One 、 Semaphore ： A piece of code can only be n Process execution Example ： # Semaphore ： Control concurrency import threading import time def run(n): semaphore.acquire() # Start event volume print(n) time.sleep(1) semaphore.release()# Release event volume if __name__ == '__main__': semaphore = threading.Semaphore(5)# Run five threads at a time , One of the threads ends the lock release for i in range(50): t = threading.Thread(target=run,args=(i,)) t.start()
Two 、Condition（ Condition variables, ） Usually associated with a lock . Need more than one Contidion When sharing a lock in , You can pass a Lock/RLock Examples to construct methods , Otherwise it will generate a RLock example .
It can be said that , except Lock With the lock out of the pool ,Condition It also contains a waiting pool , The threads in the pool are in the waiting blocking state in the state diagram , Until another thread calls notify()/notifyAll() notice ; When notified, the thread enters the lock pool and waits for the lock .
Condition():
''' Condition variables, '''
import threading
import time
def run(x):
con.acquire()
print(f' Threads {x}')
con.notify() # Tell the next thread wait end
print(f' Threads {x} Hang up ')
con.wait()
time.sleep
print(f' Threads {x} Start again ')
con.notify()
con.release()
if __name__ == '__main__':
con = threading.Lock() # Instantiate condition variables
# for i in range(10):
t = threading.Thread(target=run,args=(1,))
t.start()
Example ： Implementation scenario ： When a Classmate Wang hot pot is filled with fish balls （ most 5 individual , Notify when full b Eat it ）, notice b Students eat fish balls （ Eat to 0 Notice when a Students continue to add ）# coding=utf-8 import threading import time con = threading.Condition() num = 0 # producer class Producer(threading.Thread): def __init__(self): threading.Thread.__init__(self) def run(self): # Lock thread global num con.acquire() while True: print " Start adding ！！！" num += 1 print " The number of fish balls in the hot pot ：%s" % str(num) time.sleep(1) if num >= 5: print " The number of fish balls in the hot pot has reached 5 individual , Unable to add ！" # Wakes up the waiting thread con.notify() # Wake up the little friend and eat # Wait for a notice con.wait() # Release the lock con.release() # consumer class Consumers(threading.Thread): def __init__(self): threading.Thread.__init__(self) def run(self): con.acquire() global num while True: print " Start eating ！！！" num -= 1 print " The number of fish balls left in the hot pot ：%s" %str(num) time.sleep(2) if num <= 0: print " The bottom of the pot is out of stock , Add fish balls quickly ！" con.notify() # Wake up other threads # Wait for a notice con.wait() con.release() p = Producer() c = Consumers() p.start() c.start()
3、 ... and 、 event
（1） A single signal can put all processes into a blocking state
（2） You can also control all processes to unblock
（3） After an event is created , The default is blocking
Example ：
import threading
import time
def car():
while True:
if event.is_set(): # If the event is executing
print(' The car runs ')
else:
print(' The trolley stops ')
event.wait() # Events wait
def set_event():
while True:
event.set() # Start events
time.sleep(1)
event.clear() # Clear the event
time.sleep(1)
if __name__ == '__main__':
event = threading.Event() # Create an event
car1 = threading.Thread(target=car)
car1.start()
set_e = threading.Thread(target=set_event)
set_e.start()Video Explanation ：https://space.bilibili.com/196858266/channel/collectiondetail?sid=57996
Condition() Condition variables, :
Official documents ：threading --- Thread based parallelism — Python 3.9.9 file
