自定义迭代器

Python 面向对象编程进阶

Jake Roach

Data Engineer

迭代器

允许遍历对象集合或数据流，并且一次返回一个项目的类

类似于 list、tuple，但行为不同
可用于导航、转换、生成
用 for 循环迭代
next() 函数

迭代器协议…

# Collection
chuck = NameIterator("Charles Carmicheal")
for letter in chuck:
    print(letter)

C
h
u
...

# Data stream
fun_game = DiceGame(rolls=3)
next(fun_game)  # ... and so on

迭代器协议

__iter__()

返回一个迭代器，此处返回自身引用
... return self

__next__()

返回集合或数据流中的下一个值
在此进行迭代、转换、生成

类要被视为迭代器，必须同时定义 __iter__() 和 __next__()！

示例迭代器

class CoinFlips:
    def __init__(self, number_of_flips):
        self.number_of_flips = number_of_flips  # Store the total number of flips
        self.counter = 0

    def __iter__(self):
        return self  # Return a reference of the iterator

    # Flip the next coin, return the output
    def __next__(self):
        if self.counter < self.number_of_flips:
            self.counter += 1
            return random.choice(["H", "T"])

使用示例迭代器

three_flips = CoinFlips(3)

# Flip the coin three times
next(three_flips)
next(three_flips)
next(three_flips)

H
H
T

遍历迭代器

three_flips = CoinFlips(3)

# Now, try to loop through every element of the iterator
for flip in three_flips:
    print(flip)

T
H
T
None
None
None
...

StopIteration

...
    def __next__(self):
        # Only do this if the coin hasn't been flipped "number_of_flips" times
        if self.counter < self.number_of_flips:
            self.counter += 1
            return random.choice(["H", "T"])

        else:  # Otherwise, stop execution with StopIteration
            raise StopIteration

标记集合/数据流结束
防止无限循环
易于处理

遍历迭代器

three_flips = CoinFlips(3)

# Now, try to loop through every element of the iterator
for flip in three_flips:
    print(flip)

H
T
H

处理 StopIteration 异常

while True:
    try:
        next(three_flips)  # Pull the next element of three_flips

    # Catch a stop iteration exception
    except StopIteration:
        print("Completed all coin flips!")
        break

H
H
H
Completed all coin flips!

练习时间！

Python 面向对象编程进阶