import multiprocessing, threading
import time
import random

ARROW_NUMBER = 150
MAX_CYCLES = 20000
SPACE_SIZE = 10000
PREEMPTION_GLOBAL_SPACE = [True] * SPACE_SIZE
ARROW_LIST = [(random.randint(0,SPACE_SIZE - 1),0) for x in range(ARROW_NUMBER)]
TRANSITIONS_TREE = None
MAX_THREAD = 10


class CurrentlyComputing(Exception):
    pass

class SuccessfulOperation(Exception):
    pass

class FailedOperation(Exception):
    pass

## Local Threads

class LocalThread(multiprocessing.Process):

    def __init__(self, id, shared_memory, address, orientation, transitions):

        multiprocessing.Process.__init__(self)

        self.id = id
        self.address = address
        self.orientation = orientation
        self.transitions = transitions
        self.namespace = shared_memory.Namespace()
        self.namespace.returncode = CurrentlyComputing()

    def run(self):
        try:
            # Actual code
            print("Thread local n°{} parle depuis {} !".format(self.id, self.address))

            L = [random.randint(0,3000) for x in range(random.randint(0, 300))]

            for i in xrange(10000000):
                sum(range(100))

            self.namespace.returncode = SuccessfulOperation()

        except Exception as exception:
             self.namespace.returncode = exception

## Master Thread
class GreatScheduler(threading.Thread):

    def __init__(self, preemption_space, arrow_list, n_thread, nmax_cycles):

        threading.Thread.__init__(self)

        self.preemption_space = preemption_space
        self.arrow_list = arrow_list
        self.n_thread = n_thread
        self.cur_id = -1
        self.n_cycle = -1
        self.nmax_cycles = nmax_cycles
        self.stopped = False

    def run(self):
        thread_list = []
        shared_memory = multiprocessing.Manager()

        while not self.stopped and self.n_cycle != self.nmax_cycles:

            self.n_cycle += 1

            #print("--- Mesures et écoute ---")
            #print("Espace global de préemption :\t {}".format(self.preemption_space))
            #print("Liste des flèches :\t {}".format(self.arrow_list))

            ## Tirer une flêche au hasard
            n = int(random.random()*len(self.arrow_list))
            elected_arrow = self.arrow_list[n]

            ## Vérification de l'espace global de préemption autour de la flêche
            if not(len(thread_list) >= self.n_thread) and self.preemption_space[elected_arrow[0]]:
                print("Espace local libre à {}".format(elected_arrow[0]))
                self.preemption_space[elected_arrow[0]] = False

                ## Création du thread local

                self.cur_id += 1

                thread_list.append(
                    LocalThread(self.cur_id,
                        shared_memory,
                        elected_arrow[0],
                        elected_arrow[1],
                        TRANSITIONS_TREE
                    )
                )

                ## Lancer le thread local
                thread_list[-1].start()

            for thread in thread_list:
                if not thread.is_alive():
                    if not thread.namespace.returncode.__class__ is SuccessfulOperation:
                        if thread.namespace.returncode.__class__ is FailedOperation:
                            print("A optimiser : créer une règle")
                        else:
                            raise thread.namespace.returncode
                    print("Thread local n°{} est terminé".format(thread.id))

                    self.preemption_space[thread.address] = True
                    thread_list.pop(thread_list.index(thread))

        ## Sortie du scheduler
        print("Attente de la fin des thread locaux")

        for thread in thread_list:
            thread.join()
            if not thread.namespace.returncode.__class__ is SuccessfulOperation:
                if thread.namespace.returncode.__class__ is FailedOperation:
                    print("A optimiser : créer une règle")
                else:
                    raise thread.namespace.returncode
            print("Thread local n°{} est terminé".format(thread.id))

class ServerCLI:

    def start():
        stopped = False
        while not stopped:
            stopped = input()
        return

#multiprocessing.set_start_method('fork')
scheduler = GreatScheduler(PREEMPTION_GLOBAL_SPACE, ARROW_LIST, MAX_THREAD, MAX_CYCLES)
scheduler.start()
ServerCLI.start()
scheduler.stopped = True