server.py

from mcp.server.fastmcp import FastMCP, Context
from datetime import datetime
from llama_index.core import VectorStoreIndex
from llama_index.core import (
    StorageContext,
    load_index_from_storage,
)
from llama_index.tools.arxiv import ArxivToolSpec
from llama_index.core import Settings
from llama_index.llms.azure_openai import AzureOpenAI
from llama_index.embeddings.azure_openai import AzureOpenAIEmbedding
from llama_index.llms.openai import OpenAI
from llama_index.embeddings.openai import OpenAIEmbedding
from typing import Optional, List, Dict, Any
from pathlib import Path
import json
import os
import aiohttp  # Necesario para las peticiones HTTP asíncronas

import asyncio

import logging
logger = logging.getLogger(__name__)
logging.basicConfig(level=logging.INFO)



##### OPENAI #####
# Context:
openai_api_key = os.environ.get('OPENAI_API_KEY')


llm = OpenAI(
    model="gpt-4.1",
    api_key=openai_api_key,
)
embed_model = OpenAIEmbedding(
    model="text-embedding-ada-002",
    api_key=openai_api_key,
)

Settings.llm = llm
Settings.embed_model = embed_model




# Configuración de paths
DOCUMENTS_BASE_PATH = "./"
RETRIEVERS_JSON_PATH = Path("./retrievers.json")

# Cargar metadatos
def load_retrievers_metadata():
    try:
        with open(RETRIEVERS_JSON_PATH, 'r', encoding='utf-8') as f:
            return json.load(f)
    except Exception as e:
        logger.error(f"Error cargando retrievers.json: {str(e)}", exc_info=True)
        return {}

retrievers_metadata = load_retrievers_metadata()
SOURCES = {source: f"{source.lower()}/" for source in retrievers_metadata.keys()}

# Cargar índices
indices: Dict[str, VectorStoreIndex] = {}

for source, rel_path in SOURCES.items():
    full_path = os.path.join(DOCUMENTS_BASE_PATH, rel_path)
    if not os.path.exists(full_path):
        logger.warning(f"No se encontró la ruta para {source}")
        continue

    for root, dirs, files in os.walk(full_path):
        if "storage_nodes" in dirs:
            try:
                storage_path = os.path.join(root, "storage_nodes")
                storage_context = StorageContext.from_defaults(persist_dir=storage_path)
                index_name = os.path.basename(root)
                indices[index_name] = load_index_from_storage(storage_context) #, index_id="vector_index"
                logger.info(f"Índice cargado correctamente: {index_name}")
            except Exception as e:
                logger.error(f"Error cargando índice {index_name}: {str(e)}", exc_info=True)




            


port = int(os.getenv("PORT", 7860))
mcp = FastMCP("OnBase", port=port)


arxiv_tool = ArxivToolSpec(max_results=5).to_tool_list()[0]
arxiv_tool.return_direct = True

@mcp.tool()
async def search_arxiv(
    query: str,
    max_results: int = 5
) -> Dict[str, Any]:
    """
    Busca artículos académicos en ArXiv.
    
    Args:
        query: Términos de búsqueda (ej. "deep learning")
        max_results: Número máximo de resultados (1-10, default 5)
    
    Returns:
        Dict: Resultados de la búsqueda con metadatos de los papers
    """
    try:
        # Configurar máximo de resultados
        max_results = min(max(1, max_results), 10)
        arxiv_tool.metadata.max_results = max_results
        
        # Ejecutar búsqueda y obtener resultados
        tool_output = arxiv_tool(query=query)
        
        # Procesar documentos
        papers = []
        for doc in tool_output.raw_output:  # Acceder correctamente a los documentos
            content = doc.text_resource.text.split('\n')
            papers.append({
                'title': content[0].split(': ')[1] if ': ' in content[0] else content[0],
                'abstract': '\n'.join(content[1:]).strip(),
                'pdf_url': content[0].split(': ')[0].replace('http://', 'https://'),
                'arxiv_id': content[0].split(': ')[0].split('/')[-1].replace('v1', '')
            })
        
        return {
            'papers': papers,
            'count': len(papers),
            'query': query,
            'status': 'success'
        }
        
    except Exception as e:
        return {
            'papers': [],
            'count': 0,
            'query': query,
            'status': 'error',
            'error': str(e)
        }



@mcp.tool()
async def list_retrievers(source: str = None) -> dict:
    """
    Devuelve la lista de retrievers disponibles. 
    Si se especifica una source y existe, filtra por ella; si no existe, devuelve todas.
    
    Args:
        source (str, optional): Fuente para filtrar. Si no existe, se ignorará. Defaults to None.
    
    Returns:
        dict: {
            "retrievers": Lista de retrievers (filtrados o completos),
            "count": Número total,
            "status": "success"|"error",
            "source_requested": source,  # Muestra lo que se solicitó
            "source_used": "all"|source  # Muestra lo que realmente se usó
        }
    """
    try:
        available = []
        source_exists = source in retrievers_metadata if source else False
        
        for current_source, indexes in retrievers_metadata.items():
            # Solo filtrar si el source existe, sino mostrar todo
            if source_exists and current_source != source:
                continue
                
            for index_name, metadata in indexes.items():
                available.append({
                    "name": index_name,
                    "source": current_source,
                    "title": metadata.get("title", ""),
                    "description": metadata.get("description", "")
                })
        
        return {
            "retrievers": available,
            "count": len(available),
            "status": "success",
            "source_requested": source,
            "source_used": source if source_exists else "all"
        }
    except Exception as e:
        return {
            "retrievers": [],
            "count": 0,
            "status": "error",
            "error": str(e),
            "source_requested": source,
            "source_used": "none"
        }


@mcp.tool()
async def list_retrievers(ctx: Context, source: str = None) -> dict:
    """
    Devuelve la lista de retrievers disponibles, opcionalmente filtrados por source.
    
    Args:
        source (str, optional): Fuente para filtrar. Default None.
    
    Returns:
        dict: Lista de retrievers con metadatos.
    """
    # Obtenemos todos los retrievers del resource
    result = await ctx.resources.read("data://retrievers/list")
    
    # Si hay error en el resource, lo propagamos
    if result.get("status") == "error":
        return result
    
    # Filtramos por source si se especificó
    if source:
        filtered = [r for r in result["retrievers"] if r["source"] == source]
        return {
            "retrievers": filtered,
            "count": len(filtered),
            "status": "success",
            "source_requested": source,
            "source_used": source if filtered else "none"
        }
    
    return {
        **result,
        "source_requested": None,
        "source_used": "all"
    }




# Función de retrievers
@mcp.tool()
def retrieve_docs(
    query: str,
    retrievers: List[str],
    top_k: int = 3
) -> dict:
    """
    Realiza búsqueda semántica en documentos indexados.
    
    Parámetros:
        query (str): Texto de búsqueda (requerido)
        retrievers (List[str]): Nombres de retrievers a consultar (requerido)
        top_k (int): Número de resultados por retriever (opcional, default=3)
    """
    logger.info(f"Iniciando búsqueda para query: '{query}'")
    logger.debug(f"Parámetros - retrievers: {retrievers}, top_k: {top_k}")
    
    results = {}
    invalid = []
    
    for name in retrievers:
        if name not in indices:
            logger.warning(f"Retriever no encontrado: {name}")
            invalid.append(name)
            continue
            
        try:
            logger.info(f"Procesando retriever: {name}")
            
            # 1. Obtener el índice y realizar la búsqueda
            logger.debug(f"Creando retriever para {name} con top_k={top_k}")
            retriever = indices[name].as_retriever(similarity_top_k=top_k)
            nodes = retriever.retrieve(query)
            logger.info(f"Retrieved {len(nodes)} documentos de {name}")
            
            # 2. Buscar metadatos COMPLETOS
            metadata = {}
            source = "unknown"
            for src, indexes in retrievers_metadata.items():
                if name in indexes:
                    metadata = indexes[name]
                    source = src
                    break
            logger.debug(f"Metadatos encontrados para {name}: {metadata.keys()}")
            
            # 3. Construir respuesta
            results[name] = {
                "title": metadata.get("title", name),
                "documents": [
                    {
                        "content": node.get_content(),
                        "metadata": node.metadata,
                        "score": node.score
                    }
                    for node in nodes
                ],
                "description": metadata.get("description", ""),
                "source": source,
                "last_updated": metadata.get("last_updated", "")
            }
            logger.info(f"Retriever {name} procesado exitosamente")
            
        except Exception as e:
            logger.error(f"Error procesando retriever {name}: {str(e)}", exc_info=True)
            results[name] = {
                "error": str(e),
                "retriever": name
            }
    
    # Construir respuesta final
    response = {
        "query": query,
        "results": results,
        "top_k": top_k,
    }
    
    if invalid:
        logger.warning(f"Retrievers inválidos: {invalid}. Opciones válidas: {list(indices.keys())}")
        response["warnings"] = {
            "invalid_retrievers": invalid,
            "valid_options": list(indices.keys())
        }
    
    logger.info(f"Búsqueda completada. Total resultados: {len(results)}")
    return response

    

@mcp.tool()
async def search_tavily(
    query: str,
    days: int = 7,
    max_results: int = 1,
    include_answer: bool = False
) -> dict:
    """Perform a web search using the Tavily API.
    
    Args:
        query: Search query string (required)
        days: Restrict search to last N days (default: 7)
        max_results: Maximum results to return (default: 1)
        include_answer: Include a direct answer only when requested by the user (default: False)
        
    Returns:
        dict: Search results from Tavily
    """
    # Obtener la API key de las variables de entorno
    tavily_api_key = os.environ.get('TAVILY_API_KEY')
    if not tavily_api_key:
        raise ValueError("TAVILY_API_KEY environment variable not set")
    
    headers = {
        "Authorization": f"Bearer {tavily_api_key}",
        "Content-Type": "application/json"
    }
    
    payload = {
        "query": query,
        "search_depth": "basic",
        "max_results": max_results,
        "days": days if days else None,
        "include_answer": include_answer
    }
    
    try:
        async with aiohttp.ClientSession() as session:
            async with session.post(
                "https://api.tavily.com/search",
                headers=headers,
                json=payload
            ) as response:
                response.raise_for_status()
                result = await response.json()
                return result
                
    except Exception as e:
        return {
            "error": str(e),
            "status": "failed",
            "query": query
        }

if __name__ == "__main__":
    mcp.run("sse")