Update app.py

app.py

@@ -1,39 +1,225 @@
-"""
-Jajabor - Absolute Minimal Working Version
-"""
+import streamlit as st
+import os
+import sqlite3
+from datetime import datetime
+import PyPDF2
+from sentence_transformers import SentenceTransformer
+import faiss
+import numpy as np
+from transformers import pipeline
+import pytesseract
+from PIL import Image
+import sympy as sp
+import io
 
-import gradio as gr
+# App configuration
+st.set_page_config(
+    page_title="Jajabor – SEBA Class 10 Tutor",
+    page_icon="🧭",
+    layout="wide"
+)
 
-def chat_fn(message, history):
-    if history is None:
-        history = []
-    
-    if not message.strip():
-        history.append(("", "অনুগ্ৰহ কৰি প্ৰশ্ন লিখক।"))
-        return history
-    
-    # Simple echo with Assamese response
-    response = f"ধন্যবাদ! আপোনাৰ প্ৰশ্ন: '{message}'\n\nমই জাজাবৰ, আপোনাৰ SEBA Class 10 AI টিউটাৰ। PDF আপলোড কৰি RAG চিস্টেম সক্ৰিয় কৰক।"
-    history.append((message, response))
-    return history
+# Initialize session state
+if 'chat_history' not in st.session_state:
+    st.session_state.chat_history = []
+if 'username' not in st.session_state:
+    st.session_state.username = ""
+if 'tutor' not in st.session_state:
+    st.session_state.tutor = None
 
-# Minimal interface
-with gr.Blocks() as demo:
-    gr.Markdown("# 🧭 জাজাবৰ - SEBA Class 10 Tutor")
-    chatbot = gr.Chatbot()
-    msg = gr.Textbox(label="প্ৰশ্ন লিখক")
-    clear = gr.Button("পৰিষ্কাৰ কৰক")
-    
-    def user(user_message, history):
-        return "", history + [[user_message, None]]
-    
-    def bot(history):
-        response = "জাজাবৰই আপোনাক সহায় কৰিবলৈ সাজু! PDF পাঠ্যপুথি আপলোড কৰি সম্পূৰ্ণ RAG চিস্টেম সক্ৰিয় কৰক।"
-        history[-1][1] = response
-        return history
-    
-    msg.submit(user, [msg, chatbot], [msg, chatbot], queue=False).then(bot, chatbot, chatbot)
-    clear.click(lambda: None, None, chatbot, queue=False)
+class SimpleTutor:
+    def __init__(self):
+        self.llm = None
+        self.embedding_model = None
+        self.index = None
+        self.corpus_chunks = []
+        self._load_models()
+        self.load_pdfs()
+    
+    def _load_models(self):
+        try:
+            self.embedding_model = SentenceTransformer("all-MiniLM-L6-v2")
+            self.llm = pipeline("text2text-generation", model="google/flan-t5-small", device=-1)
+        except Exception as e:
+            st.error(f"Model loading error: {e}")
+    
+    def load_pdfs(self):
+        pdf_dir = "pdfs/class10"
+        if not os.path.exists(pdf_dir):
+            return
+        
+        all_texts = []
+        for fname in os.listdir(pdf_dir):
+            if fname.lower().endswith('.pdf'):
+                path = os.path.join(pdf_dir, fname)
+                try:
+                    reader = PyPDF2.PdfReader(path)
+                    text = ""
+                    for page in reader.pages:
+                        text += page.extract_text() or ""
+                    if text.strip():
+                        all_texts.append(text)
+                except Exception as e:
+                    st.error(f"Error reading {fname}: {e}")
+        
+        self.corpus_chunks = []
+        for text in all_texts:
+            chunks = self._split_text(text)
+            self.corpus_chunks.extend(chunks)
+        
+        if self.corpus_chunks and self.embedding_model:
+            try:
+                embs = self.embedding_model.encode(self.corpus_chunks).astype("float32")
+                dim = embs.shape[1]
+                self.index = faiss.IndexFlatL2(dim)
+                self.index.add(embs)
+            except Exception as e:
+                st.error(f"FAISS error: {e}")
+    
+    def _split_text(self, text, chunk_size=400):
+        if not text:
+            return []
+        return [text[i:i+chunk_size] for i in range(0, len(text), chunk_size) if text[i:i+chunk_size].strip()]
+    
+    def answer_question(self, question):
+        if not question.strip():
+            return "অনুগ্ৰহ কৰি এটা প্ৰশ্ন সোধক।"
+        
+        if self._is_math_question(question):
+            return self._solve_math(question)
+        
+        context = ""
+        if self.index and self.corpus_chunks:
+            relevant_chunks = self._find_relevant_chunks(question)
+            if relevant_chunks:
+                context = "\n".join(relevant_chunks[:2])
+        
+        if self.llm:
+            try:
+                prompt = f"প্ৰশ্ন: {question}\n\nসংদৰ্ভ: {context}\n\nসহায়ক উত্তৰ:" if context else f"প্ৰশ্ন: {question}\n\nউত্তৰ:"
+                response = self.llm(prompt, max_new_tokens=150, temperature=0.3)
+                return response[0]['generated_text']
+            except Exception as e:
+                return f"উত্তৰ তৈয়াৰ কৰোঁতে সমস্যা: {e}"
+        else:
+            return "মই আপোনাৰ প্ৰশ্নটো বুজিলোঁ। অধ্যয়নৰ বাবে শুভেচ্ছা!"
+    
+    def _is_math_question(self, text):
+        math_indicators = ['+', '-', '*', '/', '=', 'x', 'y', 'গণিত', 'সমীকৰণ']
+        return any(indicator in text.lower() for indicator in math_indicators)
+    
+    def _solve_math(self, expr):
+        try:
+            expr = expr.strip().replace('^', '**')
+            if '=' in expr:
+                parts = expr.split('=')
+                if len(parts) == 2:
+                    left = sp.sympify(parts[0].strip())
+                    right = sp.sympify(parts[1].strip())
+                    equation = sp.Eq(left, right)
+                    solutions = sp.solve(equation)
+                    if solutions:
+                        return f"সমীকৰণ: {equation}\n\nসমাধান: {solutions}"
+            else:
+                expr_sym = sp.sympify(expr)
+                simplified = sp.simplify(expr_sym)
+                return f"প্ৰকাশ: {expr}\n\nসৰলীকৃত: {simplified}"
+        except Exception as e:
+            return f"গণিত সমাধানত সমস্যা: {e}"
+    
+    def _find_relevant_chunks(self, question, k=3):
+        if not self.corpus_chunks:
+            return []
+        
+        if self.index and self.embedding_model:
+            try:
+                q_vec = self.embedding_model.encode([question]).astype("float32")
+                D, I = self.index.search(q_vec, k)
+                return [self.corpus_chunks[i] for i in I[0] if 0 <= i < len(self.corpus_chunks)]
+            except Exception:
+                pass
+        return []
+
+def extract_text_from_image(uploaded_file):
+    try:
+        image = Image.open(uploaded_file)
+        text = pytesseract.image_to_string(image)
+        return text.strip()
+    except Exception as e:
+        return ""
+
+# Main app
+def main():
+    st.title("🧭 জাজাবৰ – SEBA Class 10 AI Tutor")
+    
+    # Sidebar
+    with st.sidebar:
+        st.header("👤 লগিন")
+        username = st.text_input("আপোনাৰ নাম", value=st.session_state.username)
+        if username and username != st.session_state.username:
+            st.session_state.username = username
+            st.success(f"লগিন successful: {username}")
+        
+        st.header("📷 ছবিৰ পৰা পাঠ")
+        uploaded_image = st.file_uploader("ছবি আপলোড কৰক", type=['png', 'jpg', 'jpeg'])
+        
+        st.header("💡 টিপছ")
+        st.info("""
+        - নাম লিখি প্ৰশ্ন সোধক
+        - ছবি আপলোড কৰিলে OCR ৰ সহায়ত পাঠ পঢ়িব
+        - বিষয়সমূহ:
+          - অসমীয়া
+          - ইংৰাজী  
+          - গণিত
+          - বিজ্ঞান
+          - সামাজিক বিজ্ঞান
+        """)
+    
+    # Initialize tutor
+    if st.session_state.tutor is None:
+        with st.spinner('জাজাবৰক সাজু কৰি থকা হৈছে...'):
+            st.session_state.tutor = SimpleTutor()
+    
+    # Main chat area
+    st.header("💬 জাজাবৰৰ সৈতে কথোপকথন")
+    
+    # Display chat history
+    for i, (question, answer) in enumerate(st.session_state.chat_history):
+        with st.chat_message("user"):
+            st.write(question)
+        with st.chat_message("assistant"):
+            st.write(answer)
+    
+    # Chat input
+    if prompt := st.chat_input("আপোনাৰ প্ৰশ্ন ইয়াত লিখক..."):
+        if not st.session_state.username:
+            st.error("⚠️ প্ৰথমে আপোনাৰ নাম লিখক")
+            st.stop()
+        
+        # Add user message to chat
+        st.session_state.chat_history.append((prompt, ""))
+        
+        # Process OCR if image uploaded
+        full_question = prompt
+        if uploaded_image:
+            ocr_text = extract_text_from_image(uploaded_image)
+            if ocr_text:
+                full_question += f"\n[ছবিৰ পাঠ: {ocr_text}]"
+        
+        # Get AI response
+        with st.spinner('জাজাবৰে চিন্তা কৰি আছে...'):
+            response = st.session_state.tutor.answer_question(full_question)
+        
+        # Update chat history
+        st.session_state.chat_history[-1] = (prompt, response)
+        
+        # Rerun to update display
+        st.rerun()
+    
+    # Clear chat button
+    if st.button("🧹 কথোপকথন পৰিষ্কাৰ কৰক"):
+        st.session_state.chat_history = []
+        st.rerun()
 
 if __name__ == "__main__":
-    demo.launch(share=False)
+    main()