I’m trying to run a function in an update loop

def get_screencap_raw(self, emulator_id):
    # Construct the ADB command to capture a screenshot in raw format
    adb_command = [self.adb_path, "-s", emulator_id, "exec-out", "screencap"]

    # Directly execute the adb command
    process = subprocess.Popen(
        adb_command,
        stdout=subprocess.PIPE,
        startupinfo=self.startupinfo,
    )
    try:
        stdout, stderr = process.communicate(timeout=3)
    except subprocess.TimeoutExpired:
        process.kill()
        return None
    finally:
        # Ensure the pipes are closed
        process.stdout.close()
        process.terminate()

    if stderr:
        print(
            f"Error: Unable to capture screenshot for device {emulator_id}. Details: {stderr.decode('utf-8')}"
        )
        return None

    # Return the raw screenshot data
    return stdout

But when I continuously run this function, I notice the memory keeps going up and up. I am not sure if my theory is correct, but I believe continuously creating a subprocess.popen, then closing and terminating it is the cause for the memory overhead.

My solution was to try and create an instance of subprocess.popen once, save it in a self variable, and then execute a function that will execute the command for the subprocess. But I have not succeeded in this as it only returns the first output data.

I’d like to know if my theory is correct and if there is a solution to my problem to continuously run a command without creating a new subprocess with every call.