handle empty strings more consistently

rapidfuzz · Apr 17, 2023 · c8be909 · c8be909
1 parent fd4ba10
commit c8be909
Show file tree

Hide file tree

Showing 6 changed files with 18 additions and 3 deletions.
diff --git a/extern/rapidfuzz-cpp b/extern/rapidfuzz-cpp
diff --git a/src/rapidfuzz/distance/Jaro_py.py b/src/rapidfuzz/distance/Jaro_py.py
@@ -97,6 +97,9 @@ def similarity(
         s1 = processor(s1)
         s2 = processor(s2)
 
+    if not s1 and not s2:
+        return 1.0
+
     if score_cutoff is None:
         score_cutoff = 0
 

diff --git a/tests/distance/test_Jaro.py b/tests/distance/test_Jaro.py
@@ -10,7 +10,7 @@ def test_hash_special_case():
 
 
 def test_edge_case_lengths():
-    assert pytest.approx(Jaro.similarity("", "")) == 0
+    assert pytest.approx(Jaro.similarity("", "")) == 1
     assert pytest.approx(Jaro.similarity("0", "0")) == 1
     assert pytest.approx(Jaro.similarity("00", "00")) == 1
     assert pytest.approx(Jaro.similarity("0", "00")) == 0.833333

diff --git a/tests/distance/test_JaroWinkler.py b/tests/distance/test_JaroWinkler.py
@@ -10,7 +10,7 @@ def test_hash_special_case():
 
 
 def test_edge_case_lengths():
-    assert pytest.approx(JaroWinkler.similarity("", "")) == 0
+    assert pytest.approx(JaroWinkler.similarity("", "")) == 1.0
     assert pytest.approx(JaroWinkler.similarity("0", "0")) == 1
     assert pytest.approx(JaroWinkler.similarity("00", "00")) == 1
     assert pytest.approx(JaroWinkler.similarity("0", "00")) == 0.85

diff --git a/tests/distance/test_distance.py b/tests/distance/test_distance.py
@@ -43,6 +43,15 @@ def test_nan(scorer):
     assert scorer.normalized_similarity("test", float("nan")) == 0.0
 
 
+@pytest.mark.parametrize("scorer", all_scorer_modules)
+def test_empty_strings(scorer):
+    """
+    Test behavior when comparing two empty strings
+    """
+    assert scorer.normalized_distance("", "") == 0.0
+    assert scorer.normalized_similarity("", "") == 1.0
+
+
 @pytest.mark.parametrize("scorer", all_scorer_modules)
 def test_similar_array(scorer):
     """

diff --git a/tests/test_hypothesis.py b/tests/test_hypothesis.py
@@ -76,6 +76,9 @@ def normalize_distance(dist, s1, s2, weights=(1, 1, 1)):
 
 
 def jaro_similarity(pattern, text):
+    if not pattern and not text:
+        return 1.0
+
     P_flag = [0] * (len(pattern) + 1)
     T_flag = [0] * (len(text) + 1)