diff options
| -rw-r--r-- | fuzzycat/utils.py | 4 | ||||
| -rw-r--r-- | tests/test_utils.py | 1 |
2 files changed, 3 insertions, 2 deletions
diff --git a/fuzzycat/utils.py b/fuzzycat/utils.py index 5ded48c..56aa467 100644 --- a/fuzzycat/utils.py +++ b/fuzzycat/utils.py @@ -1,5 +1,5 @@ -import itertools import io +import itertools import string printable_no_punct = string.digits + string.ascii_letters + string.whitespace @@ -26,6 +26,7 @@ def cut(f: int = 0, sep: str = '\t', ignore_missing_column: bool = True): return func + def author_similarity_score(u, v): """ Given two author strings, return a similarity score between 0 and 1. @@ -66,4 +67,3 @@ def nwise(iterable, n=2): while piece: yield piece piece = tuple(itertools.islice(i, n)) - diff --git a/tests/test_utils.py b/tests/test_utils.py index 1d27e96..46ffd0e 100644 --- a/tests/test_utils.py +++ b/tests/test_utils.py @@ -22,6 +22,7 @@ def test_cut(): with pytest.raises(ValueError): cut(3, sep=',', ignore_missing_column=False)("a,b,c") == "" + def test_author_similarity_score(): assert author_similarity_score("", "") == 0.0 assert author_similarity_score("Gregor Samsa", "G. Samsa") == 0.42857142857142855 |