fix: date parse error and fix memreader (#132)

* fix n4j cypher query

* feat: add llm extra body

* feat: update memory extraction prompt and result parser

* fix: evaluation locomo search

* ci: fix format and update test

* feat: update result json parser

* feat: recursively cluster nodes to max_cluster_size

* fix: fix template

* feat: keep default min-group-size 3

* feat: keep default min-group-size 3

* feat: update doc mem reader

* test: fix test

* fix: date parse error when python<3.11

* fix: update dependencies

* fix: TextualMemoryItem pydantic error

* fix: TextualMemoryItem pydantic error

* fix: TextualMemoryItem pydantic error

* fix: TextualMemoryItem pydantic error

* fix: prompt example

* fix: robuster memory parser

---------

Co-authored-by: chunyu li <[email protected]>

Author: Nyakult

poetry.lock

@@ -44,6 +44,7 @@ dependencies = [
     "sqlalchemy (>=2.0.41,<3.0.0)",  # SQL toolkit
     "scikit-learn (>=1.7.0,<2.0.0)",  # Machine learning
     "fastmcp (>=2.10.5,<3.0.0)",
+    "python-dateutil (>=2.9.0.post0,<3.0.0)",
 ]
 
 [project.urls]

pyproject.toml

@@ -58,9 +58,13 @@ def _process_chat_data(self, scene_data_info, info):
                 metadata=TreeNodeTextualMemoryMetadata(
                     user_id=info.get("user_id"),
                     session_id=info.get("session_id"),
-                    memory_type=memory_i_raw.get("memory_type", ""),
+                    memory_type=memory_i_raw.get("memory_type", "")
+                    .replace("长期记忆", "LongTermMemory")
+                    .replace("用户记忆", "UserMemory"),
                     status="activated",
-                    tags=memory_i_raw.get("tags", ""),
+                    tags=memory_i_raw.get("tags", [])
+                    if type(memory_i_raw.get("tags", [])) is list
+                    else [],
                     key=memory_i_raw.get("key", ""),
                     embedding=self.embedder.embed([memory_i_raw.get("value", "")])[0],
                     usage=[],
@@ -214,7 +218,7 @@ def _process_doc_data(self, scene_data_info, info):
                         session_id=info.get("session_id"),
                         memory_type="LongTermMemory",
                         status="activated",
-                        tags=chunk_res["tags"],
+                        tags=chunk_res["tags"] if type(chunk_res["tags"]) is list else [],
                         key=chunk_res["key"],
                         embedding=self.embedder.embed([chunk_res["value"]])[0],
                         usage=[],

src/memos/mem_reader/simple_struct.py

@@ -2,7 +2,7 @@
 import re
 
 from datetime import datetime
-
+from dateutil import parser
 from memos.embedders.base import BaseEmbedder
 from memos.graph_dbs.neo4j import Neo4jGraphDB
 from memos.llms.base import BaseLLM
@@ -133,8 +133,8 @@ def _hard_update(self, memory_a: TextualMemoryItem, memory_b: TextualMemoryItem)
         """
         Hard update: compare updated_at, keep the newer one, overwrite the older one's metadata.
         """
-        time_a = datetime.fromisoformat(memory_a.metadata.updated_at)
-        time_b = datetime.fromisoformat(memory_b.metadata.updated_at)
+        time_a = parser.isoparse(memory_a.metadata.updated_at)
+        time_b = parser.isoparse(memory_b.metadata.updated_at)
 
         newer_mem = memory_a if time_a >= time_b else memory_b
         older_mem = memory_b if time_a >= time_b else memory_a

src/memos/memories/textual/tree_text_memory/organize/conflict.py

@@ -150,7 +150,7 @@
   "summary": "Tom is currently focused on managing a new project with a tight schedule. After a team meeting on June 25, 2025, he realized the original deadline of December 15 might not be feasible due to backend delays. Concerned about insufficient testing time, he welcomed Jerry’s suggestion of proposing an extension. Tom plans to raise the idea of shifting the deadline to January 5, 2026 in the next morning’s meeting. His actions reflect both stress about timelines and a proactive, team-oriented problem-solving approach."
 }
 
-Another Example in Chinese (注意: 你的输出必须和输入的user语言一致)：
+Another Example in Chinese (注意: 当user的语言为中文时，你就需要也输出中文)：
 {
   "memory list": [
     {