tests/test_scrub.py


1
2
3
4
5
6
7
8
9
10
11
12
13
14

from fatcat_scholar.schema import *


def test_scrub():
    vectors = [
        (
            "“Please clean this piece… of text</b>„",
            '"Please clean this piece... of text"',
        ),
        ("<jats:p>blah", "blah"),
    ]

    for raw, fixed in vectors:
        assert fixed == scrub_text(raw)