如何使用正则表达式从文件中提取文本？

Question

Hi I'm looking for a way to extract a part of a text file with Python using a Regex: here is my code:

    texfile=open("texte.txt", "r")
        for line in texfile:
            if re.match("^text(.*)", line):
               print line,

I'm seaching for the text following the word text until it reaches the end of paragraph or when it reach a white space,but my code return just the words which follow the word "text" on 1 line.

For Example :

bla bla hhhhhhhh text bla blajjjjjjjjjjjjjjjjjjjjj
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
ffff

Must return :

bla blajjjjjjjjjjjjjjjjjjjjj
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
ffff

---

thank you I tried all your codes but no one works as I want:I want now to render it simple,extract the stream which follow a certain "text" until it reach a blank line:

          text
    sssssssssssssssss
     ssssssss
    kkkk
    lllmmm

    kkkk

   ;must return 
    sssssssssssssssss
    ssssssss
    kkkk
    lllmmm
    ;because of the blank line

Answer 1

If you want to detect a part of a file extending on several lines, and if the file isn't too much gigantic, it's not a particularly good method to limit the power of regexes by examining one line at a time. When the file can be read and put entirely in RAM, it is better to analyze it with a regex exploring the text as a unique whole.

Note that '^' has significance "beginning of the string" if flag re.MULTILINE isn't used, and "beginning of a line" if this flag is used.

By the way, if you employ the match() method, you don't need to add "^" at the start of the RE pattern, since match() try to match from the very beginning of the string.

So, here's a manner to analyze the whole text as you seem to want it (I use splitlines(True) to obtain a list of the lines in the string ss , this list simulates a file° :

import re

ss = """   first line
    bli bli hhhhhhhh TEXT bla blajjjjjjjjj
hhhhhhhh  VVVVV
ZZZZZZ
    tttt
bolo bolo TEXTrumunu and badad
yyyyyyyyyyyyyyyy
kkkkkkkkkkk
jjjjjjjjjjjjjjj
   nnnn    uytr
      poiurrr
ahahahah bobobo
  ppppp TEXT aaaabbbbb cccccg    
      kmsms
TEXT fedex redex bidex
pududadi
A

no-whitespace-before-that
   hhrhezipo"""

regx = re.compile('TEXT *(.+(?<! )(?<!\r)(?:\n[^ ]+(?<!\n))?)')

for fnd in regx.findall(ss):
    print '\n'.join(map(repr,fnd.splitlines(True)))
    print '---------------------------------'

result:

'bla blajjjjjjjjj\n'
'hhhhhhhh'
---------------------------------
'rumunu and badad\n'
'yyyyyyyyyyyyyyyy\n'
'kkkkkkkkkkk\n'
'jjjjjjjjjjjjjjj'
---------------------------------
'aaaabbbbb cccccg'
---------------------------------
'fedex redex bidex\n'
'pududadi\n'
'A\n'
'\n'
'no-whitespace-before-that'
---------------------------------

.

If the file is gigantic and can't be charged in only one chunk in RAM, you can do:

import re

ss = """   first line
    bli bli hhhhhhhh TEXT bla blajjjjjjjjj
hhhhhhhh  VVVVV
ZZZZZZ
    tttt
bolo bolo TEXTrumunu and badad
yyyyyyyyyyyyyyyy
kkkkkkkkkkk
jjjjjjjjjjjjjjj
   nnnn    uytr
      poiurrr
ahahahah bobobo
  ppppp TEXT aaaabbbbb cccccg    
      kmsms
TEXT fedex redex bidex
pududadi
A

no-whitespace-before-that
   hhrhezipo"""

rigx = re.compile('TEXT *(.+\n?)')
li = []
for line in ss.splitlines(True):
    mat = rigx.search(line)
    if 'TEXT' in line:
        li.append(mat.group(1))
    elif ' ' in line and li:
        if not line.startswith(' '):
            li.append(line.split(' ')[0])
        li[-1] = li[-1].rstrip(' \r\n')
        print '\n'.join(map(repr,li))
        print '====================='
        li = []
    elif li:
        li.append(line)

This code gives the same result as the former one. You see it's less simple. That's because big-big files are more problematic.

Answer 2

This worked for me in python3:

for line in texfile:
    x = re.search("(.*?)(text)", line)
    try:
        print (x.group(1))
    except:
        print(line)

Answer 3

if you're not forced into using regexes you could use this...

Load the file as a list:

with open("texte.txt", "r") as fileInput:
    listLines = fileInput.readlines()

Get the index of the line that contains your keyword, if it exists more than once you might not get the expected result, but it's an easy fix:

listIndex = [i for i, item in enumerate(listLines) if "text" in item]

This are the lines that follow your keyword without blank lines, by slicing the list:

lines = [line for line in listLines[listIndex[0]:] if line]

You might want to get any text following you keyword with:

lineMatched = listLines[listIndex].split("text")[1].strip()

& print the result:

print "\n".join([lineMatched] + lines if lineMatched else lines)