Update app.py

app.py

@@ -164,12 +164,18 @@ def analyze_segment_with_gemini(cluster_text, is_full_text=False):
         prompt = f"""
             Analyze the following text (likely a transcript or document) and:
             
-            1. First, do text segmentation and identify DISTINCT key topics within the text
+            FIRST ASSESS THE TEXT:
+            - Check if it's primarily self-introduction, biographical information, or conclusion
+            - Check if it's too short or lacks meaningful content (less than 100 words of substance)
+            - If either case is true, respond with a simple JSON: {{"status": "insufficient", "reason": "Brief explanation"}}
+            
+            IF THE TEXT HAS SUFFICIENT MEANINGFUL CONTENT:
+            1. Identify DISTINCT key topics within the text
             2. For each segment/topic you identify:
                - Provide a SPECIFIC and UNIQUE topic name (3-5 words) that clearly differentiates it from other segments
                - List 3-5 key concepts discussed in that segment
                - Write a brief summary of that segment (3-5 sentences)
-               - Create 5 quiz questions based DIRECTLY on the content in that segment
+               - Create 5 quiz questions based DIRECTLY on the content in that segment, ONLY if the segment contains factual information
             
             For each quiz question:
             - Create one correct answer that comes DIRECTLY from the text
@@ -182,6 +188,7 @@ def analyze_segment_with_gemini(cluster_text, is_full_text=False):
             
             Format your response as JSON with the following structure:
             {{
+                "status": "processed",
                 "segments": [
                     {{
                         "topic_name": "Name of segment 1",
@@ -211,14 +218,27 @@ def analyze_segment_with_gemini(cluster_text, is_full_text=False):
                     // More segments...
                 ]
             }}
+            
+            OR if the text is just introductory, concluding, or insufficient:
+            {{
+                "status": "insufficient",
+                "reason": "Brief explanation of why (e.g., 'Text is primarily self-introduction', 'Text is too short', etc.)"
+            }}
         """
     else:
         prompt = f"""
-            Analyze the following text segment and provide:
+            Analyze the following text segment and:
+            
+            FIRST ASSESS THE TEXT:
+            - Check if it's primarily self-introduction, biographical information, or conclusion
+            - Check if it's too short or lacks meaningful content (less than 100 words of substance)
+            - If either case is true, respond with a simple JSON: {{"status": "insufficient", "reason": "Brief explanation"}}
+            
+            IF THE TEXT HAS SUFFICIENT MEANINGFUL CONTENT:
             1. A SPECIFIC and DESCRIPTIVE topic name (3-5 words) that precisely captures the main focus
             2. 3-5 key concepts discussed
             3. A brief summary (6-7 sentences)
-            4. Create 5 quiz questions based DIRECTLY on the text content (not from your summary)
+            4. Create 5 quiz questions based DIRECTLY on the text content (not from your summary), ONLY if the segment contains factual information
             
             For each quiz question:
             - Create one correct answer that comes DIRECTLY from the text
@@ -231,6 +251,7 @@ def analyze_segment_with_gemini(cluster_text, is_full_text=False):
             
             Format your response as JSON with the following structure:
             {{
+                "status": "processed",
                 "topic_name": "Name of the topic",
                 "key_concepts": ["concept1", "concept2", "concept3"],
                 "summary": "Brief summary of the text segment.",
@@ -255,50 +276,19 @@ def analyze_segment_with_gemini(cluster_text, is_full_text=False):
                     // More questions...
                 ]
             }}
+            
+            OR if the text is just introductory, concluding, or insufficient:
+            {{
+                "status": "insufficient",
+                "reason": "Brief explanation of why (e.g., 'Text is primarily self-introduction', 'Text is too short', etc.)"
+            }}
         """
-    
-    response = llm.invoke(prompt)
-    
-    response_text = response.content
-    
-    try:
-        json_match = re.search(r'\{[\s\S]*\}', response_text)
-        if json_match:
-            response_json = json.loads(json_match.group(0))
-        else:
-            response_json = json.loads(response_text)
-        
-        return response_json
-    except json.JSONDecodeError as e:
-        print(f"Error parsing JSON response: {e}")
-        print(f"Raw response: {response_text}")
-        
-        if is_full_text:
-            return {
-                "segments": [
-                    {
-                        "topic_name": "JSON Parsing Error",
-                        "key_concepts": ["Error in response format"],
-                        "summary": f"Could not parse the API response. Raw text: {response_text[:200]}...",
-                        "quiz_questions": []
-                    }
-                ]
-            }
-        else:
-            return {
-                "topic_name": "JSON Parsing Error",
-                "key_concepts": ["Error in response format"],
-                "summary": f"Could not parse the API response. Raw text: {response_text[:200]}...",
-                "quiz_questions": []
-            }
-
-
 
 def process_document_with_quiz(text):
     token_count = len(tokenizer.encode(text))
     print(f"Text contains {token_count} tokens")
     
-    if token_count < 7000:
+    if token_count < 8000:
         print("Text is short enough to analyze directly without text segmentation")
         full_analysis = analyze_segment_with_gemini(text, is_full_text=True)