python 精华梳理(已理解并手写)--全是干货--已结

基础部分

map,reduce,filter,sort,推导式，匿名函数lambda , 协程，异步io，上下文管理

自定义字符串转数字方法一不使用reduce

import re
def str2int(s):
    if not re.match(r'^\d+$',s):
        raise ValueError('请检查数字格式是否正确')
    else:
        digital={'0':0,"1":1,"2":2,"3":3,"4":4,"5":5,"6":6,"7":7,"8":8,"9":9}
        int_list = list(map(lambda k:digital[k] or 0,s))
        sum=0
        for i,v in enumerate(int_list):
            sum+=v*(10**(len(int_list)-i-1))

    return sum
print(str2int('456'))

自定义字符串转数字方法二使用reduce

from functools import reduce
import re
def str2int(s):
    if not re.match(r'^\d+$',s):
        raise ValueError('请检查数字格式是否正确')
    else:
        digital={'0':0,"1":1,"2":2,"3":3,"4":4,"5":5,"6":6,"7":7,"8":8,"9":9}
        int_list = list(map(lambda k:digital[k] or 0,s))

        return reduce(lambda a,b:a+b,[v*(10**(len(int_list)-i-1))for i,v in enumerate(int_list)])

print(str2int('456'))

import re
from functools import reduce
def digital(c):
    return {'0':0,'1':1,'2':2,'3':3,'4':4,'5':5,'6':6,'7':7,'8':8,'9':9}.get(c,0)

def str2int(s):
    if not re.match(r'^\d+$',s):
        raise TypeError('类型不是数字!')
    return reduce(lambda x,y:x+y,[v*(10**i) for i,v in enumerate(list(map(lambda c:digital(c),s))[::-1])])

print('%s-->%d' %(type(str2int('123')),str2int('123')))

assert isinstance(str2int('123'),int)

<class 'int'>-->123

让人眼前一亮的匿名函数

>>> for i in range(10):
...     print(list(map(lambda x:x(i),[lambda x:x+x,lambda x:x**2])))

列表推导式与 filter 相互等价

[i for i in range(30) if i%2==0]
list(filter(lambda x:x%2==0,range(30)))

# 0,1,2,3,4,5...
# 累加
accumulate = lambda x:x+fib(x-1) if x!=0 else 0
accumulate(3)
# 0,1,2,3,4,5...

# 阶乘
factorical = lambda x:x*factorical(x-1) if x!=1 else 1
factorical(6)

# 斐波拉契数列
# 0,1,1,2,3,5...
fib = lambda x: 1 if x==1 or x==2 else fib(x-1)+fib(x-2)
fib(6)

#生成器斐波拉契
def fib(x):
    n=0
    a,b = 0,1
    while n<x:
        b,a = a,a+b
        yield a
        n+=1

for i in fib(6):
    print(i,end=',')

#协程写斐波拉契
>>> def fib():
...     a,b = 0,1
...     r = 'start...'
...     yield r
...     while True:
...         yield b
...         a,b = b,a+b
...
>>> f = fib()
>>> f.send(None)
'start...'
>>> for i in range(1,10):
...     print('index {}-->{}'.format(i,f.send(i)))
...
index 1-->1
index 2-->1
index 3-->2
index 4-->3
index 5-->5
index 6-->8
index 7-->13
index 8-->21
index 9-->34
>>> f.close()

异步 io

>>> import asyncio
>>> @asyncio.coroutine
... def say(greet_str):
...     print(greet_str)
...     r = yield from asyncio.sleep(1)
...     print(greet_str,'again')
...     print('r:{}'.format(r))
...
>>> loop = asyncio.get_event_loop()
>>> tasks = [say('hello world!'),say('hello frank')]
>>> loop.run_until_complete(asyncio.wait(tasks))
hello frank
hello world!
hello frank again
r:None
hello world! again
r:None
({<Task finished coro=<say() done, defined at <stdin>:1> result=None>, <Task finished coro=<say() done, defined at <stdin>:1> result=None>}, set())
>>> loop.close()

3.5 及以后
把@asyncio.coroutine替换为async；
把yield from替换为await。
通过async定义的协程被称为原生协程跟
通过@asyncio.coroutine定义的协程函数被称为基于生成器的协程
两者不是一个概念

查找文件名包含某字符串的文件并输出器文件名

import os
def findstr_in_file(path):
    for child in os.listdir(path):
        qualified_child = os.path.join(path,child)
        if os.path.isdir(qualified_child):
                tmp = findstr_in_file(qualified_child)
                for item in tmp:
                    yield item
        else:
            yield qualified_child

for name in findstr_in_file('.'):
    if 'diabetes' in name:
        print(name)

os os.path 学习

#
os.environ.get('PATH')
if 'Linux' in  os.name:
    os.system('clear')
elif 'nt' in os.name:
    os.system('cls')

# 查看当前目录的绝对路径:
os.path.abspath('.')
os.path.join('/Users/michael', 'testdir')
os.mkdir('/Users/michael/testdir')
# 删掉一个目录:
os.rmdir('/Users/michael/testdir')
os.path.split('/Users/michael/testdir/demo.txt')
os.path.splitext('/Users/michael/testdir/demo.txt')
os.rename('text.txt','test.py')
os.remove('test.py')
[x for x in os.listdir('.') if os.path.isdir(x)] # 查出当前目录下子目录
[x for x in os.listdir('.') if os.path.isfile(x) and os.path.splitext(x)[1]=='.py'] # 查出当前目录下 py 文件

序列化与反序列化

import json

class Student(object):
    def __init__(self,name,age,score):
        self._name = name
        self._age = age
        self._score = score

    def __repr__(self):
        return 'Student(name={},age={},score={})'.format(self._name,self._age,self._score)

    def __str__(self):
        return self.__repr__()

    def student2dict(self,std):
        return {'name':std._name
                ,'age':std._age
                ,'score':std._score}

s = Student('frank','18',100)
json.dumps(s,default=s.student2dict)
json.dumps(s,default=lambda obj:obj.__dict__)

print(eval(s.__repr__()))

json_str = '{"name":"frank","age":18,"score":100}'

def dict2student(d):
    return Student(d['name'],d['age'],d['score'])
print(json.loads(json_str,object_hook=dict2student))

另一种形式的 迭代器，注意字符串也是序列

a = 'baidu'
try:
    g = iter(a)
    while True:
        print(next(g))
except StopIteration as stopIter:
    print(stopIter)
    print('no more!')

map filter lambda 综合点运用

from functools import reduce
import re
exp = '99 + 88 + 7.6 ='
reduce(lambda x,y:x+y,list(map(int,filter(lambda x: x if x else None,map(lambda s:s.replace('.','').strip(),re.split(r'[\+|=]',exp))))))

sorted 排序

>>> sorted(d.items(),key=lambda t:t[1])
[('g', 20), ('f', 21), ('e', 22), ('d', 23), ('c', 24), ('b', 25), ('a', 26)]
>>> dict(sorted(d.items(),key=lambda t:t[1]))
{'g': 20, 'f': 21, 'e': 22, 'd': 23, 'c': 24, 'b': 25, 'a': 26}
>>> {}.fromkeys([1,2,2,3,1,4,5])
{1: None, 2: None, 3: None, 4: None, 5: None}
>>> set([1,2,3,1,2,1,4,5])
{1, 2, 3, 4, 5}
>>> import os
>>> with open(os.path.join(baseDir,'test.txt')) as f:
...     for line in f:
...         print(line)

自己写冒泡+快速排序

import time
from functools import wraps
def time_count(func):
    @wraps(func)
    def execute_func(*args,**kw):
        t1 = time.time()
        result = func(*args,**kw)
        t2 = time.time()
        print('共花费:{} 秒'.format(t2-t1))
        return result
    return execute_func

l = [9,8,7,6,5,4,3,2,1]
@time_count
def bubble_sort(l):
    for i in range(len(l)):
        for j in range(len(l)-i-1):
            if l[j]>l[j+1]:
                l[j],l[j+1] = l[j+1],l[j]

    return l

bubble_sort(l)

@time_count
def quick_sort(array,low=0,high=None):
    high = len(array)-1 if high is None else high
    if low>=high:
        return
    l,r = low,high
    base_value = array[l]
    while l<r:
        while l<r and array[r]>=base_value:
            r-=1
        array[l] = array[r]
        while l<r and array[r]<base_value:
            l+=1
        array[r] = array[l]

        array[l] = base_value

        quick_sort(array,low,l)
        quick_sort(array,l+1,high)
    return array

l = [9,8,7,6,5,4,3,2,1]
quick_sort(l)

较为奢华版

from functools import wraps
from time import time
def time_count(func):
    @wraps(func)
    def wrapper_func(*args,**kw):
        start = time()
        result = func(*args,**kw)
        end = time()
        print('{} cost {} seconds in total...'.format(func.__name__,(end-start)))
        return result
    return wrapper_func

@time_count
def quick_sort(array,low=0,high=None):
    high = len(array)-1 if high is None else high
    if low>=high:
        return
    l,r = low,high
    base = array[l]
    while l<r:
        while all([l<r,base<=array[r]]):
            r-=1

        array[l] = array[r]

        while all([l<r,base>=array[l]]):
            l+=1

        array[r] = array[l]

        array[l] = base

        quick_sort(array,low,l-1)
        quick_sort(array,l+1,high)

    return array

array = [9,8,7,6,5,4,3,2,1]
quick_sort(array)

from functools import reduce
import re
out_str = """quick_sort cost 1.9073486328125e-06 seconds in total...
quick_sort cost 2.384185791015625e-06 seconds in total...
quick_sort cost 1.9073486328125e-06 seconds in total...
quick_sort cost 1.6689300537109375e-06 seconds in total...
quick_sort cost 1.9073486328125e-06 seconds in total...
quick_sort cost 0.0007488727569580078 seconds in total...
quick_sort cost 1.9073486328125e-06 seconds in total...
quick_sort cost 0.0014128684997558594 seconds in total...
quick_sort cost 0.002044200897216797 seconds in total...
quick_sort cost 1.9073486328125e-06 seconds in total...
quick_sort cost 0.0027604103088378906 seconds in total...
quick_sort cost 0.003480195999145508 seconds in total...
quick_sort cost 1.9073486328125e-06 seconds in total...
quick_sort cost 0.003975629806518555 seconds in total...
quick_sort cost 0.005151271820068359 seconds in total...
quick_sort cost 1.9073486328125e-06 seconds in total...
quick_sort cost 0.005714893341064453 seconds in total..."""

reduce(lambda x,y:float(x)+float(y),[item.strip() for item in re.findall(r'(\d.*)(?:second)',out_str)])

一句话简单版

quick_sort = lambda array: array if len(array) <= 1 else quick_sort([item for item in array[1:] if item <= array[0]]) + [array[0]] + quick_sort([item for item in array[1:] if item > array[0]])

l = [9,8,7,6,5,4,3,2,1]
quick_sort(l)

# 等价于：
def quick_sort2(arr):
    if len(arr) <= 1:
        return arr
    else:
        base = arr[0]
        left = [item for item in arr[1:] if item<base]
        right = [item for item in arr[1:] if item>=base]
        return quick_sort2(left)+[base]+quick_sort2(right)

l = [9,8,7,6,5,4,3,2,1]
quick_sort2(l)

比较精简版水仙花数(原创)

from functools import reduce
[x for x in range(100,1000) if reduce(lambda x,y:x+y,map(lambda i:int(i)**3,str(x))) == x]

杨辉三角之前准备，先打印一下99乘法表

>>> def g_mul():
...     for i in range(1,10):
...         l = []
...         for j in range(1,i+1):
...             l.extend('{}*{}={}\t'.format(j,i,j*i))
...         yield l
...
>>> for i in g_mul():
...     print(''.join(i))

一句话版99乘法表

print ('\n'.join([' '.join(['%s*%s=%-2s' % (y,x,x*y) for y in range(1,x+1)]) for x in range(1,10)]))
print('\n'.join([ '\t'.join([str(j)+'*'+str(i)+'='+str(j*i) for j in range(1,i+1)]) for i in range(1,10)]))
print('\n'.join([ '\t'.join(['{}*{}={}'.format(j,i,j*i) for j in range(1,i+1)]) for i in range(1,10)]))

杨辉三角

>>> def triangle(x):
...     l=[1]
...     while x>0:
...         yield l
...         l.append(0)
...         l = [l[i-1]+l[i] for i in range(len(l))]
...         x-=1
...
>>> for i in triangle(10):
...     print(i)

三元运算符的另一种写法

>>> a = 1
>>> b = (2,3)[a==1]
>>> b
3
>>> b = (2,3)[a==13]
>>> b
2

注意

reload(sys)
sys.setdefaultencoding("utf-8")12

在Python 3.x中不好使了 提示 name ‘reload’ is not defined

在3.x中已经被毙掉了被替换为

import importlib
importlib.reload(sys)

上下文管理

#方式一：
# 第1个参数保存异常类型；
# 第2个参数保存异常对象的值；
# 第3个参数保存异常的traceback信息。

class OpenObj(object):
    def __init__(self,filename,model):
        self.fp = open(filename,model)
    def __enter__(self):
        return self.fp
    def __exit__(self,exc_type,exc_val,exc_traceback):
        self.fp.close()

with OpenObj('./test.py','r') as f:
    for line in f:
        print(line)

#方式二：
from contextlib import contextmanager

@contextmanager
def make_open_context(filename,model):
    fp = open(filename,model)
    try:
        yield fp
    finally:
        fp.close()

with make_open_context('./test.py','r') as f:
    for line in f:
        print(line)

# 扩展 mysql 数据库游标
from pymysql import connect

class DBHelper:
    def __init__(self):
        self.conn = connect(host='localhost', port=3306, user='user', password='password', database='database', charset='utf8')
        self.csr = self.conn.cursor()

    def __enter__(self):
        return self.csr

    def __exit__(self, exc_type, exc_val, exc_tb):
        self.csr.close()
        self.conn.close()

with DBHelper() as csr:
    sql = """select * from table;"""
    csr.execute(sql)
    all_datas = csr.fetchall()

for item in all_datas:
    print(item)

@contextmanager
def conn_db():
    conn = connect(host='localhost', port=3306, user='user', password='password', database='database', charset='utf8')
    csr = conn.cursor()

    yield csr

    csr.close()
    conn.close()

with conn_db() as csr:
    sql = """select * from table"""
    csr.execute(sql)
    all_datas = csr.fetchall()

for item in all_datas:
    print(item)

深度遍历与宽度遍历

class Node(object):
    def __init__(self,sName):
        self._lChildren = []
        self.sName = sName
    def __repr__(self):
        return "<Node '{}'>".format(self.sName)
    def append(self,*args,**kwargs):
        self._lChildren.append(*args,**kwargs)
    def print_all_1(self):
        print(self)
        for oChild in self._lChildren:
            oChild.print_all_1()
    def print_all_2(self):
        def gen(o):
            lAll = [o,]
            while lAll:
                oNext = lAll.pop(0)
                lAll.extend(oNext._lChildren)
                yield oNext
        for oNode in gen(self):
            print(oNode)

oRoot = Node("root")
oChild1 = Node("child1")
oChild2 = Node("child2")
oChild3 = Node("child3")
oChild4 = Node("child4")
oChild5 = Node("child5")
oChild6 = Node("child6")
oChild7 = Node("child7")
oChild8 = Node("child8")
oChild9 = Node("child9")
oChild10 = Node("child10")

oRoot.append(oChild1)
oRoot.append(oChild2)
oRoot.append(oChild3)
oChild1.append(oChild4)
oChild1.append(oChild5)
oChild2.append(oChild6)
oChild4.append(oChild7)
oChild3.append(oChild8)
oChild3.append(oChild9)
oChild6.append(oChild10)

# 说明下面代码的输出结果

oRoot.print_all_1()
oRoot.print_all_2()

验证下parquet 与 csv 读写效率

import psycopg2
from contextlib import contextmanager
import pandas as pd
import numpy as np
import time
from functools import wraps
@contextmanager
def get_cur(**kw):
        conn = psycopg2.connect(**kw)
        crs = conn.cursor()
        yield crs
        crs.close
        conn.close()

field_list = ['person_id',xxx
                      'survival_36m', 'survival_60m', 'survival_120m', 'survival_months', 'recurrence_months',
                      'recurrence_site']

kw = dict(dbname='demodb'
     , host='devxxxom'
     , port=5439
     , user='user'
     , password='password')

def get_frame():
    with get_cur(**kw) as crs:
        sql = 'select "'+'","'.join(field_list)+'" from ml_sample limit 10000'
        print(sql)
        crs.execute(sql)
        df = pd.DataFrame(list(crs.fetchall()))
        df.columns = field_list
    return df

def save_file(df,filename,file_ex):
    if file_ex.strip()=='parquet':
#         from os import path
#         if path.exits(filename):
#             os.remove(filename)
        df.to_parquet(filename, compression='gzip')
    elif file_ex.strip() == 'csv':
        df.to_csv(filename)
    else:
        print('not implement~...')

#
#     pd.read_parquet('df.parquet.gzip')

def magnify_file(src,target,tn):
    for i in range(tn):
        print(src)
        with open(src,'rb') as fr:
            with open(target,'ab') as fw:
                for blk in fr:
                    fw.write(blk)

def time_count(func):
    @wraps(func)
    def func_wrapper(*args,**kw):
        start = time.time()
        result = func(*args,**kw)
        end = time.time()
        print('{} cost {} seconds...'.format(func.__name__,(end-start)))
        return result
    return func_wrapper

@time_count
def parquet_test(tn):
    df = get_frame()
    parquet_file = 'dataset.gzip'
    save_file(df,parquet_file,'parquet')
    from os import path
    parquet_file_magnif,_= path.splitext(parquet_file)
    magnify_file(parquet_file,parquet_file_magnif+'_magnif.gzip',tn)
    df = pd.read_parquet(parquet_file_magnif+'_magnif.gzip')
    df.head()

@time_count
def csv_test(tn):
    df = get_frame()
    csv_file = 'dataset.csv'
    save_file(df,csv_file,'csv')
    from os import path
    csv_file_magnif,_ = path.splitext(csv_file)
    print(csv_file_magnif)
    magnify_file(csv_file,csv_file_magnif+'_magnif.csv',tn)
    print('.....herer ....')
    df = pd.read_csv(csv_file_magnif+'_magnif.csv')
    df.head()

csv_test(10)
parquet_test(10)

csv_test(100)
parquet_test(100)

回文数的n种写法

def palindrome(s):
    return s == s[::-1]

def palindrome2(s):
    flag = True
    l,r = 0,len(s)-1
    while l<=r:
        if s[l] != s[r]:
            flag = False
        l+=1
        r-=1
    return flag

# 存在闭包问题，下面这种是错误的
def palindrome3(s,low=0,high=None):
    flag = True
    high = len(s)-1 if high is None else high
    l,r  = low, high
    if l<r:
        if s[l] == s[r]:
            palindrome3(s,l+1,r-1)
        else:
            flag = False

    return flag

palindrome3('aba')
True

# 解决了闭包问题
def palindrome3(s,low=0,high=None):
    global flag
    high = len(s)-1 if high is None else high
    l,r  = low, high
    print(l,'--',r)
    if l<r:
        if s[l] == s[r]:
            flag = True
            palindrome3(s,l+1,r-1)
        else:
            flag = False
            print(flag)
    return flag

argparse 模块用来做命令行工具挺方便

import argparse
import time
if __name__ == '__main__':
    start_time = time.time()
    #PATIENT SIMILARITY REQUEST INPUT
    parser = argparse.ArgumentParser(description="input to experiments")
    parser.add_argument('--organization_id', type=int, help='org id for filter')
    parser.add_argument('--save_cluster', type=int, help='if save cluster, please provide org_model_id')
    parser.add_argument('--threshold', type=float, help='initial threshold ')
    parser.add_argument('--bin', type=float, help='degree of freedom to do chi-squared test')
    parser.add_argument('--list_of_features', type=str, nargs='*',help='features could be demographic, tumor-related and gene related')

    parser.add_argument('--organization_model_id', type=int, help='organization_model_id from ml_organization_model table')
    args = parser.parse_args()
    print(args.list_of_features)
    end_time = time.time()
    print('cost {} s in total...'.format(end_time-start_time))

实用代码片段

# 屏蔽标准输出流标准错误流

def shield_print(func):
    @wraps(func)
    def main_func(*args,**kw):
        with open('outprint.log','w') as fw:
            old_stdout = sys.stdout
            old_stderr = sys.stderr
            sys.stdout = sys.stderr = fw
            result = func(*args,**kw)
            sys.stdout = old_stdout
            sys.stderr = old_stderr
            return result
    return main_func

自定义 精度条显示

import sys,time

def progress_bar(num,total):
    rate = float(num)/total
    rate_num = int(rate*100)
    out = '\r[{}{}] 已完成 {} %'.format('*'*rate_num,' '*(100-rate_num),rate_num)
    stdout = sys.stdout
    stdout.write(out)
    stdout.flush()

def test_progress_bar(start,end):
    i = start
    for i in range(end):
        time.sleep(0.1)
        progress_bar(i+1,end)

if __name__ == '__main__':
    start = 0
    end = 200
    test_progress_bar(start,end)

类似于 log4j

# -*- coding: utf-8 -*-
"""
 %(levelno)s:         打印日志级别的数值
 %(levelname)s:    打印日志级别名称
 %(pathname)s:    打印当前执行程序的路径，其实就是sys.argv[0]
 %(filename)s:      打印当前执行程序名
 %(funcName)s:    打印日志的当前函数
 %(lineno)d:         打印日志的当前行号
 %(asctime)s:      打印日志的时间
 %(thread)d:        打印线程ID
 %(threadName)s: 打印线程名称
 %(process)d:      打印进程ID
 %(message)s:    打印日志信息 

%y 两位数的年份表示（00-99）
%Y 四位数的年份表示（000-9999）
%m 月份（01-12）
%d 月内中的一天（0-31）
%H 24小时制小时数（0-23）
%I 12小时制小时数（01-12）
%M 分钟数（00=59）
%S 秒（00-59）
%a 本地简化星期名称
%A 本地完整星期名称
%b 本地简化的月份名称
%B 本地完整的月份名称
%c 本地相应的日期表示和时间表示
%j 年内的一天（001-366）
%p 本地A.M.或P.M.的等价符
%U 一年中的星期数（00-53）星期天为星期的开始
%w 星期（0-6），星期天为星期的开始
%W 一年中的星期数（00-53）星期一为星期的开始
%x 本地相应的日期表示
%X 本地相应的时间表示
%Z 当前时区的名称
%% %号本身

"""
import logging
import datetime
def getLog(path='.'):  

    dt=datetime.datetime.now()
    dtstr=dt.strftime('%Y-%m-%d')
    #dtstr=dt.strftime('%Y-%m-%d_%H_%M_%S')
    #创建一个logger实例
    logger = logging.getLogger("mllogger")
    logging.basicConfig(filemode='wa')
    hds=logger.handlers
    for h in hds:
        logger.removeHandler(h)
    logger.setLevel(logging.INFO)
    #创  建一个handler,用于写入日志文件，handler可以把日志内容写到不同的地方
    logName ="%s/%s.ml.log"%(path,dtstr)
    fh = logging.FileHandler(logName)
    fh.setLevel(logging.ERROR)
    #再  创建一个handler，用于输出控制台
    ch = logging.StreamHandler()
    ch.setLevel(logging.INFO)
    log_format=logging.Formatter("[%(asctime)s-%(filename)s:%(lineno)d]\n%(message)s",datefmt='%d %H:%M:%S')
    fh.setFormatter(log_format) #setFormatter() selects a Formatter object for this handler to use
    ch.setFormatter(log_format)
    logger.addHandler(fh)
    logger.addHandler(ch)  

    return logger

if __name__=='__main__':
    log=getLog()
    try:
        with open('./test.txt') as f:
            pass
    except (FileNotFoundError,EOFError) as e:
        log.error(e)

    log.info(a)
    log.info('test1')
    log.info('test2')
    log.info('test3')

自己写好的 工具 py 脚本 放到对应目录下去

有时候我们在liunx上想修改查看python的包路径可以试试以下命令

from distutils.sysconfig import get_python_lib

print(get_python_lib())

1024 == 2**10
65535 == 2**16-1
4294967295 == 2**32-1
[chr(i) for i in range(65,91)]+[chr(i) for i in range(97,123)]
b'\xe4\xb8\xad\xff'.decode('utf-8',errors='ignore')

不可变的tuple有什么意义？因为tuple不可变，所以代码更安全。如果可能，能用tuple代替list就尽量用tuple。

思考

eval   exec zip 的应用场景

不可变参数通过值传递，可变参数通过引用传递 list=[] 与 list =None 做参数的例子

上下文管理器+各种编码问题+异常处理深入

第三方模块

谷歌的 fire 模块
import dis 分析函数过程等...
from __future__ import printfunction 等 用来兼容 py2 与 py3
代码统计 cloc
excel 读写 pandas + xlrd , xlsxwriter
lxml
shutil
f-string

框架部分

相见恨晚