###########################################################
# JPython Standard Library
# Author: Alexander Tsepkov
# Copyright 2013 Pyjeon Software LLC
# License: Apache License    2.0
# This library is covered under Apache license, so that
# you can distribute it with your JPython applications.
###########################################################

# basic implementation of Python's 'random' library

# JavaScript's Math.random() does not allow seeding its random generator.
# To bypass that, this module implements its own version that can be seeded.
# I decided on RC4 algorithm for this.

_seed_state = {'key': [], 'key_i': 0, 'key_j': 0}


def _get_random_byte():
    _seed_state.key_i = (_seed_state.key_i + 1) % 256
    _seed_state.key_j = (_seed_state.key_j +
                         _seed_state.key[_seed_state.key_i]) % 256
    _seed_state.key[_seed_state.key_i], _seed_state.key[_seed_state.key_j] = \
            _seed_state.key[_seed_state.key_j], _seed_state.key[_seed_state.key_i]
    return _seed_state.key[(_seed_state.key[_seed_state.key_i] + \
            _seed_state.key[_seed_state.key_j]) % 256]


def seed(x=Date().getTime()):
    _seed_state.key_i = _seed_state.key_j = 0
    if jstype(x) is 'number':
        x = x.toString()
    elif jstype(x) is not 'string':
        raise TypeError("unhashable type: '" + jstype(x) + "'")
    for i in range(256):
        _seed_state.key[i] = i
    j = 0
    for i in range(256):
        j = (j + _seed_state.key[i] + x.charCodeAt(i % x.length)) % 256
        _seed_state.key[i], _seed_state.key[j] = _seed_state.key[
            j], _seed_state.key[i]


seed()


def random():
    n = 0
    m = 1
    for i in range(8):
        n += _get_random_byte() * m
        m *= 256
    return n / 0x10000000000000000


# unlike the python version, this DOES build a range object, feel free to reimplement
def randrange():
    return choice(range.apply(this, arguments))


def randint(a, b):
    return int(random() * (b - a + 1) + a)


def uniform(a, b):
    return random() * (b - a) + a


def choice(seq):
    if len(seq) > 0:
        return seq[Math.floor(random() * len(seq))]
    else:
        raise IndexError()


# uses Fisher-Yates algorithm to shuffle an array
def shuffle(x, random_f=random):
    x = list(x)
    for i in range(len(x)):
        j = Math.floor(random_f() * (i + 1))
        x[i], x[j] = x[j], x[i]
    return x


# similar to shuffle, but only shuffles a subset and creates a copy
def sample(population, k):
    x = list(population)[:]
    for i in range(len(population) - 1, len(population) - k - 1, -1):
        j = Math.floor(random() * (i + 1))
        x[i], x[j] = x[j], x[i]
    return x[-k:]