after fixing bugs

utils.py

@@ -1,4 +1,4 @@
-# utils.py
+import re
 from fuzzywuzzy import fuzz
 
 def extract_keyword(text: str, symptoms: list = None) -> str:
@@ -13,8 +13,10 @@ def extract_keyword(text: str, symptoms: list = None) -> str:
         words = question.split()
         if not words:
             return "Unknown"
+
         # Common words to skip
-        common_words = {'what', 'is', 'are', 'how', 'why', 'can', 'do', 'does', 'i', 'have', 'my', 'a', 'an', 'the', 'in', 'of', 'and', 'or', 'for', 'with', 'from', 'about', 'some', 'any', 'this', 'that', 'there', 'be', 'to', 'me', 'am', 'feel', 'feeling', 'experiencing', 'symptoms', 'issue', 'problem', 'cause', 'causes'}
+        common_words = {'what', 'is', 'are', 'how', 'why', 'can', 'do', 'does', 'i', 'have', 'my', 'a', 'an', 'the', 'in', 'of', 'and', 'or', 'for', 'with', 'from', 'about', 'some', 'any', 'this', 'that', 'there', 'be', 'to', 'me', 'am', 'feel', 'feeling', 'experiencing', 'symptoms', 'issue', 'problem', 'cause', 'causes', 'having', 'severe'} # Added 'having', 'severe'
+
         # Fuzzy matching against mb_symptoms (if provided)
         if symptoms:
             best_match = None
@@ -31,10 +33,12 @@ def extract_keyword(text: str, symptoms: list = None) -> str:
                         highest_score = score
             if best_match:
                 return best_match.capitalize()
+
         # Fallback: pick the first non-common word longer than 3 characters
         for word in words:
             if word.lower() not in common_words and len(word) > 3:
                 return word.capitalize()
+
         return words[0].capitalize() if words else "Unknown"
     else:
         # For symptoms or diseases, take the first word (or primary symptom if it's a list)