NLP-正则表达式 Regular Expressions_nlp正则表达式

用实现定义好的字符串组成“规则字符串”，对文本进行过滤

1. Basic Patterns:

search() 返回object，且是找到的第一个例子

import re
 
text='Today is a nice day, and today is Monday.'
pattern='today'
match=re.search(pattern,text) #search()返回一个object

<re.Match object; span=(56, 62), match='London'>

findall() 找到所有符合对象，返回str

finditer() 返回object

for match in re.finditer(pattern,text):
    print(match) #return object
    print(match.span()) #return index

group() 返回actual text that matched

2. Simple Patterns:

[] Disjunction 

pattern='[Tt]oday' #Today or today

^ NOT 和 $

注：^主要有两个用处

a) 在[]开头表示否定(注意要在[]里面)

pattern='[^a-z]' #非小写的所有字符串

b) match the start of the line

 
text='Today is happy day. Today is Monday.'
pattern1='^The' # Start of Line match
pattern2='Monday$' # End of Line match
re.findall(pattern1,text) #return 第一个Today
re.findall(pattern2,text) #return Monday
 

？Question mark

# match one or zero optional character
text='apple and apples'
pattern='apples?' #?是跟着前面的char的
re.findall(pattern,text) #return 'apple' and 'apples' ?是跟着前面的char的

* Kleene

# match zero or many optional characters
text='b and baaaa and ba'
pattern='ba*' #*跟着前面的char,所以表示没有a或者后面很多重复a
re.findall(pattern,text) #return 'b' 'baaaa' 'ba'

+ Kleene

# match one or many optional characters
text='b and baaaa and ba'
pattern='ba*' #*跟着前面的char,所以表示有一个a或者后面很多重复a
re.findall(pattern,text) #return 'baaaa' 'ba'

. Wild Card

# matches any single character
text = 'care and core'
pattern= 'c.re'
re.findall(pattern,text) #return 'care' 和'core'

| OR 

re.search('apple|orange','I like apple')
re.search('cat(fish|nap|claw)','I like catfish and catnap')

\b 

text='the and other'
pattern=r'\bthe\b' # Word Boundary
re.findall(pattern,text) #只返回the

3. Identifiers for Characters

对于一些重复出现的字符，有特殊量词指代符号