正則表達式 Python正則表達式指南

1. 正則表達式基礎

1.1. 簡單介紹

正則表達式並非Python的一部分。正則表達式是用於處理字符串的強大工具,擁有本身獨特的語法以及一個獨立的處理引擎,效率上可能不如str自帶的方法,但功能十分強大。得益於這一點,在提供了正則表達式的語言裏,正則表達式的語法都是同樣的,區別只在於不一樣的編程語言實現支持的語法數量不一樣;但不用擔憂,不被支持的語法一般是不經常使用的部分。若是已經在其餘語言裏使用過正則表達式,只須要簡單看一看就能夠上手了。html

下圖展現了使用正則表達式進行匹配的流程: 
re_simplegit

正則表達式的大體匹配過程是:依次拿出表達式和文本中的字符比較,若是每個字符都能匹配,則匹配成功;一旦有匹配不成功的字符則匹配失敗。若是表達式中有量詞或邊界,這個過程會稍微有一些不一樣,但也是很好理解的,看下圖中的示例以及本身多使用幾回就能明白。正則表達式

下圖列出了Python支持的正則表達式元字符和語法:   
pyre編程

1.2. 數量詞的貪婪模式與非貪婪模式

正則表達式一般用於在文本中查找匹配的字符串。Python裏數量詞默認是貪婪的(在少數語言裏也多是默認非貪婪),老是嘗試匹配儘量多的字符;非貪婪的則相反,老是嘗試匹配儘量少的字符。例如:正則表達式"ab*"若是用於查找"abbbc",將找到"abbb"。而若是使用非貪婪的數量詞"ab*?",將找到"a"。編程語言

1.3. 反斜槓的困擾

與大多數編程語言相同,正則表達式裏使用"\"做爲轉義字符,這就可能形成反斜槓困擾。假如你須要匹配文本中的字符"\",那麼使用編程語言表示的正則表達式裏將須要4個反斜槓"\\\\":前兩個和後兩個分別用於在編程語言裏轉義成反斜槓,轉換成兩個反斜槓後再在正則表達式裏轉義成一個反斜槓。Python裏的原生字符串很好地解決了這個問題,這個例子中的正則表達式可使用r"\\"表示。一樣,匹配一個數字的"\\d"能夠寫成r"\d"。有了原生字符串,你不再用擔憂是否是漏寫了反斜槓,寫出來的表達式也更直觀。工具

1.4. 匹配模式

正則表達式提供了一些可用的匹配模式,好比忽略大小寫、多行匹配等,這部份內容將在Pattern類的工廠方法re.compile(pattern[, flags])中一塊兒介紹。post

2. re模塊

2.1. 開始使用re

Python經過re模塊提供對正則表達式的支持。使用re的通常步驟是先將正則表達式的字符串形式編譯爲Pattern實例,而後使用Pattern實例處理文本並得到匹配結果(一個Match實例),最後使用Match實例得到信息,進行其餘的操做。url

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
# encoding: UTF-8
import  re
 
# 將正則表達式編譯成Pattern對象
pattern =  re. compile (r 'hello' )
 
# 使用Pattern匹配文本,得到匹配結果,沒法匹配時將返回None
match =  pattern.match( 'hello world!' )
 
if  match:
     # 使用Match得到分組信息
     print  match.group()
 
### 輸出 ###
# hello

re.compile(strPattern[, flag]):spa

這個方法是Pattern類的工廠方法,用於將字符串形式的正則表達式編譯爲Pattern對象。 第二個參數flag是匹配模式,取值可使用按位或運算符'|'表示同時生效,好比re.I | re.M。另外,你也能夠在regex字符串中指定模式,好比re.compile('pattern', re.I | re.M)與re.compile('(?im)pattern')是等價的。 
可選值有:code

  • re.I(re.IGNORECASE): 忽略大小寫(括號內是完整寫法,下同)
  • M(MULTILINE): 多行模式,改變'^'和'$'的行爲(參見上圖)
  • S(DOTALL): 點任意匹配模式,改變'.'的行爲
  • L(LOCALE): 使預約字符類 \w \W \b \B \s \S 取決於當前區域設定
  • U(UNICODE): 使預約字符類 \w \W \b \B \s \S \d \D 取決於unicode定義的字符屬性
  • X(VERBOSE): 詳細模式。這個模式下正則表達式能夠是多行,忽略空白字符,並能夠加入註釋。如下兩個正則表達式是等價的:
1
2
3
4
a =  re. compile (r """\d +  # the integral part
                    \.    # the decimal point
                    \d *  # some fractional digits""" , re.X)
b =  re. compile (r "\d+\.\d*" )

re提供了衆多模塊方法用於完成正則表達式的功能。這些方法可使用Pattern實例的相應方法替代,惟一的好處是少寫一行re.compile()代碼,但同時也沒法複用編譯後的Pattern對象。這些方法將在Pattern類的實例方法部分一塊兒介紹。如上面這個例子能夠簡寫爲:

1
2
m =  re.match(r 'hello' , 'hello world!' )
print  m.group()

re模塊還提供了一個方法escape(string),用於將string中的正則表達式元字符如*/+/?等以前加上轉義符再返回,在須要大量匹配元字符時有那麼一點用。

2.2. Match

Match對象是一次匹配的結果,包含了不少關於這次匹配的信息,可使用Match提供的可讀屬性或方法來獲取這些信息。

屬性:

  1. string: 匹配時使用的文本。
  2. re: 匹配時使用的Pattern對象。
  3. pos: 文本中正則表達式開始搜索的索引。值與Pattern.match()和Pattern.seach()方法的同名參數相同。
  4. endpos: 文本中正則表達式結束搜索的索引。值與Pattern.match()和Pattern.seach()方法的同名參數相同。
  5. lastindex: 最後一個被捕獲的分組在文本中的索引。若是沒有被捕獲的分組,將爲None。
  6. lastgroup: 最後一個被捕獲的分組的別名。若是這個分組沒有別名或者沒有被捕獲的分組,將爲None。

方法:

  1. group([group1, …]): 
    得到一個或多個分組截獲的字符串;指定多個參數時將以元組形式返回。group1可使用編號也可使用別名;編號0表明整個匹配的子串;不填寫參數時,返回group(0);沒有截獲字符串的組返回None;截獲了屢次的組返回最後一次截獲的子串。
  2. groups([default]): 
    以元組形式返回所有分組截獲的字符串。至關於調用group(1,2,…last)。default表示沒有截獲字符串的組以這個值替代,默認爲None。
  3. groupdict([default]): 
    返回以有別名的組的別名爲鍵、以該組截獲的子串爲值的字典,沒有別名的組不包含在內。default含義同上。
  4. start([group]): 
    返回指定的組截獲的子串在string中的起始索引(子串第一個字符的索引)。group默認值爲0。
  5. end([group]): 
    返回指定的組截獲的子串在string中的結束索引(子串最後一個字符的索引+1)。group默認值爲0。
  6. span([group]): 
    返回(start(group), end(group))。
  7. expand(template): 
    將匹配到的分組代入template中而後返回。template中可使用\id或\g<id>、\g<name>引用分組,但不能使用編號0。\id與\g<id>是等價的;但\10將被認爲是第10個分組,若是你想表達\1以後是字符'0',只能使用\g<1>0。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
import  re
m =  re.match(r '(\w+) (\w+)(?P<sign>.*)' , 'hello world!' )
 
print  "m.string:" , m.string
print  "m.re:" , m.re
print  "m.pos:" , m.pos
print  "m.endpos:" , m.endpos
print  "m.lastindex:" , m.lastindex
print  "m.lastgroup:" , m.lastgroup
 
print  "m.group(1,2):" , m.group( 1 , 2 )
print  "m.groups():" , m.groups()
print  "m.groupdict():" , m.groupdict()
print  "m.start(2):" , m.start( 2 )
print  "m.end(2):" , m.end( 2 )
print  "m.span(2):" , m.span( 2 )
print  r "m.expand(r'\2 \1\3'):" , m.expand(r '\2 \1\3' )
 
### output ###
# m.string: hello world!
# m.re: <_sre.SRE_Pattern object at 0x016E1A38>
# m.pos: 0
# m.endpos: 12
# m.lastindex: 3
# m.lastgroup: sign
# m.group(1,2): ('hello', 'world')
# m.groups(): ('hello', 'world', '!')
# m.groupdict(): {'sign': '!'}
# m.start(2): 6
# m.end(2): 11
# m.span(2): (6, 11)
# m.expand(r'\2 \1\3'): world hello!

2.3. Pattern

Pattern對象是一個編譯好的正則表達式,經過Pattern提供的一系列方法能夠對文本進行匹配查找。

Pattern不能直接實例化,必須使用re.compile()進行構造。

Pattern提供了幾個可讀屬性用於獲取表達式的相關信息:

  1. pattern: 編譯時用的表達式字符串。
  2. flags: 編譯時用的匹配模式。數字形式。
  3. groups: 表達式中分組的數量。
  4. groupindex: 以表達式中有別名的組的別名爲鍵、以該組對應的編號爲值的字典,沒有別名的組不包含在內。
1
2
3
4
5
6
7
8
9
10
11
12
13
import  re
p =  re. compile (r '(\w+) (\w+)(?P<sign>.*)' , re.DOTALL)
 
print  "p.pattern:" , p.pattern
print  "p.flags:" , p.flags
print  "p.groups:" , p.groups
print  "p.groupindex:" , p.groupindex
 
### output ###
# p.pattern: (\w+) (\w+)(?P<sign>.*)
# p.flags: 16
# p.groups: 3
# p.groupindex: {'sign': 3}

實例方法[ | re模塊方法]:

    1. match(string[, pos[, endpos]]) | re.match(pattern, string[, flags]): 
      這個方法將從string的pos下標處起嘗試匹配pattern;若是pattern結束時仍可匹配,則返回一個Match對象;若是匹配過程當中pattern沒法匹配,或者匹配未結束就已到達endpos,則返回None。 
      pos和endpos的默認值分別爲0和len(string);re.match()沒法指定這兩個參數,參數flags用於編譯pattern時指定匹配模式。 
      注意:這個方法並非徹底匹配。當pattern結束時若string還有剩餘字符,仍然視爲成功。想要徹底匹配,能夠在表達式末尾加上邊界匹配符'$'。 
      示例參見2.1小節。
    2. search(string[, pos[, endpos]]) | re.search(pattern, string[, flags]): 
      這個方法用於查找字符串中能夠匹配成功的子串。從string的pos下標處起嘗試匹配pattern,若是pattern結束時仍可匹配,則返回一個Match對象;若沒法匹配,則將pos加1後從新嘗試匹配;直到pos=endpos時仍沒法匹配則返回None。 
      pos和endpos的默認值分別爲0和len(string));re.search()沒法指定這兩個參數,參數flags用於編譯pattern時指定匹配模式。 
      1
      2
      3
      4
      5
      6
      7
      8
      9
      10
      11
      12
      13
      14
      15
      16
      # encoding: UTF-8
      import  re
       
      # 將正則表達式編譯成Pattern對象
      pattern =  re. compile (r 'world' )
       
      # 使用search()查找匹配的子串,不存在能匹配的子串時將返回None
      # 這個例子中使用match()沒法成功匹配
      match =  pattern.search( 'hello world!' )
       
      if  match:
           # 使用Match得到分組信息
           print  match.group()
       
      ### 輸出 ###
      # world
    3. split(string[, maxsplit]) | re.split(pattern, string[, maxsplit]): 
      按照可以匹配的子串將string分割後返回列表。maxsplit用於指定最大分割次數,不指定將所有分割。 
      1
      2
      3
      4
      5
      6
      7
      import  re
       
      p =  re. compile (r '\d+' )
      print  p.split( 'one1two2three3four4' )
       
      ### output ###
      # ['one', 'two', 'three', 'four', '']
    4. findall(string[, pos[, endpos]]) | re.findall(pattern, string[, flags]): 
      搜索string,以列表形式返回所有能匹配的子串。 
      1
      2
      3
      4
      5
      6
      7
      import  re
       
      p =  re. compile (r '\d+' )
      print  p.findall( 'one1two2three3four4' )
       
      ### output ###
      # ['1', '2', '3', '4']
    5. finditer(string[, pos[, endpos]]) | re.finditer(pattern, string[, flags]): 
      搜索string,返回一個順序訪問每個匹配結果(Match對象)的迭代器。 
      1
      2
      3
      4
      5
      6
      7
      8
      import  re
       
      p =  re. compile (r '\d+' )
      for  m in  p.finditer( 'one1two2three3four4' ):
           print  m.group(),
       
      ### output ###
      # 1 2 3 4
    6. sub(repl, string[, count]) | re.sub(pattern, repl, string[, count]): 
      使用repl替換string中每個匹配的子串後返回替換後的字符串。 
      當repl是一個字符串時,可使用\id或\g<id>、\g<name>引用分組,但不能使用編號0。 
      當repl是一個方法時,這個方法應當只接受一個參數(Match對象),並返回一個字符串用於替換(返回的字符串中不能再引用分組)。 
      count用於指定最多替換次數,不指定時所有替換。 
      1
      2
      3
      4
      5
      6
      7
      8
      9
      10
      11
      12
      13
      14
      15
      import  re
       
      p =  re. compile (r '(\w+) (\w+)' )
      s =  'i say, hello world!'
       
      print  p.sub(r '\2 \1' , s)
       
      def  func(m):
           return  m.group( 1 ).title() +  ' '  +  m.group( 2 ).title()
       
      print  p.sub(func, s)
       
      ### output ###
      # say i, world hello!
      # I Say, Hello World!
    7. subn(repl, string[, count]) |re.sub(pattern, repl, string[, count]): 
      返回 (sub(repl, string[, count]), 替換次數)。 
      1
      2
      3
      4
      5
      6
      7
      8
      9
      10
      11
      12
      13
      14
      15
      import  re
       
      p =  re. compile (r '(\w+) (\w+)' )
      s =  'i say, hello world!'
       
      print  p.subn(r '\2 \1' , s)
       
      def  func(m):
           return  m.group( 1 ).title() +  ' '  +  m.group( 2 ).title()
       
      print  p.subn(func, s)
       
      ### output ###
      # ('say i, world hello!', 2)
      # ('I Say, Hello World!', 2)
      轉載來自園內:  Python正則表達式指南