Update app.py

app.py

@@ -98,13 +98,13 @@ def create_mask_dict(entities, additional_masks=None):
     entity_counters = {}
     
     for entity in entities:
-            if entity['entity_group'] not in ['CARDINAL', 'EVENT', 'PERCENT', 'QUANTITY']:
-            if entity['word'] not in mask_dict:
-                if entity['entity_group'] not in entity_counters:
-                    entity_counters[entity['entity_group']] = 1
-                else:
-                    entity_counters[entity['entity_group']] += 1
-                mask_dict[entity['word']] = f"{entity['entity_group']}_{entity_counters[entity['entity_group']]}"
+        if entity['entity_group'] not in ['CARDINAL', 'EVENT', 'PERCENT', 'QUANTITY']:
+        if entity['word'] not in mask_dict:
+            if entity['entity_group'] not in entity_counters:
+                entity_counters[entity['entity_group']] = 1
+            else:
+                entity_counters[entity['entity_group']] += 1
+            mask_dict[entity['word']] = f"{entity['entity_group']}_{entity_counters[entity['entity_group']]}"
     
     if additional_masks:
         for word, replacement in additional_masks.items():
@@ -129,7 +129,9 @@ def mask_patterns(text):
         masks[email] = "<EMAIL>"
     
     # Phone pattern (Turkish)
-    phone_pattern = r"\+90\d{10}|\b\d{3}[-.\s]?\d{3}[-.\s]?\d{2}[-.\s]?\d{2}\b"
+    #phone_pattern = r"\+90\d{10}|\b\d{3}[-.\s]?\d{3}[-.\s]?\d{2}[-.\s]?\d{2}\b"
+    phone_pattern = r"\b(0?5\d{2}[-.\s]?\d{3}[-.\s]?\d{2}[-.\s]?\d{2}|\b5\d{3}[-.\s]?\d{3}[-.\s]?\d{2}[-.\s]?\d{2}|\b\d{3}[-.\s]?\d{3}[-.\s]?\d{2}[-.\s]?\d{2})\b"
+
     phones = re.findall(phone_pattern, text)
     for phone in phones:
         masks[phone] = "<PHONE>"