Fix readability_stats

assayer/scorer.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 from assayer.models import ModelResult
+import re
 
 _model = None
 
@@ -40,11 +41,7 @@ def compute_similarity(results: list[ModelResult]) -> dict[tuple[str, str], floa
 
 
 def readability_stats(text: str) -> dict[str, float]:
-    sentences = [
-        s
-        for s in text.replace("!", ".").replace("?", ".").split(".")
-        if s.strip()
-    ]
+    sentences = [s for s in re.split(r"[.!?]+(?:\s|$)", text) if s.strip()]
     words = text.split()
     word_count = len(words)
     sentence_count = len(sentences) or 1

tests/test_scorer.py

@@ -78,3 +78,21 @@ def test_readability_stats_empty():
     assert stats["word_count"] == 0
     assert stats["sentence_count"] == 1
     assert stats["avg_sentence_length"] == 0.0
+
+
+def test_readability_stats_abbreviations():
+    stats = readability_stats("Dr. Smith scored 3.5. Well done.")
+    assert stats["word_count"] == 6
+    assert stats["sentence_count"] == 2
+
+
+def test_readability_stats_urls():
+    stats = readability_stats("Visit example.com for details.")
+    assert stats["word_count"] == 4
+    assert stats["sentence_count"] == 1
+
+
+def test_readability_stats_exclamations():
+    stats = readability_stats("The price is $3.99. Cheap!")
+    assert stats["word_count"] == 5
+    assert stats["sentence_count"] == 2