A regex to capture XML tags from a string. Useful for those times when accessing the DOM or using a full XML parser is not feasible (e.g. dirty scraped web data).

capture_tags_regex.py

import re

#capture a word chararacter or a / inside a <> tag, followed by anything
#except <, any number of times, non-greedy, plus any spaces around the tag
STANDALONE_TAG_REGEX = re.compile(r'(\s*<(?:[A-Za-z]+|/)[^<]*?>\s*)')

#a dictionary where keys are strings with fake xml tags in them,
#and values are the ordered list of tags that the regex should capture
REGEX_TESTS = {
    #wrapped
    'hello <a>world</a>!':            [' <a>', '</a>'],
    'hello<a>world</a>':              ['<a>', '</a>'],
    '<a>hello world</a>':             ['<a>', '</a>'],
    'hello <a>world</a> out there':   [' <a>', '</a> '],
    'hello <a>world</a>':             [' <a>', '</a>'],
    'hello <a href="blah">world</a>': [' <a href="blah">', '</a>'],

    #nested
    'hello <a><b>world</b></a>':                [' <a>', '<b>', '</b>', '</a>'],
    'hello <a><b href="blah">world</b></a>':    [' <a>', '<b href="blah">', '</b>', '</a>'],
    'hello <a href="blah"><b>world</b></a>':    [' <a href="blah">', '<b>', '</b>', '</a>'],
    'hello <a><a>world</a></a>':                [' <a>', '<a>', '</a>', '</a>'],
    '<a><b>word1</b>word2</a> word3 word4':     ['<a>', '<b>', '</b>', '</a> '],
    'hello <a><b><a>world</a></b></a>':         [' <a>', '<b>', '<a>', '</a>', '</b>', '</a>'],
    'hello <a>world</a> hello <b>everyone</b>': [' <a>', '</a> ', ' <b>', '</b>'],
    'hello <a></a>world':                       [' <a>', '</a>'],
    'hello <a href="blah"><b href="blah">world</b></a>': 
        [' <a href="blah">', '<b href="blah">', '</b>', '</a>'],

    #less than / greater than 
    'hello <3 <a>world</a>':      [' <a>', '</a>'],
    'hello <a>world</a> <3':      [' <a>', '</a> '],
    'hello E> <a>world</a>':      [' <a>', '</a>'],
    'hello <a>world</a> E>':      [' <a>', '</a> '],
    'hello <3 <a> world E> </a>': [' <a> ', ' </a>'],
    'hello <a> <3 world </a> E>': [' <a> ', ' </a> '],
    'hello <a> world </a> E>':    [' <a> ', ' </a> '],
    'hello <a> world E> </a>':    [' <a> ', ' </a>'],
    'hello <a> <3 world </a>':    [' <a> ', ' </a>'],
    'hello <a> <3 world E> </a>': [' <a> ', ' </a>'], 

    #empty
    'hello <a><b></b></a>world':  [' <a>', '<b>', '</b>', '</a>'],
    'hello <a><a></a></a>world':  [' <a>', '<a>', '</a>', '</a>'],

    #self-closing
    'hello <br /> world':         [' <br /> '],
    'hello <br /><a>world</a>':   [' <br />', '<a>', '</a>'],
    'hello<br/>world':            ['<br/>'],
  
    #realistic
    '<a>Shelf</a>life at <b>28</b><br />°C:<br />': 
        ['<a>', '</a>', ' <b>', '</b>', '<br />', '<br />'],
    '< 5 LEU/µL<x id=“locked568” xid=“lockTU_e1354968-d068-4cca-8203-7ef71227fad1"/> <x id=“1432”/>5-20 LEU/µl borderline':
        ['<x id=“locked568” xid=“lockTU_e1354968-d068-4cca-8203-7ef71227fad1"/> ', '<x id=“1432”/>'],
}

if __name__ == '__main__':
    for i, s in enumerate(REGEX_TESTS):
        result = re.findall(STANDALONE_TAG_REGEX, s)
        answer = REGEX_TESTS[s]
    if result != answer: 
        print('FAILURE.\nString:   {}\nResult:   {}\nExpected: {}\n'.format(s, result, answer))
    else:
        print('Success.\nString: {}\nResult: {}\n'.format(s, result))