Python 技巧

 一. 列表、字典、集合、元组的使用

from random import randint, sample

# 列表解析
data = [randint(-10, 10) for _ in xrange(10)]

filter(lambda x: x >= 0, data)
[x for x in data if x >= 0]      #最快速

# 字典解析
d = {x: randint(60, 100) for x in xrange(1,21)}

{k : v for k, v in d.iteritems() if v > 90}

# 集合解析
s = set(data)
{x for x in s if x % 3 ==0}


# 元组
student = ('Jim', 16, 'male', '[email protected]')
# 1. enum
NAME, AGE, SEX, EMAIL = xrange(4)
print student[NAME]
# 2. 
from collections import namedtuple
Student = namedtuuple('Student', ['name', 'age', 'sex', 'email'])
s2 = Student('Tom', 16, 'mail', '[email protected]')
print s2.name


# 统计列表的重复元素
li = [randint(0, 20) for _ in xrange(30)]
d = dict.fromkeys(li, 0)
# 1. 
for x in li: d[x] += 1
# 2. 
from collections import Counter
d2 = Counter(li)
d2.most_common(3) #重复数最高的三个元素

# 字典根据值value排序
sorted([3, 1, 5]) #排序列表
(97, 'a') > (88, 'b') # 元组的比较，每个元素从开始比较

d = {x: randint(60, 100) for x in 'abcde'}
# 1.
data = zip(d.itervalues(), d.iterkeys())
sorted(data)
# 2. 
sorted(d.items(), key=lambda x: x[1])


# 多个字典中的公共键
sample('abcdefg', 3)
sample('abcdefg', randint(3,6)) # 随机取出几个元素

s1 = {x: randint(1,4) for x in sample('abcdefg', randint(3.6))}
s2 = {x: randint(1,4) for x in sample('abcdefg', randint(3.6))}
s3 = {x: randint(1,4) for x in sample('abcdefg', randint(3.6))}

# 1. 
res = []
for k in s1: if k in s2 and k in s3: res.append(k) # res.pop(k)

# 2. 使用集合
res = s1.viewkeys() & s2.viewkeys() & s3.viewkeys()

# 3. 
m1 = map(dict.viewkeys, [s1, s2, s3])
res = reduce(lambda a, b: a & b, m1)


# 保持字典有序
d =  {'Jim':(1, 35), 'Leo':(2, 38), 'Tom':(3, 44)}

from collections import OrderedDict
d = OrderedDict() #按进入字典的顺序打印
d['Jim'] = (1,35)
d['Leo'] = (2,38)
d['Tom'] = (3,44)

from time import time
start = time()
raw_input() #等待输入
# ...
timesecond = time() - start


# 历史记录
# 1. 用队列存储
from collections import deque
q = deque([], 5)
q.append(1) # 达到长度后，先进先出

li = list(q) # 转换成列表类型

# 2. 将q存到文件中
import pickle
pickle.dump(q, open('test','w'))

q2 = pickle.load(open('test','r'))

二. 迭代器、生成器

# 实现可迭代对象、迭代器对象
# 用时访问 并封装到一个对象中

# 可迭代对象
li = [1,2,3,4]
str= 'abcde'

# 迭代器
iterl = iter(li)    # li.__iter__()
iters = iter(str)   # str.__getitem__()

iterl.next()

# 1. 城市天气的迭代器和可迭代对象
from collections import Iterable, Iterator
class WIterator(Iterator):
    def __init__(self, cities):
        self.cities = cities
        self.index  = 0

    def getWeather(city):
        import requests
        r = requests.get(u'http://wthrcdn.etouch.cn/weather_mini?city=' + city)
        data = r.json()['data']['forecast'][0]
        return '%s: %s, %s' % (city, data['low'], data['high'])

    def next(self):
        if self.index == len(self.cities)
            raise StopIteration
        city = self.cities[self.index]
        self.index += 1
        return self.getWeather(city)

class WIterable(Iterable):
    def __init__(self, cities):
        self.cities = cities

    def __iter__(self):
        return WIterator(self.cities)

for x in WIterable([u'北京', u'长沙', u'广州'])：
    print x


# 2. 使用生成器函数实现可迭代对象
def f():
    print 'in f(), 1'
    yield 1
    print 'in f(), 2'
    yield 2
g = f()   # g.__iter__()

for i in g: print i

class PrintNumbers:
    def __init__(self, start, end):
        self.start = start
        self.end   = end

    def isPrimeNum(self, k):
        if k % 2 == 0:
            return True
        else:
            return False

    def __iter__(self):
        for k in xrange(self.start, self.end+1):
            if self.isPrimeNum(k):
                yield k

for x in PrintNumbers(1, 100): print x


# 进行反向迭代
li = [1,2,3,4,5]
li.reverse()     # 改变原来列表
li[::-1]         # 切片，和原来列表等大的新列表

ll = li.reversed(li)  # 列表反向迭代 
for i in ll: print i


class FloatRange:
    def __init__(self, start, end, step=0.1)
        self.start = start
        self.end   = end
        self.step  = step
    def __iter__(self):
        t = self.start
        while t <= self.end:
            yield t
            t += self.step
    def __reversed__(self):
        t = self.enf
        while t >= self.start:
            yield t
            t -= self.step
# 正向迭代
for i in FloatRange(1.0, 4.0, 0.5): print x
# 反向迭代
for i in reversed(FloatRange(1.0, 4.0, 0.5)): print x



# 对迭代器做切片操作
from itertools import islice
# islice()

li = range(20)
t  = iter(li)
for x in islice(t, 5, 10): print x # 会消耗原来的迭代对象





# 在一个for中迭代多个可迭代对象

chinese = [randint(60,100) for _ in xrange(40)]
math    = [randint(60,100) for _ in xrange(40)]
english = [randint(60,100) for _ in xrange(40)]

for in in xrange(len(math)):
    print chinese[i] + math[i] + english[i]

total = []
# 并行多个可迭代对象
for c, m, e in zip(chiness, math, english)
    print c+m+e


# 1. 串连多个迭代对象
from itertools import chain
c1 = [randint(60,100) for _ in xrange(40)]
c2 = [randint(60,100) for _ in xrange(42)]
c3 = [randint(60,100) for _ in xrange(45)]

for s in chain(c1, c2, c3):
    if s > 90: print s

三. 字符串

# 拆分含多种分隔符的字符串

s = "fwerf sd123 ^sdf dfdsf*d dsf 123"
s.split(“xy”)         #默认以空格分割，或以参数分割

res = s.split(";")
map(lambda x: x.split("|"), res) # 以";"和"|"分割的二维数组
t = []
map(lambda x: t.extend(x.split("|")), res) # 二维元素放到t中 

# 1. 
def aSplit(s, ds):
    res = [s]

    for d in ds:
        t = []
        map(lambda x: t.extend(x.split(d)), res)
        res = t
    return res
print aSplit(s, " ^*")       # 会存在空的元素

# 2. 正则表达式
import re
re.split(r'[,;|]+', s)



# 判断字符串a是否以b开头或结尾
#s.startswith() s.endswith() 接收单个字符串或字符串元组
import os, stat
files = [name for name in os.listdir(".") if name.endswith(('.sh', '.py'))]



# 调整字符串中文本的格式
#日志中'yyyy-mm-dd' 改为 'mm/dd/yyyy'
import re
log = open("/var/log/dpkg.log").read()
re.sub('(\d{4})-(\d{2})-(\d{2})', r'\2/\3/\1', log)
re.sub('(?P\d{4})-(?P\d{2})-(?P\d{2})', r'\g/\g/\g', log)




# 多个小字符串拼接成大字符串
s1 = "abcde"
s2 = "12345"
s1 + s2                 # str.__add__(s1, s2) str.__gt__(s1, s2) 运算符重载

s = ""
for p in pl: s += p     # 变量多时，临时变量开销大，资源浪费


s.join(s1)              # 参数可为字符串，可为列表

li = ['avc', 123, 'xyz', 456]
''.join([str(x) for x in li]) #列表解析，会生成一个列表，开销大
''.join(str(x) for x in li)   #生成器, (str(x) for x in li) 作为参数是括号省略


 
# 字符串格式对齐
# str.ljust() str.rjust() str.center()

s = "abc"
s.ljust(10 ,'=')         # 左对齐，填充=
s.center(10)

format(s, '<20')         # 左对齐
format(s, '>20')         # 右对齐
format(s, '^20')         # 居中




# 去掉字符串中不需要的字符
s = ' -------sd   dfadf    2332   +++++++++'
s.strip(' -+')                 
s.lstrip()
s.rstirp()

# 删除固定位置的字符,拼接切片
s[:3]+ s[4:]

# 替换
s.replace('\t', '')

import re
re.sub('[\t\r]', '', s)

s = 'abc123e3rxyz'
#s.translate()
import string 
tr = string.maketrans('abcxyz', 'xyzabc')
s.translate(tr)

s = 'abc\rdfd\n234234\t'
s.translate(None, '\r\t\n')

 

四. 文件读写

# python2 str   unicode
# python3 bytes str

# python2
s = u'你好'
s.encode('utf8')     #存储到文件中的格式

f = open('hello.txt', 'w')
f.write(s.encode('utf8'))
f.close()

f = open('hello.txt', 'r')
t = f.read().decode('utf8')    # 你好
f.close()

# python3  字符串就是unicode
strb = b'asdfasdfsdg'
s = '你好'
f = open('hello2.txt', 'wt', encoding='utf8') # 自动完成编解码
f.write(s)
f.close()

f = open('hello2.txt', 'rt', encoding='utf8')
s = f.read()
f.close()



# 处理二进制文件  处理音频文件，将音量调小保存
f = open('demo.wav', 'rb')
info = f.read(44)               #文件头
import struct
struct.unpack('h',info[22:24])  #处理文件头 数据运算
struct.unpack('i',infi[24:28])

f.seek(0,2)
f.tell()
n = (f.tell()-44) /2

import array
buf = array.array('h', (0 for _ in xrange(n)))

f.seek(44)
f.readinto(buf)

for i in xrange(n): buf[i] /= 8

f2 = open('demo2.wav', 'wb')
f2.write(info)
buf.tofile(f2)
f2.close()



# 使用临时文件
# 自动删除，不占内存
from tempfile import TemporaryFile, NamedTemporaryFile
f = TemporaryFile()                        # 系统文件系统找不到
f.write('abcddee'*100000)
f.seek(0)
f.read(100)

ntf = NamedTemporaryFile(delete=False)     # 能找到文件，默认关闭以后会删除文件
fname = nft.name




# 设置文件的缓冲
# I/O 操作以块为单位，如4096字节一个块
f = open('test.txt', 'w'， buffering=2048)  # 全缓冲，要写满缓冲才会写到文件中
f = open('test.txt', 'w'， buffering=1)     # 行缓冲，\n就会写文件
f = open('test.txt', 'w'， buffering=1)     # 无缓冲，实时写
f.write('abc')




# 将文件映射到内存
import mmap

f = open('demo.bn','r+b')
f.fileno()

m = mmap.mmap(f.fileno(), 0, access=mmpa.ACCESS_WRITE, offset=mmap.PAGESIZE)
# 得到字节数组
m[4:8] = '\xff'*4              # 修改直接改变文件内容




# 读写csv数据
from urllib import urlretrieve
urlretrieve('http://table.finance.yahoo.com/table.csv?s=000001.sz', 'pingan.csv')

rf = open('pingan.csv', 'rb')
import csv
reader = csv.reader(rf)
header = reader.next()

wf = open('pingan_c.csv', 'wb')
writer = csv.writeer(wf)
writer.writerow(header)
rf.close()
wf.close()





# 读写json数据
import requests
import json

from record import Record
record = Record(channel=1)
audioData = record.record(2)

from secret import API_KEY, SECRET_KEY
authUrl = "https://openapi.baidu.com/oauth/2.0/token?grant_type=client_credentials&client_id=" + API_KEY + "&client_secret=" + SECRET_KEY

response = requests.get(authUrl)
res = json.loads(response.content)
token = res['access_token']

#百度语音识别
cuid = 'xxxxxxxxxxxxx'
srvUrl = 'http://vop.baidu.com/server_api?cuid=' + cuid + '&token=' + token
heepHeader = {'Content-Type': 'audio/wav; rate = 8000'}
response = requests.post(srvUrl, headers=httpHeader, data=audioData)
res = json.loads(response.content)
text = res['result'][0]

print text


# json.dumps()  python对象(列表、字典等)转换成json字符串
# json.dumps(data, sort_keys=True)
# json.loads()  json字符串转换成python对象

with open('demo.json', 'wb') as f:
    json.dump(l, f)  # 将l数据写到文件






# 构建xml文档
from xml.etree.ElementTree import parse
with open('demo.xml') with f:
    et = parse(f)
root = et.getroot()
root.tag
root.attrib
root.text

#root.getchildren()
for child in root:
    print child.get('name')

root.find('country')
root.findall('country')                  # 直接子元素
for e in root.iterfind('country'):
    print e.get('name')


from xml.etree.ElementTree import Element, ElementTree, tostring
e = Element('Data')
e.set('name', 'abc')

e2 = Element('Row')
e3 = Element('Open')
e3.text = '8.80'
e2.append(e3)

e.append(e2)

tostring(e)

et = ElementTree(e)
et.write('demo.xml')





# 读写excel文件
import xlrd, xlwt

book = xlrd.open_workbook('demo.xls')
book.sheets()

sheet = book.sheet_by_index(0)
rows = sheet.nrows
cols = sheet.ncols

cell = sheet.cell(0,0)               #(0,0)单元格
cell.ctype
cell.value

row = sheet.row(1)                   #cell对象列表
data = sheet.row_values(1, 1)        #第1列跳过第一格的值列表

sheet.put_cell(0, cols, xlrd.XL_CELL_TEXT, u'Total', None)

wbook = xlwt.Workbook()
wsheet = wbook.add_sheet('sheet1')
style = xlwt.easyxf('align: vertical center, horizontal center')
wsheet.write(rows,cols, sheet.cell_value(rows,cols), style)
wsheet.save('output.xls')

 

五. 派生内置不可变类型并修改其实例化行为

class IntTuple(tuple):
    def __new__(cls, iterable):             #先于__init__()调用
        g = (x for x in iterable if isinstance(x, int) and x > 0)
        super(IntTuple, cls).__new__(cls, g)

    def __init__(self, iterable):
        # 此时如果过滤iterable 无法过滤成功
        super(IntTuple, self).__init__(iterable)

t = IntTuple([1, -1, 'abc', 6, ['x', 'y'], 3])
print t

 

六. 使用描述符对实例属性做类型检查

# 描述符： 包含 __get__() __set__() __delete__() 函数的类

class Attr(object):
    def __init__(self, name, type_):
        self.name = name
        self.type_= type_

    def __get__(self, instance, cls):
        return instanse.__dict__[self.name]

    def __set__(self, instance, value):
        if not isinstance(value, self.type_):
            raise TypeError('expected %s' % self.type_)
        instance.__dict__[self.name] = value

    def __delete__(self, instance):
        del instance.__dict__[self.name]


class Person(object):
    name = Attr('name', str)
    age  = Attr('age', int)
    hgt  = Attr('height', float)


p = Person()
p.name = 'Bob'
print p.name

p.age = '17'      #会抛出异常

 

七.  在环状数据结构中管理内存

import sys

class A(object):
    def __del__(self):                  # 当引用次数变为0时，调用析构函数
        print 'in A.__del__'

a = A()
a2 = a
print sys.getrefcount(a) - 1            # 查看对象a的引用次数，参数名也引用了对象，要-1
del a 
del a2


# 循环引用
class Data(object):                     # Data类保存Node对象引用
    def __init__(self, value, owner):
        self.owner = owner
        self.value = value

    def __str__(self):
        return "%s's data, value is %s" % (self.owner, self.value)


    def __del__(self):
        print 'in Data.__del__'

class Node(object):                       # Node类调用Data对象
    def __init__(self, valu):
        self.data = Data(value, self)

    def __del__(self):
        print 'in Node.__del__'

node = Node(100)
del node                                  # 此时Data Node不会被回收掉
raw_input('wait...')


# 使用弱引用
import weakref
a_wref = weakref.ref(a)
a2 = a_wref()

class Data(object):                     # Data类保存Node对象引用
    def __init__(self, value, owner):
        self.owner = weakref.ref(owner) # 弱引用
        self.value = value

    def __str__(self):
        return "%s's data, value is %s" % (self.owner(), self.value)


    def __del__(self):
        print 'in Data.__del__'

node2 = node(100)
del node2                                  # 此时Data Node将被回收

 

八. 通过实例方法名的字符串调用方法

# Circle Triangle Trctangle 求面积的方法名都不同
# 通过传方法名来调用不同的方法

# 1. getattr 获取对象属性，方法名也是属性

from lib1 import Circle
from lib2 import Triangle
from lib3 import Tectangle

def getArea(shape):
    for name in ('area', 'getArea', get_area):
        f = getattr(shape, name , None)
        if f:
            return f()

shape1 = Circle(2)
shape2 = Tirangle(3,4,5)
shape3 = Rectangle(6,4)

shapes = [shape1, shape2, shape3]
print map(getArea, shapes)


# 2. 使用opreator标准库

from opreator import methodcaller
s = "abc123abc456"
s.find('abc', 4)

methodcaller('find', 'abc', 4)(s)

 

九. 为创建大量实例节省内存

class Player(object):
    def __init__(self, uid, name, status=0, level=1):
        self.uid = uid
        self.name = name
        self.stat = status
        self.level = level

class Player2(object):
    __slots__ = ['uid', 'name', 'stat', 'level']

    def __init__(self, uid, name, status=0, level=1):
        self.uid = uid
        self.name = name
        self.stat = status
        self.level = level


p1 = Player('0001', 'Jim')
p2 = Player2('0002', 'Tom')

# p1 bi p2 多两个属性 __dict__  __weakref__
# __dict__ 字典，为实例动态绑定解除新属性

# p2 则不能动态绑定属性
# __slots__ 阻止了该功能

 

十. 让对象支持上下文管理

# 要使用上下文管理，类中要定义 __enter__ __exit__方法，分别在with开始和结束时调用
class test(object):
    ...

    def __enter__(self):
        pass

    def __exit__(self, exc_type, exc_val, exc_tb):
        pass

with test() as k:
    pass

 

十一. 类的比较操作
 

# 重定义运算符
# <    __lt__()
# >    __rt__()
# <=   __le__()
# >=   __re__()

# 添加装饰器， 只需重定义__eq__ 及以上任意一个

from functools import total_ordering

@total_ordering
class Rectangle(object):
    def __init__(self, w, h):
        self.w = w
        self.h = h

    def area(area):
        return self.w * self.h

    def __lt__(self, obj):
        print 'in __lt__'
        return self.area() < obj.area()

    def __eq__(self, obj):
        print 'in __eq__'
        return self.area() == obj.area()

class Circle(object):
    def __init__(self, r):
        self.r = r

    def area(area):
        return self.r ** 2 * 3.14

r1 = Retangle(5, 3)
r2 = Retangle(4, 4)
c1 = Circle(3)

print r1 >= r2
print r2 >= r3

 

十一. 多线程