Update app.py

app.py

@@ -6,6 +6,7 @@ import os
 import PyPDF2
 import docx
 import pandas as pd
+import random
 
 class SmartRAGBot:
     def __init__(self):
@@ -15,35 +16,65 @@ class SmartRAGBot:
         self.index = None
         self.is_ready = False
         
-        # General knowledge responses for out-of-context questions
+        # Enhanced general knowledge base
+        self.general_knowledge = {
+            "capital of india": "The capital of India is New Delhi.",
+            "capital of india?": "New Delhi is the capital of India.",
+            "what is capital of india": "The capital of India is New Delhi.",
+            "capital of usa": "The capital of USA is Washington D.C.",
+            "capital of uk": "The capital of United Kingdom is London.",
+            "capital of japan": "The capital of Japan is Tokyo.",
+            "current prime minister of india": "The current Prime Minister of India is Narendra Modi.",
+            "current president of india": "The current President of India is Droupadi Murmu.",
+            "population of india": "India's population is approximately 1.4 billion people.",
+            "currency of india": "The currency of India is Indian Rupee (₹).",
+            "largest city in india": "Mumbai is the largest city in India by population.",
+            "who is prime minister of india": "Narendra Modi is the Prime Minister of India.",
+            "who is president of india": "Droupadi Murmu is the President of India.",
+            "national animal of india": "The national animal of India is the Bengal Tiger.",
+            "national bird of india": "The national bird of India is the Peacock.",
+            "independence day of india": "India's Independence Day is on August 15th.",
+            "republic day of india": "India's Republic Day is on January 26th."
+        }
+        
+        # Enhanced general responses
         self.general_responses = {
             "greeting": [
-                "Hello! How can I help you today?",
+                "Hello! How can I assist you today?",
                 "Hi there! What would you like to know?",
-                "Greetings! I'm here to assist you."
+                "Greetings! I'm here to help you with both document questions and general knowledge.",
+                "Welcome! How can I help you today?"
             ],
             "thanks": [
                 "You're welcome!",
                 "Happy to help!",
-                "Glad I could assist you!"
+                "Glad I could assist you!",
+                "My pleasure! Let me know if you need anything else."
             ],
             "identity": [
-                "I'm an AI assistant that can answer questions about your uploaded documents.",
-                "I'm a document Q&A bot that helps you find information in your files.",
-                "I'm here to help you get information from your uploaded documents."
+                "I'm an AI assistant that can answer questions about your uploaded documents and also provide general knowledge information.",
+                "I'm a smart document Q&A bot that can answer questions about your files and general knowledge topics.",
+                "I'm here to help you find information in your files and answer general knowledge questions."
             ],
             "capabilities": [
-                "I can read text files, PDFs, Word documents, and CSV files, then answer questions about their content.",
-                "I help you find specific information in your uploaded documents.",
-                "Upload any document and I'll help you search through its content."
+                "I can: Read your files (TXT, PDF, Word, CSV) and answer questions about them, Answer general knowledge questions, Have conversations in English",
+                "I help with: Document analysis, General knowledge, Casual conversation",
+                "My capabilities include: Document Q&A, General knowledge responses, File content analysis"
+            ],
+            "how_are_you": [
+                "I'm doing great! Ready to help you with your questions.",
+                "I'm functioning perfectly! What can I help you with today?",
+                "I'm running smoothly and ready to assist you!",
+                "All systems go! How can I assist you?"
             ],
             "default": [
-                "I'm not sure about that, but I can help you with questions about your uploaded document.",
-                "That's outside my current knowledge, but I'm great at finding information in your files!",
-                "I specialize in document Q&A. Try asking me something about your uploaded file."
+                "I can help you with that! But I'm even better at answering questions about your uploaded documents.",
+                "That's an interesting question! I specialize in document Q&A, but I'll do my best to help.",
+                "I can answer that! Feel free to upload a document if you have specific questions about its content.",
+                "I'd be happy to help with that question!"
             ]
         }
-    
+
     def load_file_content(self, file):
         """Load content from various file types"""
         try:
@@ -165,68 +196,110 @@ class SmartRAGBot:
             return []
     
     def detect_query_type(self, query):
-        """Detect if query is about file content or general"""
-        query_lower = query.lower()
+        """Enhanced query type detection"""
+        query_lower = query.lower().strip()
+        
+        # Check general knowledge first
+        if query_lower in self.general_knowledge:
+            return "general_knowledge"
+        
+        # Check for general knowledge patterns
+        general_patterns = [
+            'capital of', 'population of', 'currency of', 'prime minister of', 
+            'president of', 'largest city in', 'what is', 'who is', 'when is',
+            'where is', 'how many', 'full form of', 'national animal', 
+            'national bird', 'independence day', 'republic day'
+        ]
+        if any(pattern in query_lower for pattern in general_patterns):
+            return "general_knowledge"
         
         # Greeting detection
-        greetings = ['hello', 'hi', 'hey', 'namaste', 'good morning', 'good afternoon', 'good evening']
+        greetings = ['hello', 'hi', 'hey', 'good morning', 'good afternoon', 'good evening']
         if any(greet in query_lower for greet in greetings):
             return "greeting"
         
         # Thanks detection
-        thanks_words = ['thank', 'thanks', 'dhanyavad', 'shukriya']
+        thanks_words = ['thank', 'thanks', 'thankyou', 'thank you']
         if any(word in query_lower for word in thanks_words):
             return "thanks"
         
+        # How are you detection
+        how_are_you = ['how are you', 'how do you do', 'how is it going']
+        if any(phrase in query_lower for phrase in how_are_you):
+            return "how_are_you"
+        
         # Identity questions
-        identity_words = ['who are you', 'what are you', 'your name', 'tu kaun hai', 'aap kaun ho']
+        identity_words = ['who are you', 'what are you', 'your name', 'introduce yourself']
         if any(word in query_lower for word in identity_words):
             return "identity"
         
         # Capability questions
-        capability_words = ['what can you do', 'help', 'capabilities', 'kaun kaun se kaam kar sakte ho']
+        capability_words = ['what can you do', 'help', 'capabilities', 'features']
         if any(word in query_lower for word in capability_words):
             return "capabilities"
         
         return "document"
     
-    def get_general_response(self, query_type):
+    def get_general_response(self, query_type, original_query=""):
         """Get appropriate general response"""
-        import random
+        if query_type == "general_knowledge":
+            # Try exact match first
+            if original_query.lower() in self.general_knowledge:
+                return self.general_knowledge[original_query.lower()]
+            
+            # Try to find the best match
+            for key, value in self.general_knowledge.items():
+                if key in original_query.lower():
+                    return value
+            
+            # If no specific knowledge found, use default
+            responses = self.general_responses.get("default", ["I can help with that!"])
+            return random.choice(responses)
+        
         responses = self.general_responses.get(query_type, self.general_responses["default"])
         return random.choice(responses)
     
     def generate_response(self, query):
         """Generate appropriate response based on query type"""
-        # First check if it's a general question
+        # First check query type
         query_type = self.detect_query_type(query)
         
-        if query_type != "document":
-            return self.get_general_response(query_type)
+        # Handle general questions (always answer these)
+        if query_type in ["greeting", "thanks", "how_are_you", "identity", "capabilities", "general_knowledge"]:
+            return self.get_general_response(query_type, query)
         
         # If it's a document question but no file is loaded
         if not self.is_ready:
-            return "Please upload a document file first to ask questions about its content."
+            return ("📚 Please upload a document file first to ask questions about its content.\n\n"
+                   "🌍 But I can answer general questions like:\n"
+                   "• 'What is the capital of India?'\n"
+                   "• 'How are you?'\n"
+                   "• 'What can you do?'")
         
         # Search in document content
         results = self.search_documents(query)
         
         if not results:
-            # No relevant content found in document
-            return (f"I couldn't find specific information about '{query}' in your document. "
-                   "This might be outside the document's scope. "
-                   "You can ask me general questions like 'What can you do?' or upload a different file.")
+            # No relevant content found in document - provide helpful response
+            return (f"📖 I couldn't find specific information about '{query}' in your document.\n\n"
+                   "💡 This might be outside the document's scope. You can:\n"
+                   "• Ask me general knowledge questions\n"
+                   "• Upload a different file\n"
+                   "• Try rephrasing your question\n\n"
+                   "🌍 Try asking: 'What is the capital of India?' or 'How are you?'")
         
         # Check if the results are actually relevant (score threshold)
         relevant_results = [r for r in results if r['score'] > 0.3]
         
         if not relevant_results:
-            return (f"The document doesn't seem to contain specific information about '{query}'. "
-                   "This might be outside the document's scope. "
-                   "Try rephrasing your question or ask me something else!")
+            return (f"📖 The document doesn't contain specific information about '{query}'.\n\n"
+                   "💡 But I can help with:\n"
+                   "• General knowledge questions\n"
+                   "• Other questions about your document\n"
+                   "• Casual conversation")
         
         # Build response from relevant document content
-        response = "Based on your document:\n\n"
+        response = "📚 Based on your document:\n\n"
         for i, result in enumerate(relevant_results[:2]):  # Show top 2 results
             response += f"• {result['content']}\n\n"
         
@@ -238,52 +311,65 @@ def create_interface():
     with gr.Blocks(theme=gr.themes.Soft(), title="Smart Document Assistant") as demo:
         gr.Markdown("""
         # 🤖 Smart Document Assistant
+        
         **Upload your file and ask any questions!**  
-        - File ke content se related questions ka answer milega  
-        - General questions bhi pooch sakte hain  
-        - Supported formats: TXT, CSV, PDF, DOCX
+        
+        - **📚 Document Questions**: Get answers from your uploaded files
+        - **🌍 General Knowledge**: Ask about capitals, populations, facts
+        - **💬 Casual Chat**: Have normal conversations
+        
+        **Supported Formats**: TXT, CSV, PDF, DOCX, DOC
         """)
         
         with gr.Row():
             with gr.Column(scale=1):
                 gr.Markdown("### 📁 File Upload")
                 file_input = gr.File(
-                    label="Apna file upload karein",
+                    label="Upload your file",
                     file_types=[".txt", ".csv", ".pdf", ".docx", ".doc"],
                     type="filepath"
                 )
                 upload_btn = gr.Button("Process File", variant="primary")
                 status = gr.Textbox(
                     label="Status",
-                    value="Kripya pehle file upload karein...",
+                    value="Please upload a file first...",
                     interactive=False
                 )
                 
             with gr.Column(scale=2):
-                gr.Markdown("### 💬 Questions Poochein")
+                gr.Markdown("### 💬 Ask Questions")
                 chatbot = gr.Chatbot(
                     height=400,
-                    label="Baatcheet",
-                    show_copy_button=True
+                    label="Conversation",
+                    show_copy_button=True,
+                    placeholder="Your questions and answers will appear here..."
                 )
                 with gr.Row():
                     question_input = gr.Textbox(
-                        label="Apna question",
-                        placeholder="Yahan apna question type karein aur Enter dabayein...",
+                        label="Your question",
+                        placeholder="Type any question - about your file or general knowledge...",
                         scale=4
                     )
                     send_btn = gr.Button("Send", variant="primary", scale=1)
                 
-                clear_btn = gr.Button("Chat Clear Karein", variant="secondary")
+                clear_btn = gr.Button("Clear Chat", variant="secondary")
                 
                 # Example questions
                 gr.Markdown("""
-                ### 💡 Example Questions:
-                **File-related:** 
+                ### 💡 Try These Questions:
+                
+                **📚 File-related Questions:** 
                 - "What is the main topic of this document?"
                 - "Tell me about the key points"
+                - "Summarize the content"
+                
+                **🌍 General Knowledge:**
+                - "What is the capital of India?"
+                - "Who is the Prime Minister of India?"
+                - "What is the population of India?"
+                - "What is the national animal of India?"
                 
-                **General:**
+                **💬 Casual Questions:**
                 - "Hello, how are you?"
                 - "What can you do?"
                 - "Thank you for your help!"