import os
import openai
import gradio as gr
import requests
import pandas as pd
from smolagents import CodeAgent, LiteLLMModel, DuckDuckGoSearchTool, tool

# --- Constants ---
DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"

# --- Configure OpenAI SDK ---
openai_api_key = os.getenv("OPENAI_API_KEY")
if not openai_api_key:
    raise RuntimeError("Please set OPENAI_API_KEY in your Space secrets.")
openai.api_key = openai_api_key
openai_model_id = os.getenv("OPENAI_MODEL_ID", "gpt-4.1")

# --- Tool Definitions ---

@tool
def summarize_query(query: str) -> str:
    """
    Reframes an unclear query into a better one.

    Args:
        query (str): The search query to refine.
    Returns:
        str: A concise, improved query.
    """
    return f"Summarize and reframe: {query}"

@tool
def wikipedia_search(page: str) -> str:
    """
    Fetches the summary extract of an English Wikipedia page.

    Args:
        page (str): The page title (e.g. 'Mercedes_Sosa_discography').
    Returns:
        str: The extract section text.
    """
    url = f"https://en.wikipedia.org/api/rest_v1/page/summary/{page}"
    resp = requests.get(url, timeout=10)
    resp.raise_for_status()
    return resp.json().get("extract", "")

search_tool    = DuckDuckGoSearchTool()
wiki_tool      = wikipedia_search
summarize_tool = summarize_query

# --- ReACT + Scratchpad + Auto-Retry Prompt ---

instruction_prompt = """
You are a ReACT agent with three tools: 
 • DuckDuckGoSearchTool(query: str)
 • wikipedia_search(page: str)
 • summarize_query(query: str)

Internally, for each question:
1. Thought: decide which tool to call.
2. Action: call the chosen tool.
3. Observation: record the result.
4. If empty/irrelevant:
   Thought: retry with summarize_query + DuckDuckGoSearchTool.
   Record new Observation.
5. Thought: integrate observations.

Finally, output your answer with the following template: FINAL ANSWER: [YOUR FINAL ANSWER]. YOUR FINAL ANSWER should be a number OR as few words as possible OR a comma separated list of numbers and/or strings. If you are asked for a number, don't use comma to write your number neither use units such as $ or percent sign unless specified otherwise. If you are asked for a string, don't use articles, neither abbreviations (e.g. for cities), and write the digits in plain text unless specified otherwise. If you are asked for a comma separated list, apply the above rules depending of whether the element to be put in the list is a number or a string.


"""

# --- Build the CodeAgent with the same model ---

llm = LiteLLMModel(
    model_id=openai_model_id,
    api_key=openai_api_key
)

smart_agent = CodeAgent(
    tools=[search_tool, wiki_tool, summarize_tool],
    model=llm
)

# --- Wrap in BasicAgent for Gradio ---

class BasicAgent:
    def __init__(self):
        print("SmolAgent (GPT-4.1) with ReACT & tools initialized.")

    def __call__(self, question: str) -> str:
        prompt = instruction_prompt.strip() + "\n\nQUESTION: " + question.strip()
        print(f"Agent prompt: {prompt[:120]}…")
        try:
            return smart_agent.run(prompt)
        except Exception as e:
            return f"AGENT ERROR: {e}"

# --- Gradio & Submission Logic ---

def run_and_submit_all(profile: gr.OAuthProfile | None):
    if not profile:
        return "Please log in to Hugging Face.", None
    username = profile.username
    space_id = os.getenv("SPACE_ID", "")
    agent = BasicAgent()
    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"

    try:
        resp = requests.get(f"{DEFAULT_API_URL}/questions", timeout=15)
        resp.raise_for_status()
        questions = resp.json() or []
    except Exception as e:
        return f"Error fetching questions: {e}", None

    logs, payload = [], []
    for item in questions:
        tid = item.get("task_id")
        q   = item.get("question")
        if not tid or not q:
            continue
        ans = agent(q)
        logs.append({"Task ID": tid, "Question": q, "Submitted Answer": ans})
        payload.append({"task_id": tid, "submitted_answer": ans})

    if not payload:
        return "Agent did not produce any answers.", pd.DataFrame(logs)

    submission = {"username": username, "agent_code": agent_code, "answers": payload}
    try:
        post = requests.post(f"{DEFAULT_API_URL}/submit", json=submission, timeout=60)
        post.raise_for_status()
        res = post.json()
        status = (
            f"Submission Successful!\n"
            f"User: {res.get('username')}\n"
            f"Overall Score: {res.get('score', 'N/A')}% "
            f"({res.get('correct_count', '?')}/{res.get('total_attempted', '?')})\n"
            f"Message: {res.get('message','')}"
        )
        return status, pd.DataFrame(logs)
    except Exception as e:
        return f"Submission Failed: {e}", pd.DataFrame(logs)

with gr.Blocks() as demo:
    gr.Markdown("# SmolAgent GAIA Runner 🚀")
    gr.Markdown("""
**Instructions:**
1. Clone this space.
2. Add `OPENAI_API_KEY` (and optionally `OPENAI_MODEL_ID`) in Settings → Secrets.
3. Log in to Hugging Face.
4. Click **Run Evaluation & Submit All Answers**.
""")
    gr.LoginButton()
    run_btn = gr.Button("Run Evaluation & Submit All Answers")
    status_out = gr.Textbox(label="Status", lines=5, interactive=False)
    table_out  = gr.DataFrame(label="Questions & Answers", wrap=True)

    run_btn.click(fn=run_and_submit_all, outputs=[status_out, table_out])

if __name__ == "__main__":
    demo.launch(debug=True, share=False)