Pre-submission checklist | 提交前检查
Problem Statement | 问题陈述
In polardb.py, neo4j.py, and neo4j_community.py, the methods search_by_keywords_like, search_by_keywords_tfidf, search_by_fulltext, and search_by_embedding currently only return id and score in a fixed manner. If other fields are needed, you have to use the id to call get_node for another retrieval, which causes additional latency.
I hope to add a new return_fields: List[str] parameter to these methods, allowing you to pass in the desired field names, such as return_fields=["id", "memory", "status", "tags"], and include these fields + score in the returned dict.
This new field can have a default value of return_fields=["id"], so it won't affect previous calls.
Willingness to Implement | 实现意愿
Pre-submission checklist | 提交前检查
Problem Statement | 问题陈述
In polardb.py, neo4j.py, and neo4j_community.py, the methods search_by_keywords_like, search_by_keywords_tfidf, search_by_fulltext, and search_by_embedding currently only return id and score in a fixed manner. If other fields are needed, you have to use the id to call get_node for another retrieval, which causes additional latency.
I hope to add a new return_fields: List[str] parameter to these methods, allowing you to pass in the desired field names, such as return_fields=["id", "memory", "status", "tags"], and include these fields + score in the returned dict.
This new field can have a default value of return_fields=["id"], so it won't affect previous calls.
Willingness to Implement | 实现意愿