7.1. Iterator

7.1.1. Rationale

  • Used for iterating in a for loop

7.1.2. Protocol

  • __iter__(self) -> self

  • __next__(self) -> raise StopIteration

  • iter(obj) -> obj.__iter__()

  • next(obj) -> obj.__next__()

class Iterator:
    def __iter__(self):
        self._current = 0
        return self

    def __next__(self):
        if self._current >= len(self.values):
            raise StopIteration
        element = self.values[self._current]
        self._current += 1
        return element

7.1.3. Example

class Crew:
    def __init__(self):
        self.members = list()

    def __iadd__(self, other):
        self.members.append(other)
        return self

    def __iter__(self):
        self._current = 0
        return self

    def __next__(self):
        if self._current >= len(self.members):
            raise StopIteration
        result = self.members[self._current]
        self._current += 1
        return result


crew = Crew()
crew += 'Mark Watney'
crew += 'Jose Jimenez'
crew += 'Melissa Lewis'

for member in crew:
    print(member)

# Mark Watney
# Jose Jimenez
# Melissa Lewis

7.1.4. Loop and Iterators

For loop:

DATA = [1, 2, 3]

for current in DATA:
    print(current)

Intuitive implementation of the for loop:

DATA = [1, 2, 3]
iterator = iter(DATA)

try:
    current = next(iterator)
    print(current)

    current = next(iterator)
    print(current)

    current = next(iterator)
    print(current)

    current = next(iterator)
    print(current)
except StopIteration:
    pass

Intuitive implementation of the for loop:

DATA = [1, 2, 3]
iterator = DATA.__iter__()

try:
    current = iterator.__next__()
    print(current)

    current = iterator.__next__()
    print(current)

    current = iterator.__next__()
    print(current)

    current = iterator.__next__()
    print(current)
except StopIteration:
    pass

7.1.5. Built-in Type Iteration

Iterating str:

for character in 'hello':
    print(character)

# h
# e
# l
# l
# o

Iterating sequences:

for number in [1, 2, 3]:
    print(number)

# 1
# 2
# 3

Iterating dict:

DATA = {'a': 1, 'b': 2, 'c': 3}

for element in DATA:
    print(element)

# a
# b
# c

Iterating dict:

for key, value in DATA.items():
    print(f'{key} -> {value}')

# a -> 1
# b -> 2
# c -> 3

Iterating nested sequences:

for key, value in [('a',1), ('b',2), ('c',3)]:
    print(f'{key} -> {value}')

# a -> 1
# b -> 2
# c -> 3

7.1.6. Use Cases

Iterator implementation:

class Parking:
    def __init__(self):
        self._parked_cars = list()

    def park(self, car):
        self._parked_cars.append(car)

    def __iter__(self):
        self._current = 0
        return self

    def __next__(self):
        if self._current >= len(self._parked_cars):
            raise StopIteration
        element = self._parked_cars[self._current]
        self._current += 1
        return element


parking = Parking()
parking.park('Mercedes')
parking.park('Maluch')
parking.park('Toyota')

for car in parking:
    print(car)

# Mercedes
# Maluch
# Toyota

7.1.7. Standard Library Itertools

  • import itertools

itertools.count(start=0, step=1):

from itertools import count


data = count(3, 2)

next(data)
# 3

next(data)
# 5

next(data)
# 7

itertools.cycle(iterable):

from itertools import cycle

DATA = ['white', 'gray']

for color in cycle(DATA):
    print(color)

# white
# gray
# white
# gray
# ...

itertools.cycle(iterable):

from itertools import cycle

DATA = ['even', 'odd']

for i, status in enumerate(cycle(DATA)):
    print(i, status)

# 0, even
# 1, odd
# 2, even
# 3, odd
# ...

itertools.repeat(object[, times]):

from itertools import repeat

data = repeat(10, 3)
data
# repeat(10, 3)

next(data)
# 10

next(data)
# 10

next(data)
# 10

next(data)
# Traceback (most recent call last):
# StopIteration

itertools.accumulate(iterable[, func, *, initial=None]):

from itertools import accumulate

data = accumulate([1, 2, 3, 4])

next(data)
# 1

next(data)
# 3

next(data)
# 6

next(data)
# 10

next(data)
# Traceback (most recent call last):
# StopIteration

itertools.chain(*iterables):

from itertools import chain


keys = ['a', 'b', 'c']
values = [1, 2, 3]

for x in chain(keys, values):
    print(x)

# a
# b
# c
# 1
# 2
# 3

itertools.chain(*iterables):

from itertools import chain


class Iterator:
    def __iter__(self):
        self._current = 0
        return self

    def __next__(self):
        if self._current >= len(self.values):
            raise StopIteration
        element = self.values[self._current]
        self._current += 1
        return element


class Character(Iterator):
    def __init__(self, *values):
        self.values = values


class Number(Iterator):
    def __init__(self, *values):
        self.values = values


chars = Character('a', 'b', 'c')
nums = Number(1, 2, 3)

print(chain(chars, nums))
# <itertools.chain object at 0x116166970>

print(list(chain(chars, nums)))
# [1, 2, 3, 'a', 'b', 'c']

for x in chain(chars, nums):
    print(x)

# a
# b
# c
# 1
# 2
# 3

itertools.compress(data, selectors):

from itertools import compress


data = compress('ABCDEF', [1,0,1,0,1,1])

next(data)
# 'A'

next(data)
# 'C'

next(data)
# 'E'

next(data)
# 'F'

next(data)
# Traceback (most recent call last):
# StopIteration

itertools.islice(iterable, start, stop[, step]):

from itertools import islice


data = islice('ABCDEFG', 2, 6, 2 )

next(data)
# 'C'

next(data)
# 'E'

next(data)
# Traceback (most recent call last):
# StopIteration

itertools.starmap(function, iterable):

from itertools import starmap


data = starmap(pow, [(2,5), (3,2), (10,3)])

next(data)
# 32

next(data)
# 9

next(data)
# 1000

next(data)
# Traceback (most recent call last):
# StopIteration

itertools.product(*iterables, repeat=1):

from itertools import product


data = product(['a', 'b', 'c'], [1,2])

next(data)
# ('a', 1)

next(data)
# ('a', 2)

next(data)
# ('b', 1)

next(data)
# ('b', 2)

next(data)
# ('c', 1)

next(data)
# ('c', 2)

next(data)
# Traceback (most recent call last):
# StopIteration

itertools.permutations(iterable, r=None):

from itertools import permutations


data = permutations([1,2,3])

next(data)
# (1, 2, 3)

next(data)
# (1, 3, 2)

next(data)
# (2, 1, 3)

next(data)
# (2, 3, 1)

next(data)
# (3, 1, 2)

next(data)
# (3, 2, 1)

next(data)
# Traceback (most recent call last):
# StopIteration

itertools.combinations(iterable, r):

from itertools import combinations


data = combinations([1, 2, 3, 4], 2)

next(data)
# (1, 2)

next(data)
# (1, 3)

next(data)
# (1, 4)

next(data)
# (2, 3)

next(data)
# (2, 4)

next(data)
# (3, 4)

next(data)
# Traceback (most recent call last):
# StopIteration

itertools.combinations_with_replacement(iterable, r):

from itertools import combinations_with_replacement


data = combinations_with_replacement([1,2,3], 2)

next(data)
# (1, 1)

next(data)
# (1, 2)

next(data)
# (1, 3)

next(data)
# (2, 2)

next(data)
# (2, 3)

next(data)
# (3, 3)

next(data)
# Traceback (most recent call last):
# StopIteration

itertools.groupby(iterable, key=None). Make an iterator that returns consecutive keys and groups from the iterable. Generally, the iterable needs to already be sorted on the same key function. The operation of groupby() is similar to the uniq filter in Unix. It generates a break or new group every time the value of the key function changes. That behavior differs from SQL’s GROUP BY which aggregates common elements regardless of their input order:

from itertools import groupby

data = groupby('AAAABBBCCDAABBB')

next(data)
# ('A', <itertools._grouper object at 0x1215f5c70>)

next(data)
# ('B', <itertools._grouper object at 0x12157b4f0>)

next(data)
# ('C', <itertools._grouper object at 0x120e16ee0>)

next(data)
# ('D', <itertools._grouper object at 0x1215ef4c0>)

next(data)
# ('A', <itertools._grouper object at 0x12157b3a0>)

next(data)
# ('B', <itertools._grouper object at 0x12157b790>)

next(data)
# Traceback (most recent call last):
# StopIteration

[k for k, g in groupby('AAAABBBCCDAABBB')]
# A B C D A B

[list(g) for k, g in groupby('AAAABBBCCD')]
# AAAA BBB CC D

7.1.8. Assignments

Code 7.1. Solution
"""
* Assignment: Protocol Iterator Implementation
* Complexity: easy
* Lines of code: 14 lines
* Time: 13 min

English:
    1. Use data from "Given" section (see below)
    2. Modify classes to implement iterator protocol
    3. Iterator should return instances of `Mission`
    4. All tests must pass
    5. Run doctests - all must succeed

Polish:
    1. Użyj danych z sekcji "Given" (patrz poniżej)
    2. Zmodyfikuj klasy aby zaimplementować protokół iterator
    3. Iterator powinien zwracać instancje `Mission`
    4. Wszystkie testy muszą przejść
    5. Uruchom doctesty - wszystkie muszą się powieść

Tests:
    >>> from inspect import isclass, ismethod
    >>> assert isclass(Astronaut)

    >>> astro = Astronaut('Mark', 'Watney')
    >>> assert hasattr(astro, 'firstname')
    >>> assert hasattr(astro, 'lastname')
    >>> assert hasattr(astro, 'missions')
    >>> assert hasattr(astro, '__iter__')
    >>> assert hasattr(astro, '__next__')
    >>> assert ismethod(astro.__iter__)
    >>> assert ismethod(astro.__next__)

    >>> astro = Astronaut('Jan', 'Twardowski', missions=(
    ...     Mission(1969, 'Apollo 11'),
    ...     Mission(2024, 'Artemis 3'),
    ...     Mission(2035, 'Ares 3'),
    ... ))

    >>> for mission in astro:
    ...     print(mission)
    Mission(year=1969, name='Apollo 11')
    Mission(year=2024, name='Artemis 3')
    Mission(year=2035, name='Ares 3')
"""


# Given
from dataclasses import dataclass


@dataclass
class Astronaut:
    firstname: str
    lastname: str
    missions: tuple = ()


@dataclass
class Mission:
    year: int
    name: str


Code 7.2. Solution
"""
* Assignment: Protocol Iterator Range
* Complexity: medium
* Lines of code: 14 lines
* Time: 13 min

English:
    1. Use data from "Given" section (see below)
    2. Define class `Range` with parameters: `start`, `stop`, `step`
    3. Write own implementation of a built-in `range(start, stop, step)` function
    4. Assume, that user will never giv only one argument; always it will be either two or three arguments
    5. Use Iterator protocol
    6. All tests must pass
    7. Run doctests - all must succeed

Polish:
    1. Użyj danych z sekcji "Given" (patrz poniżej)
    2. Zdefiniuj klasę `Range` z parametrami: `start`, `stop`, `step`
    3. Zaimplementuj własne rozwiązanie wbudowanej funkcji `range(start, stop, step)`
    4. Przyjmij, że użytkownik nigdy nie poda tylko jednego argumentu; zawsze będą to dwa lub trzy argumenty
    5. Użyj protokołu Iterator
    6. Wszystkie testy muszą przejść
    7. Uruchom doctesty - wszystkie muszą się powieść

Tests:
    >>> from inspect import isclass, ismethod
    >>> assert isclass(Range)

    >>> r = Range(0, 0, 0)
    >>> assert hasattr(r, '__iter__')
    >>> assert hasattr(r, '__next__')
    >>> assert ismethod(r.__iter__)
    >>> assert ismethod(r.__next__)

    >>> list(Range(0, 10, 2))
    [0, 2, 4, 6, 8]

    >>> list(Range(0, 5))
    [0, 1, 2, 3, 4]
"""