# @description: 
# @author: licanglong
# @date: 2025/12/19 16:32
from app.App import App
from app.utils.pathutils import getpath


class EmbeddingStoreApp(App):

    def run(self, *args, **kwargs):
        import uuid
        from typing import List

        from qdrant_client import QdrantClient
        from qdrant_client.models import VectorParams, Distance, PointStruct
        from sentence_transformers import SentenceTransformer

        if not kwargs or not kwargs.get("collection_name"):
            raise ValueError("miss collection_name value")

        if not kwargs or not kwargs.get("vector_data"):
            raise ValueError("miss vector_data value")

        collection_name: str = kwargs['collection_name']
        vector_size = 1792
        vector_data: dict = kwargs['vector_data']  # case_embed    rule_embed     merchants_embed     edges_embed
        client = QdrantClient(host="117.72.147.109", port=16333)
        model = SentenceTransformer(getpath(r"res\models\acge_text_embedding"))

        collections = client.get_collections().collections
        exists = any(c.name == collection_name for c in collections)

        if not exists:
            client.create_collection(
                collection_name=collection_name,
                vectors_config=VectorParams(
                    size=vector_size,
                    distance=Distance.COSINE,
                ),
            )

        points: List[PointStruct] = []

        for item in vector_data:
            vector = model.encode(item['embedding_text'])
            point_id = str(uuid.uuid4())
            points.append(
                PointStruct(
                    id=point_id,
                    vector=vector.tolist(),
                    payload=item,
                )
            )

        client.upsert(
            collection_name=collection_name,
            points=points,
        )