Extract hashtags from string (python 3.x)

extract_hashtags.py

import re
import unittest


def extract_tags(text: str) -> list:
    return re.compile(r"(?:^|\W)[＃#](?!\d\d)(?!\d$)(\w+)", re.UNICODE).findall(text)


class ExtractTagsTestCase(unittest.TestCase):

    def test_extract_tags(self):
        # start with tag
        self.assertEqual(
            extract_tags('#lorem ipsum\n\n'),
            ['lorem']
        )

        # have more tags
        self.assertEqual(
            extract_tags('#Lorem #ipsum dolor sit amet,\n\n'),
            ['Lorem', 'ipsum']
        )

        # with unicode
        self.assertEqual(
            extract_tags('#Lorem #ایپسوم dolor sit amet,\n\n'),
            ['Lorem', 'ایپسوم']
        )

        # with underscore
        self.assertEqual(
            extract_tags('#Lorem_a #ایپسوم_b dolor sit amet,\n\n'),
            ['Lorem_a', 'ایپسوم_b']
        )

        self.assertEqual(
            extract_tags('#Lorem_a #ipsum_a_and_b dolor sit amet,\n\n'),
            ['Lorem_a', 'ipsum_a_and_b']
        )

        # no special chars
        self.assertEqual(
            extract_tags('#Lorem #@ipsum dolor sit amet,\n\n'),
            ['Lorem']
        )
        
        # no punctuation
        self.assertEqual(
            extract_tags('#Lorem #2nd_ipsu\'m dolor sit amet,\n\n'),
            ['Lorem', '2nd_ipsu']
        )
        
        # tags should not fully match with digits
        self.assertEqual(
            extract_tags('#Lorem #2015 dolor sit amet,\n\n'),
            ['Lorem']
        )
        
        # tags can start with digit but should contain characters on next
        self.assertEqual(
            extract_tags('#Lorem #2nd_ipsum dolor sit amet,\n\n'),
            ['Lorem', '2nd_ipsum']
        )

        self.assertEqual(
            extract_tags('#Lorem #2nd_ipsum dolor sit amet,\n\n'),
            ['Lorem', '2nd_ipsum']
        )
        
        # tags should separate with spaces
        self.assertEqual(
            extract_tags('#Lorem#2_ipsum dolor sit amet,\n\n'),
            ['Lorem']
        )

        # next lines
        self.assertEqual(
            extract_tags('#Lorem#2_ipsum dolor sit amet,\n\n #next #line'),
            ['Lorem', 'next', 'line']
        )
        
        # Ignore brackets
        self.assertEqual(
            extract_tags('<#Lorem>(#2_ipsum) dolor sit amet,\n\n "#next" <#line> «#تگ»'),
            ['Lorem', '2_ipsum', 'next', 'line', 'تگ']
        )


if __name__ == '__main__':
    unittest.main()