import os
import sys
from langchain_community.vectorstores import FAISS
from langchain_community.embeddings import HuggingFaceEmbeddings

sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
from rag.logger import get_logger  # pylint: disable=import-error
logger = get_logger(__name__)

def load_vectorstore():
    embeddings = HuggingFaceEmbeddings(
        model_name="sentence-transformers/all-MiniLM-L6-v2"
    )

    base_dir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
    vectorstore_path = os.path.join(base_dir, "data", "vectorstores")

    if not os.path.exists(vectorstore_path):
        raise FileNotFoundError(
            f"Vectorstore not found at: {vectorstore_path}\nRun ingest.py first."
        )

    vectorstore = FAISS.load_local(
        vectorstore_path,
        embeddings,
        allow_dangerous_deserialization=True 
    )
    logger.info('Vector store loaded')
    return vectorstore

def get_retriever(top_k: int = 5):
    vectorstore = load_vectorstore()

    # LangChain wrapper - preferred
    retriever = vectorstore.as_retriever(
        search_kwargs={"k": top_k}
    )
    logger.info('Retrieval Complete')
    return retriever


def search(query: str, top_k: int = 5):
    vectorstore = load_vectorstore()

    results = vectorstore.similarity_search(query, k=top_k)
    
    return results

if __name__ == "__main__":
    query = "GITHUB REPO DATA"
    results = search(query, top_k=5)

    print("\n=== SIMILARITY RESULTS ===")
    for r in results:
        print("\n--- CHUNK ---")
        print(r.page_content)
        print("Metadata:", r.metadata)