from dotenv import load_dotenv
import os

load_dotenv()

QDRANT_URL = os.getenv("QDRANT_URL")
QDRANT_API_KEY = os.getenv("QDRANT_API_KEY")
from qdrant_client import QdrantClient
from sentence_transformers import SentenceTransformer
from dotenv import load_dotenv
import os

load_dotenv()

QDRANT_URL = os.getenv("QDRANT_URL")
QDRANT_API_KEY = os.getenv("QDRANT_API_KEY")

COLLECTION_NAME = "student_materials"   # غيّري الاسم لو عندك اسم تاني

# Connect to Qdrant
client = QdrantClient(
    url=QDRANT_URL,
    api_key=QDRANT_API_KEY,
)

# Load embedding model
model = SentenceTransformer("intfloat/e5-large")


def search(query):
    # 1) Embed query
    query_vector = model.encode(query).tolist()

    # 2) Search Qdrant
    results = client.query_points(
        collection_name=COLLECTION_NAME,
        query=query_vector,
        limit=5
    )

    return results.points   # أهم سطر


# ==========================
# Example test
# ==========================
if __name__ == "__main__":
    res = search("What is machine learning?")
    for p in res:
        print("Payload:", p.payload)
        print("Score:", p.score)
        print("-" * 50)