使用列表数组的共享内存的Python多处理

Question

I need to know if there is a way to share memory for an array of bool lists:

s = Array( 'x' , range(10000000)) 

What do I have to write instead of x to make it an array of lists whose size is 64. Then I have to manipulate "s" in two different processes like these:

#First

for i in range(0,photoCount):
    db.newDBS.insert_one({ 'photo' : s[i], 'user_id' : ids[i] })

#Second

s[photoCount].append = inPhoto

What should be the type? Any help or suggestion appreciated. Thank you!

Answer 1

Python Multiprocessing module allows 2 types of shared variables : Array , which is a simple 1D array of a single dtype, and Value , which is just a single value.

You can design your own shared variables using ctype, if you know your way in C : see the documentation here.

If you just need a 2D Array and you don't want to use ctype shared objects, maybe you could flatten your array in a single list and use multiprocessing.Array instead? Then just reshape when your processing is done.