详解Python之可迭代对象,迭代器和生成器
时间：2022-04-16 10:01:36 | 栏目：Python代码 | 点击：次
一、概念描述
可迭代对象就是可以迭代的对象，我们可以通过内置的iter函数获取其迭代器，可迭代对象内部需要实现__iter__函数来返回其关联的迭代器;
迭代器是负责具体数据的逐个遍历的，其通过实现__next__函数得以逐个的访问关联的数据元素;同时通过实现__iter__来实现对可迭代对象的兼容;
生成器是一种迭代器模式，其实现了数据的惰性生成，即只有使用的时候才会生成对应的元素；
二、序列的可迭代性
python内置的序列可以通过for进行迭代，解释器会调用iter函数获取序列的迭代器，由于iter函数兼容序列实现的__getitem__，会自动创建一个迭代器;
迭代器的
import re
from dis import dis
class WordAnalyzer:
reg_word = re.compile('\w+')
def __init__(self, text):
self.words = self.__class__.reg_word.findall(text)
def __getitem__(self, index):
return self.words[index]
def iter_word_analyzer():
wa = WordAnalyzer('this is mango word analyzer')
print('start for wa')
for w in wa:
print(w)
print('start while wa_iter')
wa_iter = iter(wa)
while True:
try:
print(next(wa_iter))
except StopIteration as e:
break;
iter_word_analyzer()
dis(iter_word_analyzer)
# start for wa
# this
# is
# mango
# word
# analyzer
# start while wa_iter
# this
# is
# mango
# word
# analyzer
# 15 0 LOAD_GLOBAL 0 (WordAnalyzer)
# 2 LOAD_CONST 1 ('this is mango word analyzer')
# 4 CALL_FUNCTION 1
# 6 STORE_FAST 0 (wa)
#
# 16 8 LOAD_GLOBAL 1 (print)
# 10 LOAD_CONST 2 ('start for wa')
# 12 CALL_FUNCTION 1
# 14 POP_TOP
#
# 17 16 LOAD_FAST 0 (wa)
# 18 GET_ITER
# >> 20 FOR_ITER 12 (to 34)
# 22 STORE_FAST 1 (w)
#
# 18 24 LOAD_GLOBAL 1 (print)
# 26 LOAD_FAST 1 (w)
# 28 CALL_FUNCTION 1
# 30 POP_TOP
# 32 JUMP_ABSOLUTE 20
#
# 20 >> 34 LOAD_GLOBAL 1 (print)
# 36 LOAD_CONST 3 ('start while wa_iter')
# 38 CALL_FUNCTION 1
# 40 POP_TOP
#
# 21 42 LOAD_GLOBAL 2 (iter)
# 44 LOAD_FAST 0 (wa)
# 46 CALL_FUNCTION 1
# 48 STORE_FAST 2 (wa_iter)
#
# 23 >> 50 SETUP_FINALLY 16 (to 68)
#
# 24 52 LOAD_GLOBAL 1 (print)
# 54 LOAD_GLOBAL 3 (next)
# 56 LOAD_FAST 2 (wa_iter)
# 58 CALL_FUNCTION 1
# 60 CALL_FUNCTION 1
# 62 POP_TOP
# 64 POP_BLOCK
# 66 JUMP_ABSOLUTE 50
#
# 25 >> 68 DUP_TOP
# 70 LOAD_GLOBAL 4 (StopIteration)
# 72 JUMP_IF_NOT_EXC_MATCH 114
# 74 POP_TOP
# 76 STORE_FAST 3 (e)
# 78 POP_TOP
# 80 SETUP_FINALLY 24 (to 106)
#
# 26 82 POP_BLOCK
# 84 POP_EXCEPT
# 86 LOAD_CONST 0 (None)
# 88 STORE_FAST 3 (e)
# 90 DELETE_FAST 3 (e)
# 92 JUMP_ABSOLUTE 118
# 94 POP_BLOCK
# 96 POP_EXCEPT
# 98 LOAD_CONST 0 (None)
# 100 STORE_FAST 3 (e)
# 102 DELETE_FAST 3 (e)
# 104 JUMP_ABSOLUTE 50
# >> 106 LOAD_CONST 0 (None)
# 108 STORE_FAST 3 (e)
# 110 DELETE_FAST 3 (e)
# 112 RERAISE
# >> 114 RERAISE
# 116 JUMP_ABSOLUTE 50
# >> 118 LOAD_CONST 0 (None)
# 120 RETURN_VALUE
三、经典的迭代器模式
标准的迭代器需要实现两个接口方法，一个可以获取下一个元素的__next__方法和直接返回self的__iter__方法；
迭代器迭代完所有的元素的时候会抛出StopIteration异常，但是python内置的for、列表推到、元组拆包等会自动处理这个异常；
实现__iter__主要为了方便使用迭代器，这样就可以最大限度的方便使用迭代器；
迭代器只能迭代一次，如果需要再次迭代就需要再次调用iter方法获取新的迭代器，这就要求每个迭代器维护自己的内部状态，即一个对象不能既是可迭代对象同时也是迭代器；
从经典的面向对象设计模式来看，可迭代对象可以随时生成自己关联的迭代器，而迭代器负责具体的元素的迭代处理；
import re
from dis import dis
class WordAnalyzer:
reg_word = re.compile('\w+')
def __init__(self, text):
self.words = self.__class__.reg_word.findall(text)
def __iter__(self):
return WordAnalyzerIterator(self.words)
class WordAnalyzerIterator:
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
def iter_word_analyzer():
wa = WordAnalyzer('this is mango word analyzer')
print('start for wa')
for w in wa:
print(w)
print('start while wa_iter')
wa_iter = iter(wa)
while True:
try:
print(next(wa_iter))
except StopIteration as e:
break;
iter_word_analyzer()
# start for wa
# this
# is
# mango
# word
# analyzer
# start while wa_iter
# this
# is
# mango
# word
# analyzer
四、生成器也是迭代器
生成器是调用生成器函数生成的，生成器函数是含有yield的工厂函数；
生成器本身就是迭代器，其支持使用next函数遍历生成器，同时遍历完也会抛出StopIteration异常；
生成器执行的时候会在yield语句的地方暂停，并返回yield右边的表达式的值；
def gen_func():
print('first yield')
yield 'first'
print('second yield')
yield 'second'
print(gen_func)
g = gen_func()
print(g)
for val in g:
print(val)
g = gen_func()
print(next(g))
print(next(g))
print(next(g))
# <function gen_func at 0x7f1198175040>
# <generator object gen_func at 0x7f1197fb6cf0>
# first yield
# first
# second yield
# second
# first yield
# first
# second yield
# second
# StopIteration
我们可以将__iter__作为生成器函数
import re
from dis import dis
class WordAnalyzer:
reg_word = re.compile('\w+')
def __init__(self, text):
self.words = self.__class__.reg_word.findall(text)
def __iter__(self):
for word in self.words:
yield word
def iter_word_analyzer():
wa = WordAnalyzer('this is mango word analyzer')
print('start for wa')
for w in wa:
print(w)
print('start while wa_iter')
wa_iter = iter(wa)
while True:
try:
print(next(wa_iter))
except StopIteration as e:
break;
iter_word_analyzer()
# start for wa
# this
# is
# mango
# word
# analyzer
# start while wa_iter
# this
# is
# mango
# word
# analyzer
五、实现惰性迭代器
迭代器的一大亮点就是通过__next__来实现逐个元素的遍历，这个大数据容器的遍历带来了可能性；
我们以前的实现在初始化的时候，直接调用re.findall得到了所有的序列元素，并不是一个很好的实现；我们可以通过re.finditer来在遍历的时候得到数据；
import re
from dis import dis
class WordAnalyzer:
reg_word = re.compile('\w+')
def __init__(self, text):
# self.words = self.__class__.reg_word.findall(text)
self.text = text
def __iter__(self):
g = self.__class__.reg_word.finditer(self.text)
print(g)
for match in g:
yield match.group()
def iter_word_analyzer():
wa = WordAnalyzer('this is mango word analyzer')
print('start for wa')
for w in wa:
print(w)
print('start while wa_iter')
wa_iter = iter(wa)
wa_iter1= iter(wa)
while True:
try:
print(next(wa_iter))
except StopIteration as e:
break;
iter_word_analyzer()
# start for wa
# <callable_iterator object at 0x7feed103e040>
# this
# is
# mango
# word
# analyzer
# start while wa_iter
# <callable_iterator object at 0x7feed103e040>
# this
# is
# mango
# word
# analyzer
六、使用生成器表达式简化惰性迭代器
生成器表达式是生成器的声明性定义，与列表推到的语法类似，只是生成元素是惰性的；
def gen_func():
print('first yield')
yield 'first'
print('second yield')
yield 'second'
l = [x for x in gen_func()]
for x in l:
print(x)
print()
ge = (x for x in gen_func())
print(ge)
for x in ge:
print(x)
# first yield
# second yield
# first
# second
#
# <generator object <genexpr> at 0x7f78ff5dfd60>
# first yield
# first
# second yield
# second
使用生成器表达式实现word analyzer
import re
from dis import dis
class WordAnalyzer:
reg_word = re.compile('\w+')
def __init__(self, text):
# self.words = self.__class__.reg_word.findall(text)
self.text = text
def __iter__(self):
# g = self.__class__.reg_word.finditer(self.text)
# print(g)
# for match in g:
# yield match.group()
ge = (match.group() for match in self.__class__.reg_word.finditer(self.text))
print(ge)
return ge
def iter_word_analyzer():
wa = WordAnalyzer('this is mango word analyzer')
print('start for wa')
for w in wa:
print(w)
print('start while wa_iter')
wa_iter = iter(wa)
while True:
try:
print(next(wa_iter))
except StopIteration as e:
break;
iter_word_analyzer()
# start for wa
# <generator object WordAnalyzer.__iter__.<locals>.<genexpr> at 0x7f4178189200>
# this
# is
# mango
# word
# analyzer
# start while wa_iter
# <generator object WordAnalyzer.__iter__.<locals>.<genexpr> at 0x7f4178189200>
# this
# is
# mango
# word
# analyzer