Merge pull request #76 from juliangilbey/jarowinkler-jellyfish

Modify JaroWinkler boosting to match behaviour of jellyfish algorithm
life4 · Oct 27, 2021 · 5785d46 · 5785d46
2 parents b5ee3c3 + a5a6ae0
commit 5785d46
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 4 deletions.
diff --git a/tests/test_edit/test_jaro_winkler.py b/tests/test_edit/test_jaro_winkler.py
@@ -14,7 +14,7 @@
 @pytest.mark.parametrize('left, right, expected', [
     ('elephant', 'hippo', 0.44166666666666665),
     ('fly', 'ant', 0.0),
-    ('frog', 'fog', 0.916666666),
+    ('frog', 'fog', 0.925),
     ('MARTHA', 'MARHTA', 0.9611111111111111),
     ('DWAYNE', 'DUANE', 0.84),
     ('DIXON', 'DICKSONX', 0.8133333333333332),

diff --git a/textdistance/algorithms/edit_based.py b/textdistance/algorithms/edit_based.py
@@ -251,8 +251,9 @@ def __call__(self, s1, s2, prefix_weight=0.1):
         if not s1_len or not s2_len:
             return 0.0
 
-        min_len = max(s1_len, s2_len)
-        search_range = (min_len // 2) - 1
+        min_len = min(s1_len, s2_len)
+        search_range = max(s1_len, s2_len)
+        search_range = (search_range // 2) - 1
         if search_range < 0:
             search_range = 0
 
@@ -294,7 +295,7 @@ def __call__(self, s1, s2, prefix_weight=0.1):
         # stop to boost if strings are not similar
         if not self.winklerize:
             return weight
-        if weight <= 0.7 or s1_len <= 3 or s2_len <= 3:
+        if weight <= 0.7:
             return weight
 
         # winkler modification