sanitieswiki.py

import re
import os
import fileinput
import uuid

def getDocList(rootFoldr):
    filelst = []

    for(dirpath, _, files) in os.walk(rootFoldr):
        for filename in files:
            filepath = os.path.join(dirpath, filename)
            filelst.append(filepath)
    return (filelst)


def textsane(rootFoldr):
    flist= getDocList(rootFoldr)
    pattern = re.compile("<([^>]+)>")

    for each_file in flist:
        for line in fileinput.input(each_file, inplace=True):
            if not re.search(pattern, line):
                print line

textsane('data')