import numpy as np
from state import create_empty_state, to_state_string
from machine_player import MachinePlayer
from human_player import HumanPlayer
from player_provider import PlayerProvider
import json

float_formatter = "{:.3f}".format
np.set_printoptions(formatter={'float_kind': float_formatter})

REPORT_INTERVAL = 1000
report_counter = REPORT_INTERVAL


def play(player_provider: PlayerProvider, k_max=10000, with_debug=False):
	player_provider.set_debug(with_debug)
	for k in range(0, k_max):
		state = create_empty_state()
		players = player_provider.choose(state)
		move = 1
		run = True
		other_player = players[-1]
		while run:
			do_stop = False
			this_reward = 0
			other_reward = 0
			for this_player in players:
				if with_debug:
					print(f"---------------------------------------------------")
					print(f"- Game {k:06d}, Move {move} -----------------------------")
					print(f"---------------------------------------------------")
				last_state = state
				state, has_moved = this_player.move(state)
				if with_debug:
					print(to_state_string(last_state, state))
				if not has_moved:
					this_reward = 0
					other_reward = 0
					do_stop = True
					if with_debug:
						print(f"{this_player.mark}: No more moves")
				if this_player.has_won(state):
					this_reward = 1
					other_reward = 0
					do_stop = True
					if with_debug:
						print(f"{this_player.mark}: Has won the game")

				if do_stop:
					this_player.end_game(state, this_reward)
					other_player.end_game(state, other_reward)
					run = False
					break

				other_player = this_player
				move += 1


do_training = 0
do_human_player = 0
do_load_values = 1

if do_load_values:
	try:
		with open("values_x.json", "r") as fp:
			x_values = json.load(fp)

		with open("values_o.json", "r") as fp:
			o_values = json.load(fp)
	except Exception:
		x_values = {}
		o_values = {}
else:
	x_values = {}
	o_values = {}

p_hx = HumanPlayer(mark='X', name="Jens")
p_mx = MachinePlayer(mark='X', params=MachinePlayer.Params(p_explore=0.1, alpha=0.1), values=x_values)
p_mo = MachinePlayer(mark='O', params=MachinePlayer.Params(p_explore=0.1, alpha=0.1), values=o_values)

K = 5
N = 1000
if do_training:
	for k in range(0, K):
		play(PlayerProvider(p_mx, p_mo), N, False)

		# Values after training
		# Convert and write JSON object to file
		with open("values_x.json", "w") as fp:
			json.dump(p_mx.values, fp, indent=0)

		with open("values_o.json", "w") as fp:
			json.dump(p_mo.values, fp, indent=0)

		print(f"Iteration: {k*N:8d}")

if do_human_player:
	pp = PlayerProvider(p_hx, p_mo)
else:
	pp = PlayerProvider(p_mx, p_mo)

play(pp, 10, True)