app.py

import os
import gradio as gr
import requests
import inspect
import pandas as pd

# Try to import Google ADK components, fallback to simple agent if not available
try:
    from google.genai import types
    from agent import session_service, APP_NAME, USER_ID, SESSION_ID, runner
    GOOGLE_ADK_AVAILABLE = True
    print("✅ Google ADK components loaded successfully")
except ImportError as e:
    print(f"⚠️ Google ADK not available: {e}")
    print("🔄 Falling back to simple HTTP-based agent")
    GOOGLE_ADK_AVAILABLE = False

# (Keep Constants as is)
# --- Constants ---
DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"

# --- Fallback Simple Agent for when Google ADK is not available ---
class SimpleAgent:
    def __init__(self):
        print("SimpleAgent initialized - using basic HTTP requests")

    def __call__(self, question: str) -> str:
        print(f"SimpleAgent received question (first 50 chars): {question[:50]}...")

        try:
            # Analyze the question to understand what's needed
            question_lower = question.lower()

            # Detect GAIA-style complex questions
            if self._is_complex_gaia_question(question):
                return self._handle_complex_question(question)

            # Check if it's a math question
            elif any(word in question_lower for word in ['calculate', 'sum', 'total', 'add', 'multiply', 'divide']):
                return self._try_basic_math(question)

            # Check if it's asking for a count or number
            elif any(word in question_lower for word in ['how many', 'count', 'number of']):
                return "I would need to analyze the data to count the items. [SimpleAgent - limited capabilities]"

            # Check if it's asking about a file
            elif 'file' in question_lower or 'excel' in question_lower or 'csv' in question_lower:
                return "I would need to download and analyze the file to answer this question. [SimpleAgent - limited capabilities]"

            # Check if it's asking about a person or entity
            elif any(word in question_lower for word in ['who is', 'who are', 'what is']):
                return "I would need to search for information about this topic. [SimpleAgent - limited capabilities]"

            # Default response
            else:
                return f"I received your question but need more advanced capabilities to answer it properly. Question: {question[:200]}... [SimpleAgent - Google ADK not available]"

        except Exception as e:
            return f"Error processing question: {str(e)} [SimpleAgent]"

    def _is_complex_gaia_question(self, question):
        """Detect if this is a complex GAIA-style question requiring multiple steps"""
        indicators = [
            'painting', 'film', 'movie', 'ocean liner', 'ship', 'menu',
            'clockwise', 'order', 'arrangement', 'position',
            'comma-separated', 'list', 'plural form',
            'served as part of', 'later used as', 'floating prop'
        ]
        question_lower = question.lower()
        return sum(1 for indicator in indicators if indicator in question_lower) >= 3

    def _handle_complex_question(self, question):
        """Handle complex GAIA questions with basic analysis"""
        question_lower = question.lower()

        # Identify what the question is asking for
        steps_needed = []

        if 'painting' in question_lower:
            steps_needed.append("🎨 Analyze painting/image")
        if any(word in question_lower for word in ['film', 'movie']):
            steps_needed.append("🎬 Research film information")
        if any(word in question_lower for word in ['ocean liner', 'ship']):
            steps_needed.append("🚢 Research ship/vessel details")
        if 'menu' in question_lower:
            steps_needed.append("📋 Find historical menu information")
        if any(word in question_lower for word in ['clockwise', 'order', 'arrangement']):
            steps_needed.append("🔄 Analyze spatial arrangement")

        analysis = f"This appears to be a complex GAIA question requiring multiple steps:\n"
        for i, step in enumerate(steps_needed, 1):
            analysis += f"{i}. {step}\n"

        analysis += "\nI would need advanced capabilities including:\n"
        analysis += "- Image analysis for visual content\n"
        analysis += "- Web search for historical/factual information\n"
        analysis += "- Multi-step reasoning to connect different pieces of information\n"
        analysis += "\n[SimpleAgent - Complex GAIA question detected but cannot solve]"

        return analysis

    def _try_basic_math(self, question):
        """Try to extract and solve basic math from the question"""
        try:
            # Very basic math extraction - look for numbers
            import re
            numbers = re.findall(r'\d+\.?\d*', question)
            if len(numbers) >= 2:
                nums = [float(n) for n in numbers[:2]]
                if 'add' in question.lower() or 'sum' in question.lower():
                    result = nums[0] + nums[1]
                    return f"Basic calculation: {nums[0]} + {nums[1]} = {result} [SimpleAgent - basic math]"
                elif 'multiply' in question.lower():
                    result = nums[0] * nums[1]
                    return f"Basic calculation: {nums[0]} × {nums[1]} = {result} [SimpleAgent - basic math]"

            return "I can see this involves math but need more advanced capabilities to solve it. [SimpleAgent - limited math]"
        except:
            return "I can see this involves math but couldn't parse it. [SimpleAgent - limited math]"

# --- Google ADK Agent Wrapper ---
# ----- USING THE ACTUAL GOOGLE ADK AGENT FROM AGENT.PY ------
class GoogleADKAgent:
    def __init__(self):
        print("GoogleADKAgent initialized with Google ADK runner and agents.")

        try:
            # Use the pre-configured runner and root_agent from agent.py
            self.runner = runner
            self.session_service = session_service
            self.app_name = APP_NAME
            self.user_id = USER_ID
            self.question_counter = 0  # To create unique session IDs for each question
            self.initialized = True
            print("✅ Google ADK Agent successfully initialized using pre-configured runner")

        except Exception as e:
            print(f"❌ Failed to initialize Google ADK Agent: {e}")
            self.initialized = False
            raise e

    def __call__(self, question: str) -> str:
        print(f"Agent received question (first 50 chars): {question[:50]}...")

        if not self.initialized:
            return "Google ADK Agent not properly initialized"

        try:
            # Create a unique session for each question to avoid state conflicts
            self.question_counter += 1
            unique_session_id = f"{SESSION_ID}_q{self.question_counter}"

            # Create the session before using it
            try:
                self.session_service.create_session(
                    app_name=self.app_name,
                    user_id=self.user_id,
                    session_id=unique_session_id
                )
                print(f"✅ Created session: {unique_session_id}")
            except Exception as session_error:
                print(f"⚠️ Session creation error: {session_error}")
                # Fallback to the default session
                unique_session_id = SESSION_ID

            # Create the query content
            query_content = types.Content(
                role='user',
                parts=[types.Part(text=question)]
            )

            # Run the agent synchronously using the runner with correct parameters
            print(f"🚀 Running agent with session: {unique_session_id}")
            events = list(self.runner.run(
                user_id=self.user_id,
                session_id=unique_session_id,
                new_message=query_content
            ))

            print(f"📊 Generated {len(events)} events")

            # Debug: Print event details
            for i, event in enumerate(events):
                print(f"Event {i}: author={getattr(event, 'author', 'unknown')}, content_type={type(getattr(event, 'content', None))}")
                if hasattr(event, 'content') and event.content and hasattr(event.content, 'parts'):
                    for j, part in enumerate(event.content.parts):
                        if hasattr(part, 'text') and part.text:
                            print(f"  Part {j}: {part.text[:100]}...")

            # Extract the final answer from the events
            final_answer = "No response generated."

            # Extract the final answer with GAIA-specific processing
            final_answer = self._extract_gaia_answer(events)

            # Clean up the answer for exact matching
            final_answer = self._clean_answer_for_exact_match(final_answer)

            print(f"Agent returning answer: {final_answer[:100]}...")
            return final_answer

        except Exception as e:
            error_msg = f"Error running Google ADK agent: {str(e)}"
            print(error_msg)
            return error_msg

    def _extract_gaia_answer(self, events):
        """Extract the final answer from events with GAIA-specific logic"""
        final_answer = "No response generated."

        # Collect all text responses from the agent
        all_responses = []
        for event in events:
            if event.content and event.content.parts:
                for part in event.content.parts:
                    if part.text and part.text.strip():
                        text = part.text.strip()
                        # Skip system messages and tool calls, but keep substantial responses
                        if (not text.startswith("I'll") and
                            not text.startswith("Let me") and
                            not text.startswith("I need to") and
                            len(text) > 10):
                            all_responses.append(text)

        # For GAIA questions, prefer the last substantial response
        if all_responses:
            # Look for responses that seem like final answers
            for response in reversed(all_responses):
                # Skip responses that are clearly intermediate steps
                if not any(phrase in response.lower() for phrase in [
                    "let me", "i need to", "first", "next", "then", "now i'll"
                ]):
                    final_answer = response
                    break

            # If no clear final answer, use the last response
            if final_answer == "No response generated.":
                final_answer = all_responses[-1]
        else:
            # Fallback: get any text response
            for event in reversed(events):
                if event.content and event.content.parts:
                    for part in event.content.parts:
                        if part.text and part.text.strip():
                            final_answer = part.text.strip()
                            break
                    if final_answer != "No response generated.":
                        break

        return final_answer

    def _clean_answer_for_exact_match(self, answer):
        """Clean the answer for exact matching requirements"""
        if not answer or answer == "No response generated.":
            return answer

        # Remove common prefixes that agents might add
        prefixes_to_remove = [
            "The answer is: ",
            "Answer: ",
            "Final answer: ",
            "FINAL ANSWER: ",
            "Based on my analysis, ",
            "The result is: ",
        ]

        cleaned = answer
        for prefix in prefixes_to_remove:
            if cleaned.startswith(prefix):
                cleaned = cleaned[len(prefix):]

        # Remove trailing explanations in brackets or parentheses
        import re
        cleaned = re.sub(r'\s*\[.*?\]\s*$', '', cleaned)
        cleaned = re.sub(r'\s*\(.*?\)\s*$', '', cleaned)

        # Clean up whitespace
        cleaned = cleaned.strip()

        return cleaned

def run_and_submit_all( profile: gr.OAuthProfile | None):
    """
    Fetches all questions, runs the BasicAgent on them, submits all answers,
    and displays the results.
    """
    # --- Determine HF Space Runtime URL and Repo URL ---
    space_id = os.getenv("SPACE_ID") # Get the SPACE_ID for sending link to the code

    if profile:
        username= f"{profile.username}"
        print(f"User logged in: {username}")
    else:
        print("User not logged in.")
        return "Please Login to Hugging Face with the button.", None

    api_url = DEFAULT_API_URL
    questions_url = f"{api_url}/questions"
    submit_url = f"{api_url}/submit"

    # 1. Instantiate Agent ( modify this part to create your agent)
    try:
        if GOOGLE_ADK_AVAILABLE:
            agent = GoogleADKAgent()
            print("✅ Using Google ADK Agent")
        else:
            agent = SimpleAgent()
            print("⚠️ Using Simple Agent (Google ADK not available)")
    except Exception as e:
        print(f"Error instantiating agent: {e}")
        # Fallback to simple agent if Google ADK fails
        try:
            agent = SimpleAgent()
            print("🔄 Fallback to Simple Agent due to error")
        except Exception as e2:
            print(f"Error with fallback agent: {e2}")
            return f"Error initializing any agent: {e}, {e2}", None
    # In the case of an app running as a hugging Face space, this link points toward your codebase ( usefull for others so please keep it public)
    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
    print(agent_code)

    # 2. Fetch Questions
    print(f"Fetching questions from: {questions_url}")
    try:
        response = requests.get(questions_url, timeout=15)
        response.raise_for_status()
        questions_data = response.json()
        if not questions_data:
             print("Fetched questions list is empty.")
             return "Fetched questions list is empty or invalid format.", None
        print(f"Fetched {len(questions_data)} questions.")
    except requests.exceptions.RequestException as e:
        print(f"Error fetching questions: {e}")
        return f"Error fetching questions: {e}", None
    except requests.exceptions.JSONDecodeError as e:
         print(f"Error decoding JSON response from questions endpoint: {e}")
         print(f"Response text: {response.text[:500]}")
         return f"Error decoding server response for questions: {e}", None
    except Exception as e:
        print(f"An unexpected error occurred fetching questions: {e}")
        return f"An unexpected error occurred fetching questions: {e}", None

    # 3. Run your Agent
    results_log = []
    answers_payload = []
    print(f"Running agent on {len(questions_data)} questions...")
    for item in questions_data:
        task_id = item.get("task_id")
        question_text = item.get("question")
        file_name = item.get("file_name", "")

        if not task_id or question_text is None:
            print(f"Skipping item with missing task_id or question: {item}")
            continue

        # Enhance question with file information if available
        enhanced_question = question_text
        if file_name:
            file_url = f"https://agents-course-unit4-scoring.hf.space/files/{task_id}"
            enhanced_question = f"{question_text}\n\nFile available at: {file_url}"

        try:
            submitted_answer = agent(enhanced_question)
            answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer})
            results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": submitted_answer})
        except Exception as e:
             print(f"Error running agent on task {task_id}: {e}")
             results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": f"AGENT ERROR: {e}"})

    if not answers_payload:
        print("Agent did not produce any answers to submit.")
        return "Agent did not produce any answers to submit.", pd.DataFrame(results_log)

    # 4. Prepare Submission 
    submission_data = {"username": username.strip(), "agent_code": agent_code, "answers": answers_payload}
    status_update = f"Agent finished. Submitting {len(answers_payload)} answers for user '{username}'..."
    print(status_update)

    # 5. Submit
    print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
    try:
        response = requests.post(submit_url, json=submission_data, timeout=60)
        response.raise_for_status()
        result_data = response.json()
        final_status = (
            f"Submission Successful!\n"
            f"User: {result_data.get('username')}\n"
            f"Overall Score: {result_data.get('score', 'N/A')}% "
            f"({result_data.get('correct_count', '?')}/{result_data.get('total_attempted', '?')} correct)\n"
            f"Message: {result_data.get('message', 'No message received.')}"
        )
        print("Submission successful.")
        results_df = pd.DataFrame(results_log)
        return final_status, results_df
    except requests.exceptions.HTTPError as e:
        error_detail = f"Server responded with status {e.response.status_code}."
        try:
            error_json = e.response.json()
            error_detail += f" Detail: {error_json.get('detail', e.response.text)}"
        except requests.exceptions.JSONDecodeError:
            error_detail += f" Response: {e.response.text[:500]}"
        status_message = f"Submission Failed: {error_detail}"
        print(status_message)
        results_df = pd.DataFrame(results_log)
        return status_message, results_df
    except requests.exceptions.Timeout:
        status_message = "Submission Failed: The request timed out."
        print(status_message)
        results_df = pd.DataFrame(results_log)
        return status_message, results_df
    except requests.exceptions.RequestException as e:
        status_message = f"Submission Failed: Network error - {e}"
        print(status_message)
        results_df = pd.DataFrame(results_log)
        return status_message, results_df
    except Exception as e:
        status_message = f"An unexpected error occurred during submission: {e}"
        print(status_message)
        results_df = pd.DataFrame(results_log)
        return status_message, results_df


# --- Build Gradio Interface using Blocks ---
with gr.Blocks() as demo:
    gr.Markdown("# 🤖 GAIA Benchmark Agent Evaluation")

    # Add dynamic status message based on agent availability
    if GOOGLE_ADK_AVAILABLE:
        status_msg = "✅ **Google ADK Agent Active** - Full capabilities for complex GAIA questions including multi-step reasoning, web search, code execution, file analysis, and multimodal understanding."
    else:
        status_msg = "⚠️ **Simple Agent Active** - Limited capabilities. Google ADK not available in this environment. Can detect GAIA question types but cannot solve them."

    gr.Markdown(f"**Agent Status:** {status_msg}")

    gr.Markdown(
        """
        ## About GAIA Benchmark

        This evaluation uses questions from the **GAIA benchmark** - a challenging dataset that tests AI agents on:
        - 🔍 **Multi-step reasoning** across different domains
        - 🖼️ **Multimodal understanding** (text, images, files)
        - 🔗 **Multi-hop information retrieval**
        - 📊 **Structured output formatting**
        - 🎯 **Exact answer matching**

        **Example GAIA Question:**
        *"Which of the fruits shown in the 2008 painting 'Embroidery from Uzbekistan' were served as part of the October 1949 breakfast menu for the ocean liner that was later used as a floating prop for the film 'The Last Voyage'?"*

        ---

        **Instructions:**
        1. **Clone this space** and customize the agent code for your approach
        2. **Log in** to your Hugging Face account using the button below
        3. **Run Evaluation** to test your agent on 20 filtered GAIA questions
        4. **Submit answers** for scoring with exact match evaluation

        **Target:** Aim for ~30% accuracy on Level 1 GAIA questions (current benchmark performance)

        ---
        **Note:** Evaluation may take several minutes as the agent processes complex multi-step questions.
        """
    )

    gr.LoginButton()

    run_button = gr.Button("Run Evaluation & Submit All Answers")

    status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
    # Removed max_rows=10 from DataFrame constructor
    results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)

    run_button.click(
        fn=run_and_submit_all,
        outputs=[status_output, results_table]
    )

if __name__ == "__main__":
    print("\n" + "-"*30 + " App Starting " + "-"*30)
    # Check for SPACE_HOST and SPACE_ID at startup for information
    space_host_startup = os.getenv("SPACE_HOST")
    space_id_startup = os.getenv("SPACE_ID") # Get SPACE_ID at startup

    if space_host_startup:
        print(f"✅ SPACE_HOST found: {space_host_startup}")
        print(f"   Runtime URL should be: https://{space_host_startup}.hf.space")
    else:
        print("ℹ️  SPACE_HOST environment variable not found (running locally?).")

    if space_id_startup: # Print repo URLs if SPACE_ID is found
        print(f"✅ SPACE_ID found: {space_id_startup}")
        print(f"   Repo URL: https://huggingface.co/spaces/{space_id_startup}")
        print(f"   Repo Tree URL: https://huggingface.co/spaces/{space_id_startup}/tree/main")
    else:
        print("ℹ️  SPACE_ID environment variable not found (running locally?). Repo URL cannot be determined.")

    print("-"*(60 + len(" App Starting ")) + "\n")

    print("Launching Gradio Interface for Basic Agent Evaluation...")
    demo.launch(debug=True, share=False)