import gradio as gr
from huggingface_hub import InferenceClient
from PyPDF2 import PdfReader
import faiss
import numpy as np
from sentence_transformers import SentenceTransformer

# --- 1. Cargar y procesar el PDF ---
pdf_path = "ley_26206.pdf"
reader = PdfReader(pdf_path)

# Extraer texto
text_chunks = []
for page in reader.pages:
    text = page.extract_text()
    if text:
        # Dividir cada página en fragmentos de 500 caracteres
        for i in range(0, len(text), 500):
            text_chunks.append(text[i:i+500])

# --- 2. Crear embeddings ---
model_emb = SentenceTransformer("all-MiniLM-L6-v2")
embeddings = model_emb.encode(text_chunks, convert_to_numpy=True)

# Índice FAISS para búsquedas rápidas
index = faiss.IndexFlatL2(embeddings.shape[1])
index.add(embeddings)

def search_pdf(query, top_k=3):
    """Busca en el PDF los fragmentos más relevantes"""
    q_emb = model_emb.encode([query], convert_to_numpy=True)
    D, I = index.search(q_emb, top_k)
    results = [text_chunks[i] for i in I[0]]
    return "\n".join(results)

# --- 3. Chatbot con búsqueda en PDF ---
def respond(
    message,
    history: list[dict[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
    hf_token: gr.OAuthToken,
):
    client = InferenceClient(token=hf_token.token, model="openai/gpt-oss-20b")

    # Buscar en el PDF
    pdf_context = search_pdf(message)

    # Si encontró información, se agrega al system message
    if pdf_context.strip():
        system_message += f"\nUsa la siguiente información de la Ley 26.206 si es relevante:\n{pdf_context}"

    messages = [{"role": "system", "content": system_message}]
    messages.extend(history)
    messages.append({"role": "user", "content": message})

    response = ""
    for msg in client.chat_completion(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        choices = msg.choices
        token = ""
        if len(choices) and choices[0].delta.content:
            token = choices[0].delta.content

        response += token
        yield response

# --- 4. Interfaz ---
chatbot = gr.ChatInterface(
    respond,
    type="messages",
    additional_inputs=[
        gr.Textbox(value="Eres un asistente experto en la Ley 26.206.", label="System message"),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p"),
    ],
)

with gr.Blocks() as demo:
    gr.Markdown("# 📘 Chatbot sobre la Ley 26.206")
    with gr.Sidebar():
        gr.LoginButton()
    chatbot.render()

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