app.py

import gradio as gr
import os
import io
# Commenting out local model import - we'll use OpenRouter API instead
# from model import pipe  # Import your model pipeline
import PyPDF2
import docx
import pandas as pd
from typing import List, Tuple, Optional
import requests

# New imports for advanced text extraction
import pytesseract
import cv2
import numpy as np
import pdfplumber
from pdf2image import convert_from_path
from PIL import Image

# OpenRouter API configuration
OPENROUTER_API_KEY = "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"


def get_openrouter_completion(messages, max_tokens=600, temperature=0.7):
    """Get completion from OpenRouter API using Mistral model."""
    url = "https://openrouter.ai/api/v1/chat/completions"
    headers = {
        "Authorization": f"Bearer {OPENROUTER_API_KEY}",
        "Content-Type": "application/json",
        "HTTP-Referer": "http://localhost:7860",  # Required for OpenRouter
        "X-Title": "AI Chatbot"  # Optional: for analytics
    }
    
    # Ensure messages are properly formatted and not too long
    formatted_messages = []
    for msg in messages:
        if isinstance(msg, dict) and "role" in msg and "content" in msg:
            content = str(msg["content"]).strip()
            # Limit very long content to prevent API errors
            if len(content) > 10000:
                content = content[:10000] + "... [content truncated]"
            formatted_messages.append({
                "role": msg["role"],
                "content": content
            })
    
    json_data = {
        "model": "mistralai/mistral-7b-instruct-v0.1",  # Using reliable model
        "messages": formatted_messages,
        "max_tokens": min(max_tokens, 800),  # Reasonable limit
        "temperature": max(0.1, min(temperature, 1.0)),  # Valid range
        "top_p": 0.9,
        "stream": False
    }
    
    try:
        print(f"🔄 Making API request to OpenRouter...")
        response = requests.post(url, headers=headers, json=json_data, timeout=30)
        
        print(f"📡 Response status: {response.status_code}")
        
        if response.status_code == 400:
            try:
                error_details = response.json()
                print(f"❌ 400 Error details: {error_details}")
                error_msg = error_details.get('error', {}).get('message', 'Bad Request')
                return f"API Error: {error_msg}. Please check the API key and request format."
            except:
                return "API Error: 400 Bad Request. Please check your API configuration."
        
        elif response.status_code == 401:
            return "API Error: Invalid API key. Please check your OpenRouter API key."
        
        elif response.status_code == 429:
            return "API Error: Rate limit exceeded. Please try again in a moment."
        
        elif response.status_code != 200:
            return f"API Error {response.status_code}: {response.text[:200]}..."
        
        response_data = response.json()
        
        if "choices" in response_data and len(response_data["choices"]) > 0:
            return response_data["choices"][0]["message"]["content"]
        else:
            return "No response generated from API"
            
    except requests.exceptions.Timeout:
        return "⏱️ Request timeout - please try again"
    except requests.exceptions.ConnectionError:
        return "🌐 Connection error - check your internet connection"
    except Exception as e:
        print(f"❌ API Exception: {str(e)}")
        return f"Unexpected error: {str(e)}"


def get_fallback_response(message, file_content=""):
    """Provide a helpful fallback response when API is unavailable."""
    if file_content:
        file_summary = f"I can see you've uploaded files with content. Here's a basic analysis:\n\n"
        file_summary += f"Content length: {len(file_content)} characters\n"
        
        # Basic content analysis
        if "price" in file_content.lower() or "$" in file_content:
            file_summary += "• I notice pricing information in the uploaded content\n"
        if "plan" in file_content.lower():
            file_summary += "• I see plan-related information\n"
        if any(word in file_content.lower() for word in ["phone", "mobile", "data", "gb", "mb"]):
            file_summary += "• This appears to contain telecommunications/mobile plan information\n"
        
        file_summary += f"\nYou asked: '{message}'\n\n"
        file_summary += "I'm currently in fallback mode due to API issues, but I can see your file content has been processed successfully. For full AI analysis, please check the API configuration."
        
        return file_summary
    else:
        return f"I understand you said: '{message}'. I'm currently in fallback mode due to API connectivity issues. I can still process your files - try uploading a document and I'll extract its content for you."


def extract_text_from_image(image_path: str) -> str:
    """Extract text from image using OCR (Tesseract)."""
    try:
        # Open image
        if isinstance(image_path, str):
            image = Image.open(image_path)
        else:
            image = image_path
        
        # Convert PIL image to OpenCV format
        img_cv = cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR)
        gray = cv2.cvtColor(img_cv, cv2.COLOR_BGR2GRAY)
        
        # Apply threshold for better OCR
        _, thresh = cv2.threshold(gray, 150, 255, cv2.THRESH_BINARY)
        
        # Extract text using Tesseract
        text = pytesseract.image_to_string(thresh)
        
        # Clean up the text
        if text.strip():
            cleaned_text = ' '.join(text.split())
            return cleaned_text if cleaned_text else "No meaningful text found in image"
        else:
            return "No text found in image"
    
    except Exception as e:
        return f"Error extracting text from image: {str(e)}"


def extract_text_from_pdf_advanced(file_path: str) -> str:
    """Extract text from PDF with fallback to OCR for image-based PDFs."""
    try:
        # First try: Extract text directly using pdfplumber (faster)
        with pdfplumber.open(file_path) as pdf:
            text_content = []
            
            for page_num, page in enumerate(pdf.pages):
                page_text = page.extract_text()
                if page_text and page_text.strip():
                    text_content.append(f"--- Page {page_num + 1} ---\n{page_text.strip()}")
            
            if text_content:
                return "\n\n".join(text_content)
        
        # Fallback: If no text found, use OCR
        print("🔄 No text found in PDF, trying OCR...")
        images = convert_from_path(file_path)
        ocr_text = []
        
        for i, image in enumerate(images):
            page_text = extract_text_from_image(image)
            if page_text and not page_text.startswith("Error"):
                ocr_text.append(f"--- Page {i + 1} (OCR) ---\n{page_text}")
        
        return "\n\n".join(ocr_text) if ocr_text else "No text could be extracted from this PDF"
    
    except Exception as e:
        return f"Error processing PDF: {str(e)}"


def extract_text_from_file(file_path: str) -> str:
    """Extract text from various file formats with advanced OCR capabilities."""
    if not file_path:
        return "No file path provided"
    
    # Handle both file paths and file objects from Gradio
    if hasattr(file_path, 'name'):
        actual_path = file_path.name
    else:
        actual_path = str(file_path)
    
    if not os.path.exists(actual_path):
        return f"File not found: {actual_path}"
    
    file_extension = os.path.splitext(actual_path)[1].lower()
    
    try:
        # Handle image files with OCR
        if file_extension in ['.png', '.jpg', '.jpeg', '.bmp', '.tiff', '.gif']:
            return extract_text_from_image(actual_path)
        
        # Handle PDFs with advanced extraction
        elif file_extension == '.pdf':
            return extract_text_from_pdf_advanced(actual_path)
        
        # Handle Word documents
        elif file_extension == '.docx':
            try:
                doc = docx.Document(actual_path)
                text = ""
                for paragraph in doc.paragraphs:
                    if paragraph.text.strip():
                        text += paragraph.text + "\n"
                return text if text.strip() else "No text found in this Word document."
            except Exception as e:
                return f"Error reading Word document: {str(e)}"
        
        # Handle Excel files
        elif file_extension in ['.xlsx', '.xls']:
            try:
                # Try to read all sheets
                excel_file = pd.ExcelFile(actual_path)
                all_text = ""
                for sheet_name in excel_file.sheet_names:
                    df = pd.read_excel(actual_path, sheet_name=sheet_name)
                    all_text += f"--- Sheet: {sheet_name} ---\n"
                    all_text += df.to_string(index=False) + "\n\n"
                return all_text if all_text.strip() else "No data found in this Excel file."
            except Exception as e:
                return f"Error reading Excel file: {str(e)}"
        
        # Handle CSV files
        elif file_extension == '.csv':
            try:
                df = pd.read_csv(actual_path)
                return df.to_string(index=False)
            except Exception as e:
                return f"Error reading CSV file: {str(e)}"
        
        # Handle text files
        elif file_extension == '.txt':
            try:
                encodings = ['utf-8', 'utf-16', 'latin-1', 'cp1252']
                for encoding in encodings:
                    try:
                        with open(actual_path, 'r', encoding=encoding) as file:
                            return file.read()
                    except UnicodeDecodeError:
                        continue
                return "Could not decode text file with any supported encoding."
            except Exception as e:
                return f"Error reading text file: {str(e)}"
        
        else:
            # Try to read as text file with multiple encodings
            try:
                encodings = ['utf-8', 'utf-16', 'latin-1', 'cp1252']
                for encoding in encodings:
                    try:
                        with open(actual_path, 'r', encoding=encoding) as file:
                            content = file.read()
                            return f"File read as text (encoding: {encoding}):\n{content}"
                    except UnicodeDecodeError:
                        continue
                return f"Unsupported file format: {file_extension}. Try converting to PDF, image, or text format."
            except Exception as e:
                return f"Error reading file: {str(e)}"
    
    except Exception as e:
        return f"Error processing file: {str(e)}"


def respond(
    message: str,
    history: List[Tuple[str, str]],
    uploaded_files: Optional[List] = None,
    system_message: str = "You are a helpful AI assistant.",
    max_tokens: int = 512,
    temperature: float = 0.7,
):
    """Generate response using the local model with file context."""
    
    # Process uploaded files
    file_content = ""
    if uploaded_files:
        for i, file in enumerate(uploaded_files):
            try:
                # Handle different ways Gradio might pass files
                if hasattr(file, 'name'):
                    file_path = file.name
                    file_name = os.path.basename(file_path)
                elif isinstance(file, str):
                    file_path = file
                    file_name = os.path.basename(file_path)
                else:
                    file_path = str(file)
                    file_name = f"file_{i+1}"
                
                content = extract_text_from_file(file_path)
                
                if content and not content.startswith("Error"):
                    file_content += f"\n\n--- Content from {file_name} ---\n{content}\n"
                else:
                    file_content += f"\n\n--- Error processing {file_name} ---\n{content}\n"
                    
            except Exception as e:
                error_msg = f"Error processing file {i+1}: {str(e)}"
                file_content += f"\n\n--- {error_msg} ---\n"
    
    # Build the conversation messages for OpenRouter API
    messages = [{"role": "system", "content": system_message}]
    
    # Add conversation history
    for user_msg, assistant_msg in history:
        if user_msg:
            messages.append({"role": "user", "content": user_msg})
        if assistant_msg:
            messages.append({"role": "assistant", "content": assistant_msg})
    
    # Add file content to the current message if available
    current_message = message
    if file_content:
        current_message = f"{message}\n\nAdditional context from uploaded files:{file_content}"
    
    messages.append({"role": "user", "content": current_message})
    
    try:
        # Generate response using OpenRouter API with Mistral model
        response = get_openrouter_completion(
            messages=messages,
            max_tokens=max_tokens,
            temperature=temperature
        )
        
        # Check if response indicates an API error
        if response.startswith("API Error") or response.startswith("❌") or response.startswith("⏱️") or response.startswith("🌐"):
            print("🔄 API failed, using fallback response...")
            return get_fallback_response(message, file_content)
        
        return response if response else "Sorry, I couldn't generate a response."
            
    except Exception as e:
        print(f"❌ Exception in respond function: {str(e)}")
        return get_fallback_response(message, file_content)


"""
ChatGPT-like interface with file upload support using Mistral AI via OpenRouter API
"""

# Create custom interface with file upload
with gr.Blocks(title="AI Chatbot with File Upload & Mistral AI", theme=gr.themes.Soft()) as demo:
    gr.Markdown("# 🤖 AI Chatbot with Advanced File Upload & OCR (Powered by Mistral AI)")
    gr.Markdown("Upload files (PDF, DOCX, TXT, CSV, XLSX, Images) and chat with AI about their content! Uses Mistral AI for intelligent responses and includes OCR for images and scanned PDFs.")
    
    with gr.Row():
        with gr.Column(scale=3):
            chatbot = gr.Chatbot(
                height=500,
                show_label=False,
                avatar_images=["👤", "🤖"]
            )
            
            with gr.Row():
                msg = gr.Textbox(
                    placeholder="Type your message here...",
                    show_label=False,
                    scale=4
                )
                send_btn = gr.Button("Send", variant="primary")
            
            file_upload = gr.Files(
                label="Upload Files (PDF, DOCX, TXT, CSV, XLSX, Images: PNG, JPG, etc.)",
                file_types=None,  # Allow all file types for now
                file_count="multiple"
            )
            
        with gr.Column(scale=1):
            gr.Markdown("### Settings")
            system_message = gr.Textbox(
                value="You are a helpful AI assistant powered by Mistral AI. You can analyze uploaded files and answer questions about their content. Provide detailed, accurate, and helpful responses.",
                label="System Message",
                lines=3
            )
            max_tokens = gr.Slider(
                minimum=50,
                maximum=2048,
                value=512,
                step=50,
                label="Max Tokens"
            )
            temperature = gr.Slider(
                minimum=0.1,
                maximum=2.0,
                value=0.7,
                step=0.1,
                label="Temperature"
            )
            clear_btn = gr.Button("Clear Chat", variant="secondary")
    
    # Chat functionality
    def user_message(message, history, files):
        if message.strip() == "":
            return "", history, files
        return "", history + [[message, None]], files
    
    def bot_response(history, files, system_msg, max_tok, temp):
        if not history or history[-1][1] is not None:
            return history
        
        user_message = history[-1][0]
        bot_reply = respond(user_message, history[:-1], files, system_msg, max_tok, temp)
        history[-1][1] = bot_reply
        return history
    
    def clear_chat():
        return [], None
    
    # Event handlers
    msg.submit(
        user_message,
        [msg, chatbot, file_upload],
        [msg, chatbot, file_upload]
    ).then(
        bot_response,
        [chatbot, file_upload, system_message, max_tokens, temperature],
        chatbot
    )
    
    send_btn.click(
        user_message,
        [msg, chatbot, file_upload],
        [msg, chatbot, file_upload]
    ).then(
        bot_response,
        [chatbot, file_upload, system_message, max_tokens, temperature],
        chatbot
    )
    
    clear_btn.click(clear_chat, outputs=[chatbot, file_upload])


if __name__ == "__main__":
    demo.launch()