我们如何运行 2 个并行共享公共变量的 python 脚本？

Question

First let me explain the scenario.

My project is built using Django.

I have the frontend wherein we can upload the image of a person. That image will be stored in our local machine in (media folder).

Ex: If the image name is jon.jpeg , then it will be stored as

media
-jon
--jon.jpeg

If another image is uploaded, say bob.jpg , then the structure will be like:

media
-jon
--jon.jpeg
-bob
--bob.jpg

Now I have one Python script (say check.py ) in which we are running one infinite loop, and every 5 seconds it is checking the number of folders in the media folder. Also, we are maintaining one global variable in check.py that will keep track of the initial_count.

Now I have another Python file running (ie camera.py ) and it has 2 functions, ie load_encodings and recognize in a class called Face_Recog .

load_encodings will convert the images in the media folder to feature encodings and it will store these encodings in a global list ( known_faces ) in the camera.py file.

recognize will first extract the image from one camera frame and encode it, and then it will compare it with the encodings in known_faces .

Now whenever the number of folders in media changes, (if we upload the image from the frontend), then in our initial_count in check.py will change. We have given a condition wherein if the initial_count value changes, it will call the encoding function of the camera.py file.

check.py

from camera import Face_Recog,object

initial_count = 0
while True:
    count = 0
    for path in os.listdir(path/to/media):
        count += 1
    global initial_count
    if count != initial_count:
        Face_Recog.load_encodings(object)
        initial_count = count
    sleep(5)

camera.py

known_faces = []

class Face_Recog(object):
    def load_encodings(self):
        temp = []
        #some code that will change the temp value , i.e it will load the image from media and then encode it and store the encoding in temp
        global known_faces
        known_faces = temp

    def recognize(self):
        while(True):
            #code that uses known_faces and recognizes


object = Face_Recog()
object.load_encodings()
object.recognize()

Now I am running these two files in 2 different shells.

Initially the media folder contains only jon , so known_faces will contain only encodings of jon .

known_faces = ['jon']

and the recognize function will not recognize bob , it will only recognize jon .

(This is working.)

Now when I upload the image of bob , the initial_count value will change in check.py and it will call the load_encodings function of camera.py (it is calling it; I have checked using print statements).

Now this function will again encode the images present, ie jon and bob , and update the global variable, ie known_faces .

(It is updating, but only in the shell in which check.py is running, not in the shell running camera.py . Ie it is not reflecting the changes in the camera.py file's known_faces variable, which is running in a different shell instance).

So even if it is changing the known_faces variable, the recognize function is not recognizing bob , ie the variable ( known_faces ) it is using is not updated automatically.

I want to do something like whenever any person uploads any image from the frontend, its encodings should be updated in the file ( camera.py ) without the camera being stopped ie recognize function should not stop its execution.

PS. NOTE: Both the files, ie camera.py and check.py are running all the time in 2 different shells.

Please tell me if it is possible, or is there any other approach to achieve this.

Answer 1

The absolutely simplest fix for your scenario is probably to have check.py monitor the file and restart the other process when necessary.