retrieval_service/datastore/providers/spanner_gsql.py [625:672]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            )

        # Check if result is None
        if result is None:
            return None, None

        # Convert query result to model instance using model_validate method
        amenities = [
            models.Amenity.model_validate(
                {key: value for key, value in zip(self.AMENITIES_COLUMNS, a)}
            )
            for a in result
        ]

        return amenities[0], None

    async def amenities_search(
        self, query_embedding: list[float], similarity_threshold: float, top_k: int
    ) -> tuple[list[Any], Optional[str]]:
        """
        Search for amenities based on similarity to a query embedding.

        Args:
            query_embedding (list[float]): The embedding representing the query.
            similarity_threshold (float): The minimum similarity threshold for results.
            top_k (int): The maximum number of results to return.

        Returns:
            list[models.Amenity]: A list of Amenity model instances matching the search criteria.
        """
        with self.__database.snapshot() as snapshot:
            # Spread SQL query for readability
            query = """
                SELECT name, description, location, terminal, category, hour
                FROM (
                    SELECT name, description, location, terminal, category, hour,
                       COSINE_DISTANCE(embedding, @query_embedding) AS similarity
                    FROM amenities
                ) AS sorted_amenities
                WHERE (1 - similarity) > @similarity_threshold
                ORDER BY similarity
                LIMIT @top_k
            """

            # Execute SQL query with parameters
            results = snapshot.execute_sql(
                sql=query,
                params={
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



retrieval_service/datastore/providers/spanner_postgres.py [626:673]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            )

        # Check if result is None
        if result is None:
            return None, None

        # Convert query result to model instance using model_validate method
        amenities = [
            models.Amenity.model_validate(
                {key: value for key, value in zip(self.AMENITIES_COLUMNS, a)}
            )
            for a in result
        ]

        return amenities[0], None

    async def amenities_search(
        self, query_embedding: list[float], similarity_threshold: float, top_k: int
    ) -> tuple[list[Any], Optional[str]]:
        """
        Search for amenities based on similarity to a query embedding.

        Args:
            query_embedding (list[float]): The embedding representing the query.
            similarity_threshold (float): The minimum similarity threshold for results.
            top_k (int): The maximum number of results to return.

        Returns:
            list[models.Amenity]: A list of Amenity model instances matching the search criteria.
        """
        with self.__database.snapshot() as snapshot:
            # Spread SQL query for readability
            query = """
                SELECT name, description, location, terminal, category, hour
                FROM (
                    SELECT name, description, location, terminal, category, hour,
                       spanner.cosine_distance(embedding, $1) AS similarity
                    FROM amenities
                ) AS sorted_amenities
                WHERE (1 - similarity) > $2
                ORDER BY similarity
                LIMIT $3
            """

            # Execute SQL query with parameters
            results = snapshot.execute_sql(
                sql=query,
                params={
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -