[TOC]
summary: Reason is the light and the light of life.
toc: show
import re
from collections.abc import Iterator, Iterable
鸭子类型，通过协议实现
1. Iterable
可迭代协议: __iter__()
import re
class Sentence1:
def __init__(self, text):
self.text = text
self.words = re.findall("\w+", self.text)
def __iter__(self):
return self
s1 = Sentence1('Hello world!')
print(f"s is Iterable: {isinstance(s1, Iterable)}")
print(f"s is Iterator: {isinstance(s1, Iterator)}")
s is Iterable: True
s is Iterator: False
2. Iterator
迭代器协议: __iter__(), __next__()
import re
class Sentence2:
def __init__(self, text):
self.text = text
self.words = re.findall("\w+", self.text)
def __iter__(self):
return self
def __next__(self):
pass
s2 = Sentence2('Hello world!')
print(f"s is Iterable: {isinstance(s2, Iterable)}")
print(f"s is Iterator: {isinstance(s2, Iterator)}")
s is Iterable: True
s is Iterator: True
-
含有 __iter__() 的类对象是可迭代的。
-
含有 __iter__() 和 __next__() 是迭代器。
-
迭代器一定是可迭代的对象，可迭代的对象不一定是迭代器。
-
__iter__()方法返回迭代器。如果是迭代器本身，直接返回self。
-
__next__()方法返回下一个可用元素，否则抛出异常StopIteration
# 内置函数iter()和next()，等价于对象直接调用__iter__()和__next__()，应该避免直接调用。
iter(s2)
s2.__iter__()
next(s2)
s2.__next__()
3. 一个典型的迭代器
class SentenceIterator:
def __init__(self, words):
self.words = words
self.index = 0
def __iter__(self):
return self
def __next__(self):
try:
word = self.words[self.index]
except IndexError:
raise StopIteration()
self.index += 1
return word
class SentenceIterator:
def __init__(self, words):
self.words = words
self.index = 0
def __iter__(self):
return self
def __next__(self):
try:
word = self.words[self.index]
except IndexError:
raise StopIteration()
self.index += 1
return word
s = Sentence("Hello world!")
print(f"s is Iterable: {isinstance(s, Iterable)}")
print(f"s is Iterator: {isinstance(s, Iterator)}")
s is Iterable: True
s is Iterator: False
# for循环会默认自动调用对象s的__iter__()方法获取迭代器。并且会捕获StopIteration异常。
for w in s:
print(w)
for w in iter(s):
print(w)
Hello
world
Hello
world
next(s)
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-94-61c30b5fe1d5> in <module>
----> 1 next(s)
TypeError: 'Sentence' object is not an iterator
next(s.__iter__())
'Hello'
next(s.__iter__())
'Hello'
next(iter(s))
'Hello'
next(iter(s))
'Hello'
ss = SentenceIterator(["Hello", "world"])
next(ss)
'Hello'
next(ss)
'world'
next(ss)
---------------------------------------------------------------------------
IndexError Traceback (most recent call last)
<ipython-input-68-8245fc14b31c> in __next__(self)
19 try:
---> 20 word = self.words[self.index]
21 except IndexError:
IndexError: list index out of range
During handling of the above exception, another exception occurred:
StopIteration Traceback (most recent call last)
<ipython-input-112-d2551d6ade9e> in <module>
----> 1 next(ss)
<ipython-input-68-8245fc14b31c> in __next__(self)
20 word = self.words[self.index]
21 except IndexError:
---> 22 raise StopIteration()
23 self.index += 1
24 return word
StopIteration:
# 没有任何输出，因为之前迭代结束，ss迭代器不能恢复，需要重新构建迭代器。
for w in ss:
print(w)
ss = SentenceIterator(["Hello", "world"])
for w in ss:
print(w)
Hello
world
区别可迭代的对象和迭代器：
-
可迭代的对象存在__iter__()方法，用于实例化一个新的迭代器返回。
-
迭代器存在__iter__()方法，返回自身这个迭代器，同时存在__next()__方法，返回单个元素。
4. 更Pythonic的实现方式
用生成器函数替代SentenceIterator迭代器
class Sentence:
def __init__(self, text):
self.text = text
self.words = re.findall("\w+", self.text)
pass
def __iter__(self):
for word in self.words:
yield word
return