app.py

# app.py
import os
import gradio as gr
import pandas as pd
from bs4 import BeautifulSoup # Keep this if your tools use it
import datetime
import pytz
import math
import re
import requests
import traceback

import sys
print(f"--- Python version: {sys.version} ---")
# print(f"--- Python sys.path (module search paths): {sys.path} ---") # Optional now

import transformers
from transformers.tools import Tool
print(f"--- Expected Transformers Version: 4.36.0 ---")
print(f"--- Actual Transformers Version: {transformers.__version__} ---")
# print(f"--- Transformers module loaded from: {transformers.__file__} ---") # Optional now
# print(f"--- Attributes of 'transformers' module (dir(transformers)): {dir(transformers)} ---") # Optional now

try:
    from transformers import HfAgent # <<< --- THE CORRECT IMPORT!
    print("--- Successfully imported HfAgent directly from transformers! ---")
except ImportError as e:
    print(f"--- FAILED to import HfAgent directly from transformers: {e} ---")
    # This should ideally not happen now
    raise
except Exception as e_gen:
    print(f"--- Some other UNEXPECTED error during HfAgent import: {e_gen} ---")
    raise

print("--- If no errors above, imports were successful. Proceeding with rest of app. ---")

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

# --- Tool Definitions ---
def get_current_time_in_timezone(timezone: str) -> str:
    """Fetches the current local time in a specified IANA timezone (e.g., 'America/New_York', 'Europe/London', 'UTC').
    Args:
        timezone (str): A string representing a valid IANA timezone name.
    """
    print(f"--- Tool: Executing get_current_time_in_timezone for: {timezone} ---")
    try:
        tz = pytz.timezone(timezone)
        # Added %Z (timezone name) and %z (UTC offset)
        local_time = datetime.datetime.now(tz).strftime("%Y-%m-%d %H:%M:%S %Z%z")
        return f"The current local time in {timezone} is: {local_time}"
    except pytz.exceptions.UnknownTimeZoneError:
        print(f"Error: Unknown timezone '{timezone}'")
        return f"Error: Unknown timezone '{timezone}'. Please use a valid IANA timezone name (e.g., 'America/Denver', 'UTC')."
    except Exception as e:
        print(f"Error fetching time for timezone '{timezone}': {str(e)}")
        return f"Error fetching time for timezone '{timezone}': {str(e)}"

def web_search(query: str) -> str:
    """
    Performs a web search using DuckDuckGo (via HTML scraping) and returns the text content of the top result snippets.
    Use this tool to find up-to-date information about events, facts, or topics when the answer isn't already known.

    Args:
        query (str): The search query string.

    Returns:
        str: A string containing the summarized search results (titles and snippets of top hits), or an error message if the search fails.
    """
    print(f"--- Tool: Executing web_search with query: {query} ---")
    try:
        search_url = "https://html.duckduckgo.com/html/"
        params = {"q": query}
        headers = {'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/100.0.4896.127 Safari/537.36'} # Common user agent

        response = requests.post(search_url, data=params, headers=headers, timeout=15) # Increased timeout
        response.raise_for_status() # Check for HTTP errors (4xx or 5xx)

        soup = BeautifulSoup(response.text, 'html.parser')
        results = soup.find_all('div', class_='result__body') # Find result containers

        snippets = []
        for i, result in enumerate(results[:3]): # Get top 3 results for brevity
            title_tag = result.find('a', class_='result__a')
            snippet_tag = result.find('a', class_='result__snippet')
            title = title_tag.get_text(strip=True) if title_tag else "No Title"
            snippet = snippet_tag.get_text(strip=True) if snippet_tag else "No Snippet"
            if snippet != "No Snippet": # Only include results with a snippet
                snippets.append(f"Result {i+1}: {title} - {snippet}")

        if not snippets:
            return "No search results with relevant snippets found."

        return "\n".join(snippets)

    except requests.exceptions.Timeout:
        print(f"Error during web search request: Timeout")
        return "Error: The web search request timed out."
    except requests.exceptions.RequestException as e:
        print(f"Error during web search request: {e}")
        return f"Error: Could not perform web search. Network issue: {e}"
    except Exception as e:
        print(f"Error processing web search results: {e}")
        return f"Error: Could not process search results. {e}"

def safe_calculator(expression: str) -> str:
    """
    Evaluates a simple mathematical expression involving numbers, +, -, *, /, %, parentheses, and the math functions: sqrt, pow.
    Use this tool *only* for calculations. Do not use it to run other code.

    Args:
        expression (str): The mathematical expression string (e.g., "(2 + 3) * 4", "pow(2, 5)", "sqrt(16)").

    Returns:
        str: The numerical result of the calculation or a descriptive error message if the expression is invalid or unsafe.
    """
    print(f"--- Tool: Executing safe_calculator with expression: {expression} ---")
    try:
        # Basic check for allowed characters/patterns first
        # Allows numbers (including scientific notation), operators, parentheses, whitespace, and known function names
        pattern = r"^[0-9eE\.\+\-\*\/\%\(\)\s]*(sqrt|pow)?[0-9eE\.\+\-\*\/\%\(\)\s\,]*$"
        if not re.match(pattern, expression):
             # Fallback simple pattern check (less precise)
             allowed_chars_pattern = r"^[0-9eE\.\+\-\*\/\%\(\)\s\,sqrtpow]+$"
             if not re.match(allowed_chars_pattern, expression):
                raise ValueError(f"Expression '{expression}' contains disallowed characters.")

        # Define allowed functions/names for eval's context
        allowed_names = {
            "sqrt": math.sqrt,
            "pow": math.pow,
            # Add other safe math functions if needed e.g. "log": math.log
        }
        # Evaluate the expression in a restricted environment
        # Limited builtins, only allowed names are accessible.
        result = eval(expression, {"__builtins__": {}}, allowed_names)

        # Ensure the result is a number before converting to string
        if not isinstance(result, (int, float)):
             raise ValueError("Calculation did not produce a numerical result.")

        return str(result)
    except Exception as e:
        # Catch potential errors during eval (SyntaxError, NameError, TypeError etc.) or from the checks
        print(f"Error during calculation for '{expression}': {e}")
        return f"Error calculating '{expression}': Invalid expression or calculation error ({e})."

# --- Basic Agent Definition ---
# ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
class HfAgentWrapper:
    def __init__(self):
        print("Initializing HfAgentWrapper...")
        model_repo = "google/gemma-1.1-7b-it"  # <<< --- CORRECTED MODEL REPO ID
        print(f"Loading HfAgent with model: {model_repo}")

        try:
            if not get_current_time_in_timezone.__doc__:
                raise ValueError("Tool 'get_current_time_in_timezone' is missing a docstring.")
            if not web_search.__doc__:
                raise ValueError("Tool 'web_search' is missing a docstring.")
            if not safe_calculator.__doc__:
                raise ValueError("Tool 'safe_calculator' is missing a docstring.")

            time_tool_obj = Tool(
                name=get_current_time_in_timezone.__name__,
                func=get_current_time_in_timezone,
                description=get_current_time_in_timezone.__doc__
            )
            search_tool_obj = Tool(
                name=web_search.__name__,
                func=web_search,
                description=web_search.__doc__
            )
            calculator_tool_obj = Tool(
                name=safe_calculator.__name__,
                func=safe_calculator,
                description=safe_calculator.__doc__
            )

            self.actual_tools_for_agent = [time_tool_obj, search_tool_obj, calculator_tool_obj]
            
            tool_names_for_log = [tool.name for tool in self.actual_tools_for_agent]
            print(f"Prepared Tool objects with names: {tool_names_for_log} and descriptions.")

            # Initialize HfAgent with the corrected model_repo and without the 'tools=[]' argument
            self.agent = HfAgent(
                model_repo, # Uses the corrected model_repo string
                additional_tools=self.actual_tools_for_agent
            )
            print(f"HfAgent successfully instantiated with model {model_repo}.")

            # --- Robust toolbox inspection (keep this as it is) ---
            print(f"Inspecting self.agent.toolbox...")
            # ... (the rest of your toolbox inspection code from the previous version) ...
            if hasattr(self.agent, 'toolbox'):
                toolbox_attr = self.agent.toolbox
                print(f"Type of self.agent.toolbox: {type(toolbox_attr)}")

                if isinstance(toolbox_attr, dict):
                    # ... (dictionary inspection logic) ...
                    print(f"Toolbox is a dict. Keys: {list(toolbox_attr.keys())}")
                    tools_info_from_dict = {}
                    for key, value in list(toolbox_attr.items())[:5]: 
                        tool_name_in_dict = key 
                        if hasattr(value, 'name'): tool_name_in_dict = value.name
                        elif isinstance(value, dict) and 'name' in value: tool_name_in_dict = value['name']
                        tools_info_from_dict[key] = {'name_attr': tool_name_in_dict, 'type': str(type(value))}
                    print(f"Sample of toolbox dict content: {tools_info_from_dict}")
                    print(f"Names of tools we passed to HfAgent: {[tool.name for tool in self.actual_tools_for_agent]}")
                elif hasattr(toolbox_attr, 'tools') and callable(toolbox_attr.tools):
                    try:
                        tools_in_toolbox = toolbox_attr.tools()
                        print(f"Toolbox has .tools() method. Tool names: {[tool.name for tool in tools_in_toolbox if hasattr(tool, 'name')]}")
                    except Exception as e_toolbox_call:
                        print(f"Error calling or processing self.agent.toolbox.tools(): {e_toolbox_call}")
                else:
                    print(f"Toolbox is of type {type(toolbox_attr)} but not a dict or recognized Toolbox object for this debug print.")
            else:
                print("self.agent does not have a 'toolbox' attribute (this would be unexpected).")
            # --- End of robust toolbox inspection ---

        except Exception as e:
            print(f"CRITICAL ERROR: Failed to initialize HfAgent: {e}")
            import traceback 
            print("Full traceback of HfAgent initialization error:")
            traceback.print_exc() 
            raise RuntimeError(f"HfAgent initialization failed: {e}") from e

    # __call__ method remains the same
    def __call__(self, question: str) -> str:
        print(f"\n--- HfAgentWrapper received question (first 100 chars): {question[:100]}... ---")
        try:
            answer = self.agent.run(question)
            print(f"--- HfAgentWrapper generated answer (first 100 chars): {str(answer)[:100]}... ---")
            return str(answer)
        except Exception as e:
            print(f"ERROR: HfAgent execution failed for question '{question[:50]}...': {e}")
            import traceback
            print("Full traceback of HfAgent execution error:")
            traceback.print_exc()
            return f"Agent Error: Failed to process the question. Details: {e}"

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:
        agent = HfAgentWrapper()
    except Exception as e:
        print(f"Error instantiating agent: {e}")
        return f"Error initializing agent: {e}", 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")
        if not task_id or question_text is None:
            print(f"Skipping item with missing task_id or question: {item}")
            continue
        try:
            submitted_answer = agent(question_text)
            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("# Basic Agent Evaluation Runner")
    gr.Markdown(
        """
        **Instructions:**

        1.  Please clone this space, then modify the code to define your agent's logic, the tools, the necessary packages, etc ...
        2.  Log in to your Hugging Face account using the button below. This uses your HF username for submission.
        3.  Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.

        ---
        **Disclaimers:**
        Once clicking on the "submit button, it can take quite some time ( this is the time for the agent to go through all the questions).
        This space provides a basic setup and is intentionally sub-optimal to encourage you to develop your own, more robust solution. For instance for the delay process of the submit button, a solution could be to cache the answers and submit in a seperate action or even to answer the questions in async.
        """
    )

    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)