正則表達(dá)式(二)

re模塊(regex)

• python中沒(méi)有正則表達(dá)式的函數(shù)，需要引入內(nèi)置的re模塊

• re模塊方法:

  • match:從頭開(kāi)始匹配，只匹配一個(gè)結(jié)果
    • 作用:接收用戶(hù)傳遞過(guò)來(lái)的url進(jìn)行匹配(web框架用到)
  • search:整個(gè)字符串進(jìn)行匹配，只匹配第一個(gè)獲得的結(jié)果
  • findall:匹配整個(gè)字符串，返回一個(gè)列表對(duì)象
  • finditer:和findall非常類(lèi)似,返回一個(gè)可迭代的對(duì)象,迭代的時(shí)候才會(huì)創(chuàng)建值
  • split:字符串切割
  • sub:替換
  • 最常用match和findall來(lái)查找

• 分組的作用:

  • 在匹配得到的結(jié)果中獲取局部的值

• re模塊查找分類(lèi)(match,search,findall):

  • 普通匹配:
    • 通過(guò)match和search匹配成功返回的對(duì)象中,group得到的匹配得到的完整結(jié)果

  regex = 'H\w+'
  ret = re.match(regex, origin)
  print(ret.group())      # Hello
  print(ret.groups())     # ()
  print(ret.groupdict())  # {}
  

  • 分組匹配:
    • 分組來(lái)獲取局部的值
  • match:從頭開(kāi)始匹配

  origin = 'Hello tom, xxx tom, yyy tom haha 19'
  regex = '(H)(\w+)'
  ret = re.match(regex, origin)
  print(ret.group())      # Hello
  print(ret.groups())     # ('H', 'ello')
  print(ret.groupdict())  # {}
  

  origin = 'Hello tom, xxx tom, yyy tom haha 19'
  regex = '(?P<n1>H)(\w+)'
  ret = re.match(regex, origin)
  print(ret.group())      # Hello
  print(ret.groups())     # ('H', 'ello')
  print(ret.groupdict())  # {'n1': 'H'}
  

  • search:知道了group方法之后，searh方法和match方法是一樣的，match是從頭開(kāi)始匹配，search是全局匹配

  • findall:返回匹配成功的列表

    • 不分組:

    regex = 'app\w+'
    origin = 'i have an apple, i have a pen, apple pen'
    ret = re.findall(regex, origin)
    print(ret) # ['apple', 'apple']
    

    • 分組:
      • findall會(huì)把groups里面的所有東西放到列表里面(分組的作用：獲取局部的東西)
      • findall對(duì)?P<key>不起作用
      • 如果分組嵌套分組，則先取外面的，再取里面的

    egex = '(app\w+)'
    origin = 'i have an apple, i have a pen, apple pen'
    ret = re.findall(regex, origin)
    print(ret) # ['apple', 'apple']
    
    regex = '(app)\w+'
    origin = 'i have an apple, i have a pen, apple pen'
    ret = re.findall(regex, origin)
    print(ret) # ['app', 'app']
    
    regex = '(app)(\w+)'
    origin = 'i have an apple, i have a pen, apple pen'
    ret = re.findall(regex, origin)
    print(ret) # [('app', 'le'), ('app', 'le')]
    
    regex = '((a)(pp))(\w+)'
    origin = 'i have an apple, i have a pen, apple pen'
    ret = re.findall(regex, origin)
    print(ret) # [('app', 'a', 'pp', 'le'), ('app', 'a', 'pp', 'le')]
    

  • finditer:只有在迭代的時(shí)候才創(chuàng)建對(duì)象

    • 迭代的對(duì)象跟match和search獲得的對(duì)象一樣

  regex = '((a)(pp))(\w+)'
  origin = 'i have an apple, i have a pen, apple pen'
  
  ret1 = re.search(regex, origin)
  print('type of search', type(ret1)) 
  # type of search <class '_sre.SRE_Match'>
  ret2 = re.finditer(regex,origin)
  for i in ret2:
      print('type of finditer', type(i)) 
      # type of finditer <class '_sre.SRE_Match'>
  

• 分組注意事項(xiàng):

  • 分組的是按照你真是寫(xiě)了多少個(gè)組來(lái)確定的，如果獲得的值多于寫(xiě)的組個(gè)數(shù)，則去最后一個(gè)

  regex = '(\w)(\w)(\w)(\w)'
  ret = re.findall(regex, 'tony')
  print(ret) # [('t', 'o', 'n', 'y')]
  
  regex = '(\w){4}'
  ret = re.findall(regex, 'tony')
  print(ret) # [y]
  

  • 第一個(gè)寫(xiě)了4個(gè)組，每個(gè)組一個(gè)字符：(\w)(\w)(\w)(\w)，所以得到的結(jié)果是4個(gè)組
  • 第二個(gè)寫(xiě)的是1個(gè)組，需要匹配4個(gè)字符，因?yàn)檎嬲龑?xiě)的只有一個(gè)組，所有得到的結(jié)果為1個(gè)組，理論上每個(gè)字符的組都是一樣的，最后的會(huì)覆蓋前面的，所以取得的是最后一個(gè)

  ret = re.findall('\dyhh*', '1yhh2yhh3andyhh4yhh')
  print(ret) # ['1yhh', '2yhh', '4yhh']
  
  ret = re.findall('(\dyhh)*', '1yhh2yhh3andyhh4yhh')
  print(ret) # ['2yhh', '', '', '', '', '', '', '', '4yhh', '']
  

  • 匹配時(shí)盡量避免正則表達(dá)式為空值

• split:分割

  • 無(wú)分組分割：分割后不包含自己(分隔符)
  • 分組分割：分割后會(huì)保留分割符內(nèi)分組的內(nèi)容

  regex = '\d+'
  origin = 'ad12ad134a34da'
  ret = re.split(regex, origin)
  print(ret) # ['ad', 'ad', 'a', 'da']
  
  regex = '(\d+)'
  origin = 'ad12ad134a34da'
  ret = re.split(regex, origin, 1)
  print(ret) # ['ad', '12', 'ad134a34da']
  

  • 使用場(chǎng)景:計(jì)算器獲取最里層的括號(hào)同時(shí)取出括號(hào)

  origin = '7+((3+4)*5-(4-9)*5)*4'
  regex = '\(([^()]+)\)' # 獲取最里層的括號(hào)，同時(shí)對(duì)括號(hào)里的數(shù)據(jù)歸到組里，切割后保留括號(hào)內(nèi)的內(nèi)容
  ret = re.split(regex, origin, maxsplit=1)
  print(ret) # ['7+(', '3+4', '*5-(4-9)*5)*4']
  

  • 再根據(jù)切割后得到的結(jié)果是否是長(zhǎng)度為3的列表，不斷循環(huán)計(jì)算其值

  while True:
      ret = re.split(regex, origin, maxsplit=1)
      if len(ret) == 3:
          計(jì)算得到括號(hào)的值，然后重新組合origin字符串
      elif:
          break同時(shí)帶出結(jié)果
  

• sub:替換 && subn：替換返回一個(gè)元組