app.py

import os
import shutil
import threading
import tempfile
from typing import Optional, List, Dict, Any

from dotenv import load_dotenv
from huggingface_hub import login
import gradio as gr

from scripts.text_inspector_tool import TextInspectorTool
from scripts.text_web_browser import (
    ArchiveSearchTool,
    FinderTool,
    FindNextTool,
    PageDownTool,
    PageUpTool,
    SimpleTextBrowser,
    VisitTool,
)
from scripts.visual_qa import visualizer
from scripts.legal_document_tool import LegalDocumentTool
from smolagents import (
    CodeAgent,
    HfApiModel,
    LiteLLMModel,
    OpenAIServerModel,
    TransformersModel,
    GoogleSearchTool,
    Tool,
)
from smolagents.agent_types import AgentText, AgentImage, AgentAudio
from smolagents.gradio_ui import pull_messages_from_step, handle_agent_output_types

# ------------------------ Configuration and Setup ------------------------
# Constants and configurations
AUTHORIZED_IMPORTS = [
    "requests",
    "zipfile",
    "pandas",
    "numpy",
    "sympy",
    "json",
    "bs4",
    "pubchempy",
    "xml",
    "yahoo_finance",
    "Bio",
    "sklearn",
    "scipy",
    "pydub",
    "PIL",
    "chess",
    "PyPDF2",
    "pptx",
    "torch",
    "datetime",
    "fractions",
    "csv",
    "clean-text",
    "langchain",
    "llama_index",
]

user_agent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36 Edg/119.0.0.0"
BROWSER_CONFIG = {
    "viewport_size": 1024 * 5,
    "downloads_folder": "downloads_folder",
    "request_kwargs": {
        "headers": {"User-Agent": user_agent},
        "timeout": 300,
    },
    "serpapi_key": os.getenv("SERPAPI_API_KEY"),
}

custom_role_conversions = {"tool-call": "assistant", "tool-response": "user"}

# Multimedia file types supported (using Gradio-compatible format)
ALLOWED_FILE_TYPES = [
    ".pdf",  # application/pdf
    ".docx",  # application/vnd.openxmlformats-officedocument.wordprocessingml.document
    ".txt",  # text/plain
    ".png",  # image/png
    ".webp",  # image/webp
    ".jpeg",  # image/jpeg
    ".jpg",  # image/jpeg
    ".gif",  # image/gif
    ".mp4",  # video/mp4
    ".mp3",  # audio/mpeg
    ".wav",  # audio/wav
    ".ogg",  # audio/ogg
]


def setup_environment():
    """Initialize environment variables and authentication."""
    load_dotenv(override=True)
    if os.getenv("HF_TOKEN"):  # Check if token is actually set
        login(os.getenv("HF_TOKEN"))
        print("HF_TOKEN (last 10 characters):", os.getenv("HF_TOKEN")[-10:])
    else:
        print("HF_TOKEN not found in environment variables.")


# ------------------------ Model and Tool Management ------------------------
class ModelManager:
    """Manages model loading and initialization with Zhou Protocol patterns."""

    _instance = None
    _lock = threading.Lock()

    @classmethod
    def get_instance(cls):
        """Thread-safe singleton access to model manager."""
        if cls._instance is None:
            with cls._lock:
                if cls._instance is None:
                    cls._instance = cls()
        return cls._instance

    def __init__(self):
        """Initialize with model cache."""
        self.model_cache = {}

    def load_model(self, chosen_inference: str, model_id: str, key_manager=None):
        """Load the specified model with appropriate configuration and caching."""
        cache_key = f"{chosen_inference}:{model_id}"

        # Return cached model if available
        if cache_key in self.model_cache:
            return self.model_cache[cache_key]

        try:
            if chosen_inference == "hf_api":
                model = HfApiModel(model_id=model_id)

            elif chosen_inference == "hf_api_provider":
                model = HfApiModel(provider="together")

            elif chosen_inference == "litellm":
                model = LiteLLMModel(model_id=model_id)

            elif chosen_inference == "openai":
                if not key_manager:
                    raise ValueError("Key manager required for OpenAI model")

                model = OpenAIServerModel(
                    model_id=model_id, api_key=key_manager.get_key("openai_api_key")
                )

            elif chosen_inference == "transformers":
                model = TransformersModel(
                    model_id="HuggingFaceTB/SmolLM2-1.7B-Instruct",
                    device_map="auto",
                    max_new_tokens=1000,
                )

            else:
                raise ValueError(f"Invalid inference type: {chosen_inference}")

            # Cache the model for future use
            self.model_cache[cache_key] = model
            return model

        except Exception as e:
            print(f"✗ Couldn't load model: {e}")
            raise


class ToolRegistry:
    """Manages tool initialization and organization with validation."""

    @staticmethod
    def validate_tools(tools: List[Tool]) -> List[Tool]:
        """Validate tools and filter out any None values."""
        return [tool for tool in tools if isinstance(tool, Tool)]

    @staticmethod
    def load_web_tools(model, browser, text_limit=20000):
        """Initialize and return web-related tools."""
        return [
            GoogleSearchTool(provider="serper"),
            VisitTool(browser),
            PageUpTool(browser),
            PageDownTool(browser),
            FinderTool(browser),
            FindNextTool(browser),
            ArchiveSearchTool(browser),
            TextInspectorTool(model, text_limit),
        ]

    @staticmethod
    def load_image_generation_tools():
        """Initialize and return image generation tools."""
        try:
            return Tool.from_space(
                space_id="xkerser/FLUX.1-dev",
                name="image_generator",
                description="Generates high-quality AgentImage using the FLUX.1-dev model based on text prompts.",
            )
        except Exception as e:
            print(f"✗ Couldn't initialize image generation tool: {e}")
            return None

    @staticmethod
    def load_legal_document_tool():
        """Initialize and return the legal document processing tool."""
        try:
            # Create a simple instance with default parameters
            return LegalDocumentTool()
        except Exception as e:
            print(f"✗ Couldn't initialize legal document tool: {e}")
            # Return None instead of raising to make this tool optional
            return None


# ------------------------ Session Management ------------------------
class SessionManager:
    """Manages agent sessions with proper cleanup and lifecycle management."""

    _instance = None
    _lock = threading.Lock()

    @classmethod
    def get_instance(cls):
        """Thread-safe singleton access."""
        if cls._instance is None:
            with cls._lock:
                if cls._instance is None:
                    cls._instance = cls()
        return cls._instance

    def __init__(self):
        """Initialize session management structures."""
        self.sessions = {}
        self.temp_files = {}
        self.last_activity = {}
        self._cleanup_thread = None
        self._running = False
        self._start_cleanup_thread()

    def _start_cleanup_thread(self):
        """Start a background thread for session cleanup."""
        if self._cleanup_thread is None:
            self._running = True
            self._cleanup_thread = threading.Thread(
                target=self._cleanup_inactive_sessions, daemon=True
            )
            self._cleanup_thread.start()

    def _cleanup_inactive_sessions(self):
        """Periodically clean up inactive sessions."""
        import time

        # Session timeout in seconds (30 minutes)
        SESSION_TIMEOUT = 30 * 60

        while self._running:
            current_time = time.time()

            # Find inactive sessions
            inactive_sessions = [
                session_id
                for session_id, last_time in self.last_activity.items()
                if (current_time - last_time) > SESSION_TIMEOUT
            ]

            # Clean up each inactive session
            for session_id in inactive_sessions:
                self.cleanup_session(session_id)

            # Sleep for a minute before next check
            time.sleep(60)

    def register_session(self, session_id):
        """Register a new session."""
        if session_id not in self.sessions:
            self.sessions[session_id] = {}
            self.temp_files[session_id] = []

        # Update activity timestamp
        self.last_activity[session_id] = time.time()

    def update_activity(self, session_id):
        """Update the last activity timestamp for a session."""
        self.last_activity[session_id] = time.time()

    def register_temp_file(self, session_id, file_path):
        """Register a temporary file with a session for later cleanup."""
        if session_id not in self.temp_files:
            self.temp_files[session_id] = []

        self.temp_files[session_id].append(file_path)

    def cleanup_session(self, session_id):
        """Clean up resources for a session."""
        # Remove temporary files
        if session_id in self.temp_files:
            for file_path in self.temp_files[session_id]:
                try:
                    if os.path.exists(file_path):
                        os.remove(file_path)
                except Exception as e:
                    print(f"Error removing temp file {file_path}: {e}")

            del self.temp_files[session_id]

        # Clean up session data
        if session_id in self.sessions:
            del self.sessions[session_id]

        # Clean up activity record
        if session_id in self.last_activity:
            del self.last_activity[session_id]

    def __del__(self):
        """Clean up all sessions when the manager is destroyed."""
        self._running = False
        if self._cleanup_thread and self._cleanup_thread.is_alive():
            self._cleanup_thread.join(timeout=1.0)

        # Clean up all remaining sessions
        for session_id in list(self.sessions.keys()):
            self.cleanup_session(session_id)


# ------------------------ Agent Creation and Execution ------------------------
class AgentFactory:
    """Factory for creating and managing agent instances with Zhou Protocol patterns."""

    _instance = None
    _lock = threading.Lock()

    @classmethod
    def get_instance(cls):
        """Thread-safe singleton access."""
        if cls._instance is None:
            with cls._lock:
                if cls._instance is None:
                    cls._instance = cls()
        return cls._instance

    def __init__(self):
        """Initialize with agent cache."""
        self.agent_cache = {}

    def create_agent(self, session_id: str = "default") -> CodeAgent:
        """Creates a fresh agent instance with properly configured tools."""
        # Return cached agent if available for this session
        if session_id in self.agent_cache:
            return self.agent_cache[session_id]

        # Initialize model
        model = LiteLLMModel(
            custom_role_conversions=custom_role_conversions,
            model_id="openrouter/perplexity/r1-1776",  # currently serving:
        )  # DEEPSEEK = openrouter/perplexity/r1-1776 <--- boss model

        # Initialize tools
        text_limit = 30000
        browser = SimpleTextBrowser(**BROWSER_CONFIG)

        # Collect all tools in a single list
        web_tools = ToolRegistry.load_web_tools(model, browser, text_limit)
        image_generator = ToolRegistry.load_image_generation_tools()
        legal_tool = ToolRegistry.load_legal_document_tool()

        # Combine and validate all tools
        all_tools = [visualizer] + web_tools

        # Only add tools that are properly initialized (not None)
        if image_generator:
            all_tools.append(image_generator)

        if legal_tool:
            all_tools.append(legal_tool)

        # Final validation to ensure all tools are valid
        all_tools = ToolRegistry.validate_tools(all_tools)

        agent = CodeAgent(
            model=model,
            tools=all_tools,  # Pass a single list containing all tools
            max_steps=10,
            verbosity_level=1,
            additional_authorized_imports=AUTHORIZED_IMPORTS,
            planning_interval=4,
        )

        # Cache the agent for future use
        self.agent_cache[session_id] = agent
        return agent

    def clear_agent(self, session_id: str):
        """Remove an agent from the cache."""
        if session_id in self.agent_cache:
            del self.agent_cache[session_id]


def stream_to_gradio(
    agent,
    task: str,
    reset_agent_memory: bool = False,
    additional_args: Optional[Dict[str, Any]] = None,
):
    """Runs an agent with the given task and streams messages as Gradio ChatMessages."""
    for step_log in agent.run(
        task, stream=True, reset=reset_agent_memory, additional_args=additional_args
    ):
        for message in pull_messages_from_step(step_log):
            yield message

    # Process final answer : Use a more comprehensive media output
    final_answer = step_log  # Last log is the run's final_answer
    final_answer = handle_agent_output_types(final_answer)

    if isinstance(final_answer, AgentText):
        yield gr.ChatMessage(
            role="assistant",
            content=f"**Final answer:**\n{final_answer.to_string()}\n",
        )
    elif isinstance(final_answer, AgentImage):
        yield gr.ChatMessage(
            role="assistant",
            content={"image": final_answer.to_string(), "type": "file"},
        )  # Send as Gradio-compatible file object:
    elif isinstance(final_answer, AgentAudio):
        yield gr.ChatMessage(
            role="assistant",
            content={"audio": final_answer.to_string(), "type": "file"},
        )  # Send as Gradio-compatible file object
    else:
        yield gr.ChatMessage(
            role="assistant", content=f"**Final answer:** {str(final_answer)}"
        )


# ------------------------ Gradio UI Components ------------------------
class GradioUI:
    """A Gradio-compliant interface to launch your agent with proper resource management."""

    def __init__(self):
        """Initialize the Gradio UI with proper session management."""
        self.session_manager = SessionManager.get_instance()
        self.agent_factory = AgentFactory.get_instance()
        self.temp_dir = tempfile.mkdtemp(prefix="gradio_")

    def __del__(self):
        """Clean up resources when the UI is destroyed."""
        try:
            # Clean up the temporary directory
            if os.path.exists(self.temp_dir):
                shutil.rmtree(self.temp_dir, ignore_errors=True)
        except Exception as e:
            print(f"Error cleaning up temporary directory: {e}")

    @staticmethod
    def _get_session_id(session_state):
        """Generate or retrieve a session ID."""
        if "session_id" not in session_state:
            session_state["session_id"] = f"session_{id(session_state)}"
        return session_state["session_id"]

    def interact_with_agent(self, prompt, messages, session_state):
        """Main interaction handler with the agent."""
        # Get or create session ID
        session_id = self._get_session_id(session_state)

        # Register/update the session
        self.session_manager.register_session(session_id)
        self.session_manager.update_activity(session_id)

        # Get or create session-specific agent
        agent = self.agent_factory.create_agent(session_id)

        try:
            # Log the existence of agent memory
            has_memory = hasattr(agent, "memory")
            print(f"Agent has memory: {has_memory}")
            if has_memory:
                print(f"Memory type: {type(agent.memory)}")

            messages.append(gr.ChatMessage(role="user", content=prompt))
            yield messages

            for msg in stream_to_gradio(agent, task=prompt, reset_agent_memory=False):
                messages.append(msg)
                self.session_manager.update_activity(session_id)
                yield messages  # Yield messages after each step

            yield messages  # Yield messages one last time

        except gr.Error as e:
            # Handle Gradio-specific errors
            messages.append(
                gr.ChatMessage(role="assistant", content=f"Error: {str(e)}")
            )
            yield messages
        except Exception as e:
            # Log the error but present a user-friendly message
            print(f"Error in interaction: {str(e)}")
            messages.append(
                gr.ChatMessage(
                    role="assistant",
                    content="I encountered an error processing your request. Please try again with a different query.",
                )
            )
            yield messages

    @gr.validate_input(file="file")
    def upload_file(self, file, file_uploads_log, session_state):
        """Handle file uploads with Gradio-compliant temporary file handling."""
        session_id = self._get_session_id(session_state)
        self.session_manager.update_activity(session_id)

        if file is None:
            return gr.Textbox("No file uploaded", visible=True), file_uploads_log

        try:
            # Create a temporary file with a secure random name
            temp_file_path = ""

            with tempfile.NamedTemporaryFile(delete=False, dir=self.temp_dir) as tmp:
                # Copy the uploaded file to the temporary file
                shutil.copyfileobj(open(file.name, "rb"), tmp)
                temp_file_path = tmp.name

            # Register the temporary file with the session manager
            self.session_manager.register_temp_file(session_id, temp_file_path)

            # Store the original filename for reference
            orig_filename = os.path.basename(file.name)

            return (
                gr.Textbox(f"File uploaded: {orig_filename}", visible=True),
                file_uploads_log + [(temp_file_path, orig_filename)],
            )

        except Exception as e:
            print(f"Error handling file upload: {e}")
            return (
                gr.Textbox(f"Error uploading file: {str(e)}", visible=True),
                file_uploads_log,
            )

    def log_user_message(self, text_input, file_uploads_log, session_state):
        """Process user message and handle file references."""
        session_id = self._get_session_id(session_state)
        self.session_manager.update_activity(session_id)

        message = text_input

        if file_uploads_log and len(file_uploads_log) > 0:
            # Include only the original filenames in the message
            filenames = [f[1] for f in file_uploads_log]
            message += f"\nYou have been provided with these files: {filenames}"

            # Include the actual file paths in the additional_args
            if "additional_args" not in session_state:
                session_state["additional_args"] = {}

            session_state["additional_args"]["file_paths"] = [
                f[0] for f in file_uploads_log
            ]

        return (
            message,
            gr.Textbox(
                value="",
                interactive=False,
                placeholder="Processing...",
            ),
            gr.Button(interactive=False),
            session_state,
        )

    def launch(self, **kwargs):
        """Launch the Gradio UI with responsive layout."""
        with gr.Blocks(theme="soft", css=self._get_responsive_css()) as demo:
            with gr.Row(equal_height=True) as main_row:
                # Sidebar (adapts to screen size via CSS)
                with gr.Column(scale=1, min_width=100) as sidebar:
                    gr.Markdown(
                        """# OpenDeepResearch
                        ## Powered by Smolagents"""
                    )
                    with gr.Group():
                        gr.Markdown("**What's on your mind?**", container=True)
                        text_input = gr.Textbox(
                            lines=3,
                            label="Your request",
                            container=False,
                            placeholder="Enter your prompt here and press Shift+Enter or press the button",
                        )
                        launch_research_btn = gr.Button("Run", variant="primary")

                    # Clean file upload with Gradio-compliant file_types
                    upload_file = gr.File(
                        label="Upload a file",
                        file_types=ALLOWED_FILE_TYPES,
                        type="file",
                    )
                    upload_status = gr.Textbox(
                        label="Upload Status", interactive=False, visible=False
                    )

                    # Footer with proper responsive behavior
                    with gr.Row(visible=True) as footer:
                        gr.HTML(
                            """
                            <div style="display: flex; align-items: center; gap: 8px; font-family: system-ui, -apple-system, sans-serif;">
                            <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/smolagents/mascot_smol.png"
                                style="width: 32px; height: 32px; object-fit: contain;" alt="logo">
                            <a target="_blank" href="https://github.com/huggingface/smolagents">
                                <b>huggingface/smolagents</b>
                            </a>
                            </div>
                            """
                        )

                # Main content area
                with gr.Column(scale=4, min_width=400) as content:
                    # Add session state to store session-specific data
                    session_state = gr.State({})
                    stored_messages = gr.State([])
                    file_uploads_log = gr.State([])

                    chatbot = gr.Chatbot(
                        label="OpenDeepResearch",
                        show_label=True,
                        type="messages",
                        avatar_images=(
                            None,
                            "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/smolagents/mascot_smol.png",
                        ),
                        height=600,
                        elem_id="research-chatbot",
                    )

            # Connect event handlers
            text_input.submit(
                self.log_user_message,
                [text_input, file_uploads_log, session_state],
                [stored_messages, text_input, launch_research_btn, session_state],
            ).then(
                self.interact_with_agent,
                [stored_messages, chatbot, session_state],
                [chatbot],
            ).then(
                lambda: (
                    gr.Textbox(
                        interactive=True,
                        placeholder="Enter your prompt here and press the button",
                    ),
                    gr.Button(interactive=True),
                ),
                None,
                [text_input, launch_research_btn],
            )

            launch_research_btn.click(
                self.log_user_message,
                [text_input, file_uploads_log, session_state],
                [stored_messages, text_input, launch_research_btn, session_state],
            ).then(
                self.interact_with_agent,
                [stored_messages, chatbot, session_state],
                [chatbot],
            ).then(
                lambda: (
                    gr.Textbox(
                        interactive=True,
                        placeholder="Enter your prompt here and press the button",
                    ),
                    gr.Button(interactive=True),
                ),
                None,
                [text_input, launch_research_btn],
            )

            upload_file.change(
                self.upload_file,
                [upload_file, file_uploads_log, session_state],
                [upload_status, file_uploads_log],
            )

            # Clean up session on page unload
            demo.load(
                lambda: None,
                None,
                None,
                _js="""
                () => {
                    window.addEventListener('beforeunload', function() {
                        // Notify backend about session end (would require additional endpoint)
                        console.log('Cleaning up session');
                    });
                }
            """,
            )

        demo.queue(max_size=20).launch(debug=True, **kwargs)

    def _get_responsive_css(self):
        """Get CSS for responsive layout."""
        return """
        /* Responsive layout */
        @media (max-width: 768px) {
            #research-chatbot {
                height: 400px !important;
            }
            
            /* Stack columns on small screens */
            .gradio-row {
                flex-direction: column;
            }
            
            /* Adjust column widths */
            .gradio-column {
                min-width: 100% !important;
                width: 100% !important;
            }
        }
        
        /* Base styling */
        .gradio-container {
            max-width: 100% !important;
        }
        """


# ------------------------ Execution ------------------------
def main():
    """Main entry point for the application."""
    # Initialize environment
    setup_environment()

    # Ensure downloads folder exists for browser
    os.makedirs(f"./{BROWSER_CONFIG['downloads_folder']}", exist_ok=True)

    # Launch UI
    GradioUI().launch(share=True)


if __name__ == "__main__":
    main()