average-stabw

2024-09-27 13:39:40 +03:00
parent 773a01cdb1
commit c9728e14a7
2 changed files with 4 additions and 2 deletions
--- a/nlp/normalize.py
+++ b/nlp/normalize.py
@@ -44,8 +44,10 @@ def normalize(text):
    """
    Normalize English text to resemble Russian characters.
    """
+    t = text.replace("  ", " ").replace("  ", " ").replace("  ", " ")
+
    # Segment the text first
-    t = segment_text(text.replace("  ", " ").replace("  ", " ").replace("  ", " "))
+    # t = segment_text(t)

    t = t.lower()