Update app.py

app.py

@@ -46,56 +46,71 @@ def parse_quiz_content(quiz_content):
     question_section = quiz_parts[0].strip()
     answer_key_section = quiz_parts[1].strip()
 
-    # --- Split into blocks by double newline ---
-    blocks = re.split(r'\n\n+', question_section) # Split by one or more double newlines
+    # --- Split into blocks by double newline, more robust splitting ---
+    blocks = re.split(r'\n\n+', question_section)
 
-    # --- Handle Potential Title (First Block) ---
+    # --- Handle Potential Title (First Block) - More robust title check ---
     if blocks:
         first_block = blocks[0].strip()
-        if not re.match(r'^Question \d+:', first_block, re.IGNORECASE): # Check if the first block DOES NOT start with "Question 1:", etc.
-            # Assume the first block is the title and remove it
-            blocks = blocks[1:] # Slice to remove the first block (title)
-        # else: # No warning needed if it starts with "Question", it's likely question 1
+        if not re.search(r'Question\s+\d+:', first_block, re.IGNORECASE): # Use re.search for more flexible title detection
+            blocks = blocks[1:] # Remove title block if it doesn't look like "Question 1:", etc.
 
 
     # --- Process Question Blocks ---
+    question_number = 1 # Keep track of expected question number for better parsing
+
     for block in blocks:
         block = block.strip()
-        if not block:
+        if not block: # Skip completely empty blocks
             continue
 
         lines = block.split('\n')
-        # --- First line might be "Question N:", extract question text from the next line ---
-        question_text = lines[0].strip() # Initially take the first line
-        if re.match(r'^Question \d+:', question_text, re.IGNORECASE): # If it's "Question N:"
-            if len(lines) > 1:
-                question_text = lines[1].strip() # Take the second line as the actual question
-            else:
-                question_text = "" # Handle case where "Question N:" line is there but no question text follows
-
-
+        question_text = ""
         options = {}
-        for line in lines[1:]: # Start from the second line to parse options (or from line 2 if "Question N:" was on line 1)
+        parsing_question_text = True # Flag to indicate we are parsing question text
+
+        for line_index, line in enumerate(lines):
             line = line.strip()
-            if re.match(r'^[A-D]\)\s(.+)', line): # Options are like "A) 4" - Capture text after "A) "
-                match = re.match(r'^([A-D])\)\s(.+)', line) # Capture letter and option text
+            if not line: # Skip empty lines within a question block
+                continue
+
+            if parsing_question_text:
+                if re.match(r'^Question\s+\d+:', line, re.IGNORECASE): # Check for "Question 1:", "Question 2:", etc.
+                    if len(lines) > line_index + 1: # Check if there's a line after "Question N:"
+                        question_text = lines[line_index + 1].strip() # Take the next line as question text
+                        parsing_question_text = False # Stop parsing question text, move to options
+                    else:
+                        st.warning(f"Warning: Question label '{line}' found, but no question text followed.")
+                        question_text = "" # No question text found after label
+                        parsing_question_text = False # Move to options parsing even if question text missing
+                else:
+                    # If the line is not "Question N:", and we are still parsing question text,
+                    # it might be the question text itself (in case of format variation)
+                    question_text = line # Consider this line as question text
+                    parsing_question_text = False # Move to options parsing
+
+
+            elif re.match(r'^[A-D]\)\s(.+)', line): # Parse options
+                match = re.match(r'^([A-D])\)\s(.+)', line)
                 if match:
                     option_letter = match.group(1)
                     option_text = match.group(2).strip()
                     options[option_letter] = option_text
 
 
-        if question_text and not question_text.lower().endswith("quiz"): # Avoid title as question, case-insensitive check
+        if question_text: # Only add question if we successfully extracted question text
             questions.append({'question': question_text, 'options': options})
+        else:
+            st.warning(f"Warning: Could not parse question text for block (potential issue in question {question_number}). Block content:\n{block}")
+
+        question_number += 1 # Increment expected question number
 
 
-    # --- Answer Key Parsing (Adjusted to only capture letter) ---
+    # --- Answer Key Parsing (No change needed) ---
     answer_lines = answer_key_section.strip().split('\n')
     for line in answer_lines:
         line = line.strip()
-        match = re.match(r'(\d+)\.\s*([A-D])\)', line) # Match "1. B)" or "1. B" - capture letter only
-        if not match: # If the above doesn't match, try to match without the closing parenthesis
-            match = re.match(r'(\d+)\.\s*([A-D])', line) # Match "1. B"
+        match = re.match(r'(\d+)\.\s*([A-D])', line)
         if match:
             question_num = int(match.group(1)) - 1
             correct_answer = match.group(2)
@@ -183,10 +198,10 @@ if topic:
                 prompt = f"""
                 Generate a multiple-choice quiz on the topic of "{topic}".
                 The quiz should have 5 questions.
-                For each question, label questions as "Question 1:", "Question 2:", etc.
-                For each question, provide four plausible options labeled A), B), C), and D).
+                Label each question as "Question 1:", "Question 2:", etc.
+                Provide four plausible options for each question, labeled A), B), C), and D).
                 Clearly indicate the correct answer for each question at the end in a section called "Answer Key:".
-                Format the quiz clearly for easy reading.
+                Format the quiz for easy reading.
                 """
                 response = model.generate_content(prompt)
                 quiz_content = response.text
@@ -197,8 +212,8 @@ if topic:
 
                 parsed_quiz_data, answer_key = parse_quiz_content(quiz_content)
 
-                # st.write("### Parsed Quiz Data:")
-                # st.write(parsed_quiz_data) # Display the parsed data structure
+                st.write("### Parsed Quiz Data:") # Added Parsed Data debug print back
+                st.write(parsed_quiz_data) # Display the parsed data structure
 
                 # --- END DEBUGGING PRINTS UNCOMMENTED ---