""" Below is an implementation of a basic LRUcache that utilizes the built in OrderedDict data structure. """ from collections import OrderedDict from mediator import SupabaseMediator import uuid from typing import List, Dict from langchain.schema import BaseMessage, Document from pydantic import BaseModel # import sentry_sdk class Conversation: "Wrapper Class for storing contexts between channels. Using an object to pass by reference avoid additional cache hits" # @sentry_sdk.trace def __init__( self, mediator: SupabaseMediator, user_id: str, session_id: str = str(uuid.uuid4()), metadata: Dict = {}, ): self.mediator: SupabaseMediator = mediator self.user_id = user_id self.session_id = session_id self.metadata: Dict = metadata # @sentry_sdk.trace def add_message( self, message_type: str, message: BaseMessage, ) -> None: self.mediator.add_message(self.session_id, message_type, message.content) # @sentry_sdk.trace def get_messages(self, message_type: str) -> List[BaseMessage]: return self.mediator.get_messages(self.session_id, message_type) # @sentry_sdk.trace def delete(self) -> None: self.mediator.delete_session(self.session_id) # @sentry_sdk.trace # def restart(self) -> None: # self.delete() # representation = self.mediator.add_session( # user_id=self.user_id, location_id=self.location_id # ) # self.session_id: str = representation["id"] # self.metadata = representation["metadata"] # vector DB fn # @sentry_sdk.trace def add_texts(self, texts: List[str]) -> None: metadatas = [ {"session_id": self.session_id, "user_id": self.user_id} for _ in range(len(texts)) ] self.mediator.vector_table.add_texts(texts, metadatas) # vector DB fn # @sentry_sdk.trace def similarity_search(self, query: str, match_count: int = 5) -> List[Document]: return self.mediator.vector_table.similarity_search( query=query, k=match_count, filter={"user_id": self.user_id} )