py):
1 ImportThread2 fromTimeImportSleep, CTime3 ?4 defloop0 ():5 Print 'start loop 0 at:', CTime ()6Sleep (4)7 Print 'Loop 0 done at:', CTime ()8 ?9 defLoop1 ():Ten Print 'start loop 1 at:', CTime () OneSleep (2) A Print 'Loop 1 done at:', CTime () - ? - defMain (): the Print 'starting at:', CTime () - Thread.start_new_thread (Loop0, ()) - Thread.start_new_thread (LOOP1, ()) -Sleep (6) + Print 'All do at :', CTime () - ? + if __name__=='__main__': AMain ()
Output Result:
1 starting At:sun Jul 21:38:00 20182 start loop 0 At:sun Jul 21:38:00 20183 start L OOP 1 At:sun 21:38:00 20184 Loop 1 Done at:sun Jul 21:38:02 20185 loops 0 done At:sun Jul 22 21:38:04 20186 all done At:sun Jul 22 21:38:06 2018
In the code of this script, a sleep (6) Call is added, why do you do this? This is because if we do not prevent the main thread from continuing, it will continue to execute the next statement, show "all done" and then exit, while the two threads Loop0 () and LOOP1 () will terminate directly.
For a simple example using threads and locks, the code is as follows (mtsleepb.py):
1 ImportThread2 fromTimeImportSleep, CTime3 ?4loops = [4,2]5 ?6 defLoop (Nloop, Nsec, lock):7 Print 'Start Loop', Nloop,'At :', CTime ()8 Sleep (nsec)9 Print 'Loop', Nloop,'Done at :', CTime ()TenLock.release ()#Release Lock One ? A defMain (): - Print 'starting at:', CTime () -Locks = [] theNloops =Range (len (loops)) - ? - forIinchNloops: -Lock = Thread.allocate_lock ()#Assigning LockType Objects +Lock.acquire ()#attempt to get lock object - Locks.append (Lock) + ? A forIinchNloops: atThread.start_new_thread (Loop, (I, loops[i], locks[i]))#deriving a new thread - - forIinchNloops: - #wait for all locks to be released after exiting the loop to continue back operation - whileLocks[i].locked ():#True when a lock is acquired, all locks are released after Flase - Pass in ? - Print 'All do at :', CTime () to ? + if __name__=='__main__': -Main ()
The output is:
1 starting At:sun Jul 22:25:45 20182 start loop 1 At:start Loop Sun Jul 22:25:45 20183
0 At:sun Jul 22:25:45 20184 Loop 1 Done at:sun Jul 22:25:47 20185 loops 0 done At:sun Jul 22 22:25:49 20186 all done At:sun Jul 22 22:25:49 2018
The main () key-related process for locking in the main () function (the first for loop) is explained:
-
First, create a lock list by using the Thread.allocate_lock () function to get the lock object;
-
Each lock is obtained by means of the Acquire () method (the effect of acquiring a lock is equivalent to "locking the lock");
-
The lock is locked and added to the lock list locks;
Why not start a thread in a locked loop?
-
First, you want to synchronize threads so that all threads are finished at the same time;
-
Second, acquiring a lock takes a bit of time to avoid the thread executing too fast, and the execution ends when the lock is acquired on the front thread.
Thread Module-python multithreaded programming