Investigate get_chat_history limit

README.md

@@ -139,7 +139,7 @@ claude mcp add wechat -- python C:\Users\你的用户名\wechat-decrypt\mcp_serv
 
 前置条件：需要先运行 `python main.py` 或 `python find_all_keys.py` 完成密钥提取。
 
-说明：`get_chat_history` 和 `search_messages` 的 `limit` 最大为 `500`。
+说明：`search_messages` 的 `limit` 最大为 `500`；`get_chat_history` 支持更大的 `limit`，但消息很多时仍建议配合 `offset` 分页读取。
 
 **[查看使用案例 →](USAGE.md)**
 

USAGE.md

@@ -191,7 +191,7 @@ Claude 可以获取大量消息后自动分析活跃度、话题分布、关键
 > 帮我分析一下██群最近一周的情况
 ```
 
-Claude 会调用 `get_chat_history(chat_name="██群", limit=500)` 获取消息，然后输出：
+Claude 会调用 `get_chat_history(chat_name="██群", limit=500)` 获取消息，然后输出。消息很多时，也可以把 `limit` 设得更大，或配合 `offset` 分页读取：
 
 ```
 ## ██群最近一周分析

mcp_server.py

@@ -224,6 +224,7 @@ _self_username = None
 _XML_UNSAFE_RE = re.compile(r'<!DOCTYPE|<!ENTITY', re.IGNORECASE)
 _XML_PARSE_MAX_LEN = 20000
 _QUERY_LIMIT_MAX = 500
+_HISTORY_QUERY_BATCH_SIZE = 500
 
 
 def _load_contacts_from(db_path):
@@ -635,11 +636,11 @@ def _find_msg_tables_for_user(username):
     return matches
 
 
-def _validate_pagination(limit, offset=0):
+def _validate_pagination(limit, offset=0, limit_max=_QUERY_LIMIT_MAX):
     if limit <= 0:
         raise ValueError("limit 必须大于 0")
-    if limit > _QUERY_LIMIT_MAX:
-        raise ValueError(f"limit 不能大于 {_QUERY_LIMIT_MAX}")
+    if limit_max is not None and limit > limit_max:
+        raise ValueError(f"limit 不能大于 {limit_max}")
     if offset < 0:
         raise ValueError("offset 不能小于 0")
 
@@ -841,8 +842,6 @@ def _build_history_line(row, ctx, names, id_to_username):
     sender, text = _format_message_text(
         local_id, local_type, content, ctx['is_group'], ctx['username'], ctx['display_name'], names
     )
-    if text and len(text) > 500:
-        text = text[:500] + '...'
 
     sender_label = _resolve_sender_label(
         real_sender_id, sender, ctx['is_group'], ctx['username'], ctx['display_name'], names, id_to_username
@@ -880,6 +879,10 @@ def _message_query_batch_size(candidate_limit):
     return candidate_limit
 
 
+def _history_query_batch_size(candidate_limit):
+    return min(candidate_limit, _HISTORY_QUERY_BATCH_SIZE)
+
+
 def _page_ranked_entries(entries, limit, offset):
     ordered = sorted(entries, key=lambda item: item[0], reverse=True)
     paged = ordered[offset:offset + limit]
@@ -891,22 +894,40 @@ def _collect_chat_history_lines(ctx, names, start_ts=None, end_ts=None, limit=20
     collected = []
     failures = []
     candidate_limit = _candidate_page_size(limit, offset)
+    batch_size = _history_query_batch_size(candidate_limit)
 
     for table_ctx in _iter_table_contexts(ctx):
         try:
             with closing(sqlite3.connect(table_ctx['db_path'])) as conn:
                 id_to_username = _load_name2id_maps(conn)
+                fetch_offset = 0
+                collected_before_table = len(collected)
                 # 当前页上的消息一定落在各分表最近的 offset+limit 条记录内。
+                while len(collected) - collected_before_table < candidate_limit:
                     rows = _query_messages(
                         conn,
                         table_ctx['table_name'],
                         start_ts=start_ts,
                         end_ts=end_ts,
-                    limit=candidate_limit,
-                    offset=0,
+                        limit=batch_size,
+                        offset=fetch_offset,
                     )
+                    if not rows:
+                        break
+                    fetch_offset += len(rows)
+
                     for row in rows:
+                        try:
                             collected.append(_build_history_line(row, table_ctx, names, id_to_username))
+                        except Exception as e:
+                            failures.append(
+                                f"{table_ctx['display_name']} local_id={row[0]} create_time={row[2]}: {e}"
+                            )
+                        if len(collected) - collected_before_table >= candidate_limit:
+                            break
+
+                    if len(rows) < batch_size:
+                        break
         except Exception as e:
             failures.append(f"{table_ctx['db_path']}: {e}")
 
@@ -1221,13 +1242,13 @@ def get_chat_history(chat_name: str, limit: int = 50, offset: int = 0, start_tim
 
     Args:
         chat_name: 聊天对象的名字、备注名或wxid，自动模糊匹配
-        limit: 返回的消息数量，默认50，最大500
+        limit: 返回的消息数量，默认50；支持较大的值，建议配合 offset 分页使用
         offset: 分页偏移量，默认0
         start_time: 起始时间，支持 YYYY-MM-DD / YYYY-MM-DD HH:MM / YYYY-MM-DD HH:MM:SS
         end_time: 结束时间，支持 YYYY-MM-DD / YYYY-MM-DD HH:MM / YYYY-MM-DD HH:MM:SS
     """
     try:
-        _validate_pagination(limit, offset)
+        _validate_pagination(limit, offset, limit_max=None)
         start_ts, end_ts = _parse_time_range(start_time, end_time)
     except ValueError as e:
         return f"错误: {e}"

tests/test_mcp_server_search.py

@@ -72,6 +72,10 @@ class SearchMessagesTests(unittest.TestCase):
         with self.assertRaisesRegex(ValueError, "limit 不能大于 500"):
             mcp_server._validate_pagination(501, 0)
 
+    def test_validate_pagination_allows_large_limit_when_limit_is_unbounded(self):
+        # get_chat_history 允许更大的 limit，只校验正数和 offset。
+        mcp_server._validate_pagination(999999, 0, limit_max=None)
+
     def test_page_search_entries_returns_chronological_results_with_offset(self):
         # 结果应先按最新时间分页，再把当前页恢复成时间正序输出。
         entries = [(1, "a"), (5, "e"), (3, "c"), (4, "d"), (2, "b")]
@@ -377,6 +381,43 @@ class SearchMessagesTests(unittest.TestCase):
         self.assertIn("new message", result)
         self.assertNotIn("old message", result)
 
+    def test_get_chat_history_large_limit_reads_all_rows_across_shards(self):
+        # 大 limit 下，跨分片历史查询不能只返回较旧分片里的少量消息。
+        older_messages = [
+            (index, 1000 + index, f"old shard message {index}")
+            for index in range(1, 18)
+        ]
+        newer_messages = [
+            (index, 2000 + index, f"new shard message {index}")
+            for index in range(1, 296)
+        ]
+        older_db = self.create_db("history_cross_shard_older.db", {"alice": older_messages})
+        newer_db = self.create_db("history_cross_shard_newer.db", {"alice": newer_messages})
+        ctx = {
+            "query": "Alice",
+            "username": "alice",
+            "display_name": "Alice",
+            "db_path": newer_db,
+            "table_name": _msg_table_name("alice"),
+            "message_tables": [
+                {"db_path": newer_db, "table_name": _msg_table_name("alice")},
+                {"db_path": older_db, "table_name": _msg_table_name("alice")},
+            ],
+            "is_group": False,
+        }
+
+        with patch.object(mcp_server, "get_contact_names", return_value={"alice": "Alice"}), patch.object(
+            mcp_server, "_resolve_chat_context", return_value=ctx
+        ):
+            result = mcp_server.get_chat_history("Alice", limit=500, offset=0)
+
+        self.assertIn("Alice 的消息记录（返回 312 条，offset=0, limit=500）", result)
+        self.assertIn("new shard message 295", result)
+        self.assertIn("old shard message 17", result)
+
+        body = result.split(":\n\n", 1)[1]
+        self.assertEqual(len(body.splitlines()), 312)
+
     def test_get_chat_history_uses_bounded_sql_pagination(self):
         # 历史查询应把 offset+limit 下推到 SQL，避免把整张消息表读出来后再切片。
         db_path = self.create_db(
@@ -419,6 +460,36 @@ class SearchMessagesTests(unittest.TestCase):
         self.assertNotIn("oldest", result)
         self.assertEqual(calls, [(_msg_table_name("alice"), 3, 0)])
 
+    def test_get_chat_history_allows_large_limit_values(self):
+        # 历史查询不应再把大 limit 直接拒绝掉。
+        db_path = self.create_db(
+            "history_large_limit.db",
+            {
+                "alice": [
+                    (1, 200, "message 1"),
+                    (2, 100, "message 2"),
+                ]
+            },
+        )
+        ctx = {
+            "query": "Alice",
+            "username": "alice",
+            "display_name": "Alice",
+            "db_path": db_path,
+            "table_name": _msg_table_name("alice"),
+            "message_tables": [{"db_path": db_path, "table_name": _msg_table_name("alice")}],
+            "is_group": False,
+        }
+
+        with patch.object(mcp_server, "get_contact_names", return_value={"alice": "Alice"}), patch.object(
+            mcp_server, "_resolve_chat_context", return_value=ctx
+        ):
+            result = mcp_server.get_chat_history("Alice", limit=999999, offset=0)
+
+        self.assertNotIn("错误:", result)
+        self.assertIn("message 1", result)
+        self.assertIn("message 2", result)
+
     def test_get_chat_history_keeps_partial_results_when_formatting_fails(self):
         # 单条坏消息不应让整个历史查询失败，已有结果仍应返回并附带失败说明。
         db_path = self.create_db(
@@ -452,6 +523,31 @@ class SearchMessagesTests(unittest.TestCase):
         self.assertIn("查询失败:", result)
         self.assertIn("bad row", result)
 
+    def test_get_chat_history_does_not_truncate_long_messages(self):
+        # 历史记录应返回完整消息内容，而不是固定截断到 500 字符。
+        long_message = "x" * 600
+        db_path = self.create_db(
+            "history_long_message.db",
+            {"alice": [(1, 200, long_message)]},
+        )
+        ctx = {
+            "query": "Alice",
+            "username": "alice",
+            "display_name": "Alice",
+            "db_path": db_path,
+            "table_name": _msg_table_name("alice"),
+            "message_tables": [{"db_path": db_path, "table_name": _msg_table_name("alice")}],
+            "is_group": False,
+        }
+
+        with patch.object(mcp_server, "get_contact_names", return_value={"alice": "Alice"}), patch.object(
+            mcp_server, "_resolve_chat_context", return_value=ctx
+        ):
+            result = mcp_server.get_chat_history("Alice", limit=1, offset=0)
+
+        self.assertIn(long_message, result)
+        self.assertNotIn(("x" * 500) + "...", result)
+
     def test_search_messages_single_chat_merges_sharded_message_tables(self):
         # 单聊搜索也要跨分片合并，否则最近消息可能查不到。
         older_db = self.create_db("search_older.db", {"alice": [(1, 100, "foo old")]})