CDATA fix, Python - rextester

Language: Layout:

#python 2.7.12

import re

def callback(match):
    # Here you can further parse and/or do subs on 'group(2)'
    # the errant set of `]]>aa]]>]]>', etc.
    # For this we just cauterize it by enclosing in a comment block.
    return  match.group(1) + ''

xml = ' <article><Date>2011-04-22T10:09:18Z</Date><Heading><![CDATA[data - content ]]></Heading><Body><![CDATA[contene - data  Normal data. ]]></Body></article> <article><Date>2011-02-26T12:48:09Z</Date><Heading><![CDATA[Content]]></Heading><Body><![CDATA[ Content-abc Help ]]>  {#  #} Content data - MORE  style="position: relative; left: -35%;" /> ]]>  More   0) { %> 0) { %> 0) { %> ]]>]]></Body></article><article><Date>2011-04-22T10:09:18Z</Date><Heading><![CDATA[data - content ]]></Heading><Body><![CDATA[contene - data  Normal data. ]]></Body></article>'
xml = '<root>' + xml + '</root>'

xmlnew = re.sub( r'(?=(<!\[CDATA\[[\S\s]*?\]\]>))\1((?:(?:(?!<!\[CDATA\[[\S\s]*?\]\]>)[\S\s])*\]\]>)+)', callback, xml )    
    
print( xmlnew )

[ + ] Show input
Absolute running time: 0.07 sec, cpu time: 0.02 sec, memory peak: 4 Mb, absolute service time: 0,07 sec