message search

2025-04-19 02:59:59 +08:00
19 changed files with 910 additions and 519 deletions
--- a/internal/chatlog/database/service.go
+++ b/internal/chatlog/database/service.go
@@ -40,8 +40,8 @@ func (s *Service) GetDB() *wechatdb.DB {
 	return s.db
 }

-func (s *Service) GetMessages(start, end time.Time, talker string, limit, offset int) ([]*model.Message, error) {
-	return s.db.GetMessages(start, end, talker, limit, offset)
+func (s *Service) GetMessages(start, end time.Time, talker string, sender string, keyword string, limit, offset int) ([]*model.Message, error) {
+	return s.db.GetMessages(start, end, talker, sender, keyword, limit, offset)
 }

 func (s *Service) GetContacts(key string, limit, offset int) (*wechatdb.GetContactsResp, error) {
--- a/internal/chatlog/http/route.go
+++ b/internal/chatlog/http/route.go
@@ -75,11 +75,13 @@ func (s *Service) NoRoute(c *gin.Context) {
 func (s *Service) GetChatlog(c *gin.Context) {

 	q := struct {
-		Time   string `form:"time"`
-		Talker string `form:"talker"`
-		Limit  int    `form:"limit"`
-		Offset int    `form:"offset"`
-		Format string `form:"format"`
+		Time    string `form:"time"`
+		Talker  string `form:"talker"`
+		Sender  string `form:"sender"`
+		Keyword string `form:"keyword"`
+		Limit   int    `form:"limit"`
+		Offset  int    `form:"offset"`
+		Format  string `form:"format"`
 	}{}

 	if err := c.BindQuery(&q); err != nil {
@@ -100,7 +102,7 @@ func (s *Service) GetChatlog(c *gin.Context) {
 		q.Offset = 0
 	}

-	messages, err := s.db.GetMessages(start, end, q.Talker, q.Limit, q.Offset)
+	messages, err := s.db.GetMessages(start, end, q.Talker, q.Sender, q.Keyword, q.Limit, q.Offset)
 	if err != nil {
 		errors.Err(c, err)
 		return
@@ -119,7 +121,7 @@ func (s *Service) GetChatlog(c *gin.Context) {
 		c.Writer.Flush()

 		for _, m := range messages {
-			c.Writer.WriteString(m.PlainText(len(q.Talker) == 0, c.Request.Host))
+			c.Writer.WriteString(m.PlainText(strings.Contains(q.Talker, ","), util.PerfectTimeFormat(start, end), c.Request.Host))
 			c.Writer.WriteString("\n")
 			c.Writer.Flush()
 		}
@@ -129,10 +131,10 @@ func (s *Service) GetChatlog(c *gin.Context) {
 func (s *Service) GetContacts(c *gin.Context) {

 	q := struct {
-		Key    string `form:"key"`
-		Limit  int    `form:"limit"`
-		Offset int    `form:"offset"`
-		Format string `form:"format"`
+		Keyword string `form:"keyword"`
+		Limit   int    `form:"limit"`
+		Offset  int    `form:"offset"`
+		Format  string `form:"format"`
 	}{}

 	if err := c.BindQuery(&q); err != nil {
@@ -140,7 +142,7 @@ func (s *Service) GetContacts(c *gin.Context) {
 		return
 	}

-	list, err := s.db.GetContacts(q.Key, q.Limit, q.Offset)
+	list, err := s.db.GetContacts(q.Keyword, q.Limit, q.Offset)
 	if err != nil {
 		errors.Err(c, err)
 		return
@@ -174,10 +176,10 @@ func (s *Service) GetContacts(c *gin.Context) {
 func (s *Service) GetChatRooms(c *gin.Context) {

 	q := struct {
-		Key    string `form:"key"`
-		Limit  int    `form:"limit"`
-		Offset int    `form:"offset"`
-		Format string `form:"format"`
+		Keyword string `form:"keyword"`
+		Limit   int    `form:"limit"`
+		Offset  int    `form:"offset"`
+		Format  string `form:"format"`
 	}{}

 	if err := c.BindQuery(&q); err != nil {
@@ -185,7 +187,7 @@ func (s *Service) GetChatRooms(c *gin.Context) {
 		return
 	}

-	list, err := s.db.GetChatRooms(q.Key, q.Limit, q.Offset)
+	list, err := s.db.GetChatRooms(q.Keyword, q.Limit, q.Offset)
 	if err != nil {
 		errors.Err(c, err)
 		return
@@ -218,10 +220,10 @@ func (s *Service) GetChatRooms(c *gin.Context) {
 func (s *Service) GetSessions(c *gin.Context) {

 	q := struct {
-		Key    string `form:"key"`
-		Limit  int    `form:"limit"`
-		Offset int    `form:"offset"`
-		Format string `form:"format"`
+		Keyword string `form:"keyword"`
+		Limit   int    `form:"limit"`
+		Offset  int    `form:"offset"`
+		Format  string `form:"format"`
 	}{}

 	if err := c.BindQuery(&q); err != nil {
@@ -229,7 +231,7 @@ func (s *Service) GetSessions(c *gin.Context) {
 		return
 	}

-	sessions, err := s.db.GetSessions(q.Key, q.Limit, q.Offset)
+	sessions, err := s.db.GetSessions(q.Keyword, q.Limit, q.Offset)
 	if err != nil {
 		errors.Err(c, err)
 		return
--- a/internal/chatlog/http/static/index.htm
+++ b/internal/chatlog/http/static/index.htm
@@ -377,12 +377,12 @@
              </p>
            </div>
            <div class="form-group">
-              <label for="chatroom-query"
+              <label for="chatroom-keyword"
                >搜索群聊：<span class="optional-param">可选</span></label
              >
              <input
                type="text"
-                id="chatroom-query"
+                id="chatroom-keyword"
                placeholder="输入关键词搜索群聊"
              />
            </div>
@@ -408,12 +408,12 @@
              </p>
            </div>
            <div class="form-group">
-              <label for="contact-query"
+              <label for="contact-keyword"
                >搜索联系人：<span class="optional-param">可选</span></label
              >
              <input
                type="text"
-                id="contact-query"
+                id="contact-keyword"
                placeholder="输入关键词搜索联系人"
              />
            </div>
@@ -459,6 +459,26 @@
                placeholder="wxid、群ID、备注名或昵称"
              />
            </div>
+            <div class="form-group">
+              <label for="sender"
+                >发送者：<span class="optional-param">可选</span></label
+              >
+              <input
+                type="text"
+                id="sender"
+                placeholder="指定消息发送者"
+              />
+            </div>
+            <div class="form-group">
+              <label for="keyword"
+                >关键词：<span class="optional-param">可选</span></label
+              >
+              <input
+                type="text"
+                id="keyword"
+                placeholder="搜索消息内容中的关键词"
+              />
+            </div>
            <div class="form-group">
              <label for="limit"
                >返回数量：<span class="optional-param">可选</span></label
@@ -603,6 +623,8 @@
                url += "chatlog";
                const time = document.getElementById("time").value;
                const talker = document.getElementById("talker").value;
+                const sender = document.getElementById("sender").value;
+                const keyword = document.getElementById("keyword").value;
                const limit = document.getElementById("limit").value;
                const offset = document.getElementById("offset").value;
                const format = document.getElementById("format").value;
@@ -617,6 +639,8 @@

                if (time) params.append("time", time);
                if (talker) params.append("talker", talker);
+                if (sender) params.append("sender", sender);
+                if (keyword) params.append("keyword", keyword);
                if (limit) params.append("limit", limit);
                if (offset) params.append("offset", offset);
                if (format) params.append("format", format);
@@ -624,23 +648,23 @@

              case "contact":
                url += "contact";
-                const contactQuery =
-                  document.getElementById("contact-query").value;
+                const contactKeyword =
+                  document.getElementById("contact-keyword").value;
                const contactFormat =
                  document.getElementById("contact-format").value;

-                if (contactQuery) params.append("query", contactQuery);
+                if (contactKeyword) params.append("keyword", contactKeyword);
                if (contactFormat) params.append("format", contactFormat);
                break;

              case "chatroom":
                url += "chatroom";
-                const chatroomQuery =
-                  document.getElementById("chatroom-query").value;
+                const chatroomKeyword =
+                  document.getElementById("chatroom-keyword").value;
                const chatroomFormat =
                  document.getElementById("chatroom-format").value;

-                if (chatroomQuery) params.append("query", chatroomQuery);
+                if (chatroomKeyword) params.append("keyword", chatroomKeyword);
                if (chatroomFormat) params.append("format", chatroomFormat);
                break;

@@ -730,4 +754,4 @@
      }
    </script>
  </body>
-</html>
+</html>
--- a/internal/chatlog/mcp/const.go
+++ b/internal/chatlog/mcp/const.go
@@ -21,12 +21,12 @@ var (
 		InputSchema: mcp.ToolSchema{
 			Type: "object",
 			Properties: mcp.M{
-				"query": mcp.M{
+				"keyword": mcp.M{
 					"type":        "string",
 					"description": "联系人的搜索关键词，可以是姓名、备注名或ID。",
 				},
 			},
-			Required: []string{"query"},
+			Required: []string{"keyword"},
 		},
 	}

@@ -36,12 +36,12 @@ var (
 		InputSchema: mcp.ToolSchema{
 			Type: "object",
 			Properties: mcp.M{
-				"query": mcp.M{
+				"keyword": mcp.M{
 					"type":        "string",
 					"description": "群聊的搜索关键词，可以是群名称、群ID或相关描述",
 				},
 			},
-			Required: []string{"query"},
+			Required: []string{"keyword"},
 		},
 	}

@@ -55,24 +55,136 @@ var (
 	}

 	ToolChatLog = mcp.Tool{
-		Name:        "chatlog",
-		Description: "查询特定时间或时间段内与特定联系人或群组的聊天记录。当用户需要回顾过去的对话内容、查找特定信息或想了解与某人/某群的历史交流时使用此工具。",
+		Name: "chatlog",
+		Description: `检索历史聊天记录，可根据时间、对话方、发送者和关键词等条件进行精确查询。当用户需要查找特定信息或想了解与某人/某群的历史交流时使用此工具。
+
+【强制多步查询流程!】
+当查询特定话题或特定发送者发言时，必须严格按照以下流程使用，任何偏离都会导致错误的结果：
+
+步骤1: 初步定位相关消息
+- 使用keyword参数查找特定话题
+- 使用sender参数查找特定发送者的消息
+- 使用较宽时间范围初步查询
+
+步骤2: 【必须执行】针对每个关键结果点分别获取上下文
+- 必须对步骤1返回的每个时间点T1, T2, T3...分别执行独立查询（时间范围接近的消息可以合并为一个查询）
+- 每次独立查询必须移除keyword参数
+- 每次独立查询必须移除sender参数
+- 每次独立查询使用"Tn前后15-30分钟"的窄范围
+- 每次独立查询仅保留talker参数
+
+步骤3: 【必须执行】综合分析所有上下文
+- 必须等待所有步骤2的查询结果返回后再进行分析
+- 必须综合考虑所有上下文信息后再回答用户
+
+【严格执行规则！】
+- 禁止仅凭步骤1的结果直接回答用户
+- 禁止在步骤2使用过大的时间范围一次性查询所有上下文
+- 禁止跳过步骤2或步骤3
+- 必须对每个关键结果点分别执行独立的上下文查询
+
+【执行示例】
+正确流程示例:
+1. 步骤1: chatlog(time="2023-04-01~2023-04-30", talker="工作群", keyword="项目进度")
+   返回结果: 4月5日、4月12日、4月20日有相关消息
+2. 步骤2:
+   - 查询1: chatlog(time="2023-04-05/09:30~2023-04-05/10:30", talker="工作群") // 注意没有keyword
+   - 查询2: chatlog(time="2023-04-12/14:00~2023-04-12/15:00", talker="工作群") // 注意没有keyword
+   - 查询3: chatlog(time="2023-04-20/16:00~2023-04-20/17:00", talker="工作群") // 注意没有keyword
+3. 步骤3: 综合分析所有上下文后回答用户
+
+错误流程示例:
+- 仅执行步骤1后直接回答
+- 步骤2使用time="2023-04-01~2023-04-30"一次性查询
+- 步骤2仍然保留keyword或sender参数
+
+【自我检查】回答用户前必须自问:
+- 我是否对每个关键时间点都执行了独立的上下文查询?
+- 我是否在上下文查询中移除了keyword和sender参数?
+- 我是否分析了所有上下文后再回答?
+- 如果上述任一问题答案为"否"，则必须纠正流程
+
+返回格式："昵称(ID) 时间\n消息内容\n昵称(ID) 时间\n消息内容"
+当查询多个Talker时，返回格式为："昵称(ID)\n[TalkerName(Talker)] 时间\n消息内容"
+
+重要提示：
+1. 当用户询问特定时间段内的聊天记录时，必须使用正确的时间格式，特别是包含小时和分钟的查询
+2. 对于"今天下午4点到5点聊了啥"这类查询，正确的时间参数格式应为"2023-04-18/16:00~2023-04-18/17:00"
+3. 当用户询问具体群聊中某人的聊天记录时，使用"sender"参数
+4. 当用户询问包含特定关键词的聊天记录时，使用"keyword"参数`,
 		InputSchema: mcp.ToolSchema{
 			Type: "object",
 			Properties: mcp.M{
 				"time": mcp.M{
-					"type":        "string",
-					"description": "查询的时间点或时间段。可以是具体时间，例如 YYYY-MM-DD，也可以是时间段，例如 YYYY-MM-DD~YYYY-MM-DD，时间段之间用\"~\"分隔。",
+					"type": "string",
+					"description": `指定查询的时间点或时间范围，格式必须严格遵循以下规则：
+
+【单一时间点格式】
+- 精确到日："2023-04-18"或"20230418"
+- 精确到分钟（必须包含斜杠和冒号）："2023-04-18/14:30"或"20230418/14:30"（表示2023年4月18日14点30分）
+
+【时间范围格式】（使用"~"分隔起止时间）
+- 日期范围："2023-04-01~2023-04-18"
+- 同一天的时间段："2023-04-18/14:30~2023-04-18/15:45"
+  * 表示2023年4月18日14点30分到15点45分之间
+
+【重要提示】包含小时分钟的格式必须使用斜杠和冒号："/"和":"
+正确示例："2023-04-18/16:30"（4月18日下午4点30分）
+错误示例："2023-04-18 16:30"、"2023-04-18T16:30"
+
+【其他支持的格式】
+- 年份："2023"
+- 月份："2023-04"或"202304"`,
 				},
 				"talker": mcp.M{
-					"type":        "string",
-					"description": "交谈对象，可以是联系人或群聊。支持使用ID、昵称、备注名等进行查询。",
+					"type": "string",
+					"description": `指定对话方（联系人或群组）
+- 可使用ID、昵称或备注名
+- 多个对话方用","分隔，如："张三,李四,工作群"
+- 【重要】这是多步查询中唯一应保留的参数`,
+				},
+				"sender": mcp.M{
+					"type": "string",
+					"description": `指定群聊中的发送者
+- 仅在查询群聊记录时有效
+- 多个发送者用","分隔，如："张三,李四"
+- 可使用ID、昵称或备注名
+【重要】查询特定发送者的消息时：
+  1. 第一步：使用sender参数初步定位多个相关消息时间点
+  2. 后续步骤：必须移除sender参数，分别查询每个时间点前后的完整对话
+  3. 错误示例：对所有找到的消息一次性查询大范围上下文
+  4. 正确示例：对每个时间点T分别执行查询"T前后15-30分钟"（不带sender）`,
+				},
+				"keyword": mcp.M{
+					"type": "string",
+					"description": `搜索内容中的关键词
+- 支持正则表达式匹配
+- 【重要】查询特定话题时：
+  1. 第一步：使用keyword参数初步定位多个相关消息时间点
+  2. 后续步骤：必须移除keyword参数，分别查询每个时间点前后的完整对话
+  3. 错误示例：对所有找到的关键词消息一次性查询大范围上下文
+  4. 正确示例：对每个时间点T分别执行查询"T前后15-30分钟"（不带keyword）`,
 				},
 			},
 			Required: []string{"time", "talker"},
 		},
 	}

+	ToolCurrentTime = mcp.Tool{
+		Name: "current_time",
+		Description: `获取当前系统时间，返回RFC3339格式的时间字符串（包含用户本地时区信息）。
+使用场景：
+- 当用户询问"总结今日聊天记录"、"本周都聊了啥"等当前时间问题
+- 当用户提及"昨天"、"上周"、"本月"等相对时间概念，需要确定基准时间点
+- 需要执行依赖当前时间的计算（如"上个月5号我们有开会吗"）
+返回示例：2025-04-18T21:29:00+08:00
+注意：此工具不需要任何输入参数，直接调用即可获取当前时间。`,
+		InputSchema: mcp.ToolSchema{
+			Type:       "object",
+			Properties: mcp.M{},
+		},
+	}
+
 	ResourceRecentChat = mcp.Resource{
 		Name:        "最近会话",
 		URI:         "session://recent",
--- a/internal/chatlog/mcp/service.go
+++ b/internal/chatlog/mcp/service.go
@@ -7,6 +7,7 @@ import (
 	"fmt"
 	"net/url"
 	"strings"
+	"time"

 	"github.com/sjzar/chatlog/internal/chatlog/ctx"
 	"github.com/sjzar/chatlog/internal/chatlog/database"
@@ -83,6 +84,7 @@ func (s *Service) processMCP(session *mcp.Session, req *mcp.Request) {
 			ToolChatRoom,
 			ToolRecentChat,
 			ToolChatLog,
+			ToolCurrentTime,
 		}})
 	case mcp.MethodToolsCall:
 		err = s.toolsCall(session, req)
@@ -130,13 +132,13 @@ func (s *Service) toolsCall(session *mcp.Session, req *mcp.Request) error {
 	buf := &bytes.Buffer{}
 	switch callReq.Name {
 	case "query_contact":
-		query := ""
-		if v, ok := callReq.Arguments["query"]; ok {
-			query = v.(string)
+		keyword := ""
+		if v, ok := callReq.Arguments["keyword"]; ok {
+			keyword = v.(string)
 		}
 		limit := util.MustAnyToInt(callReq.Arguments["limit"])
 		offset := util.MustAnyToInt(callReq.Arguments["offset"])
-		list, err := s.db.GetContacts(query, limit, offset)
+		list, err := s.db.GetContacts(keyword, limit, offset)
 		if err != nil {
 			return fmt.Errorf("无法获取联系人列表: %v", err)
 		}
@@ -145,13 +147,13 @@ func (s *Service) toolsCall(session *mcp.Session, req *mcp.Request) error {
 			buf.WriteString(fmt.Sprintf("%s,%s,%s,%s\n", contact.UserName, contact.Alias, contact.Remark, contact.NickName))
 		}
 	case "query_chat_room":
-		query := ""
-		if v, ok := callReq.Arguments["query"]; ok {
-			query = v.(string)
+		keyword := ""
+		if v, ok := callReq.Arguments["keyword"]; ok {
+			keyword = v.(string)
 		}
 		limit := util.MustAnyToInt(callReq.Arguments["limit"])
 		offset := util.MustAnyToInt(callReq.Arguments["offset"])
-		list, err := s.db.GetChatRooms(query, limit, offset)
+		list, err := s.db.GetChatRooms(keyword, limit, offset)
 		if err != nil {
 			return fmt.Errorf("无法获取群聊列表: %v", err)
 		}
@@ -160,13 +162,13 @@ func (s *Service) toolsCall(session *mcp.Session, req *mcp.Request) error {
 			buf.WriteString(fmt.Sprintf("%s,%s,%s,%s,%d\n", chatRoom.Name, chatRoom.Remark, chatRoom.NickName, chatRoom.Owner, len(chatRoom.Users)))
 		}
 	case "query_recent_chat":
-		query := ""
-		if v, ok := callReq.Arguments["query"]; ok {
-			query = v.(string)
+		keyword := ""
+		if v, ok := callReq.Arguments["keyword"]; ok {
+			keyword = v.(string)
 		}
 		limit := util.MustAnyToInt(callReq.Arguments["limit"])
 		offset := util.MustAnyToInt(callReq.Arguments["offset"])
-		data, err := s.db.GetSessions(query, limit, offset)
+		data, err := s.db.GetSessions(keyword, limit, offset)
 		if err != nil {
 			return fmt.Errorf("无法获取会话列表: %v", err)
 		}
@@ -190,16 +192,29 @@ func (s *Service) toolsCall(session *mcp.Session, req *mcp.Request) error {
 		if v, ok := callReq.Arguments["talker"]; ok {
 			talker = v.(string)
 		}
+		sender := ""
+		if v, ok := callReq.Arguments["sender"]; ok {
+			sender = v.(string)
+		}
+		keyword := ""
+		if v, ok := callReq.Arguments["keyword"]; ok {
+			keyword = v.(string)
+		}
 		limit := util.MustAnyToInt(callReq.Arguments["limit"])
 		offset := util.MustAnyToInt(callReq.Arguments["offset"])
-		messages, err := s.db.GetMessages(start, end, talker, limit, offset)
+		messages, err := s.db.GetMessages(start, end, talker, sender, keyword, limit, offset)
 		if err != nil {
 			return fmt.Errorf("无法获取聊天记录: %v", err)
 		}
+		if len(messages) == 0 {
+			buf.WriteString("未找到符合查询条件的聊天记录")
+		}
 		for _, m := range messages {
-			buf.WriteString(m.PlainText(len(talker) == 0, ""))
+			buf.WriteString(m.PlainText(strings.Contains(talker, ","), util.PerfectTimeFormat(start, end), ""))
 			buf.WriteString("\n")
 		}
+	case "current_time":
+		buf.WriteString(time.Now().Local().Format(time.RFC3339))
 	default:
 		return fmt.Errorf("未支持的工具: %s", callReq.Name)
 	}
@@ -228,7 +243,6 @@ func (s *Service) resourcesRead(session *mcp.Session, req *mcp.Request) error {
 	buf := &bytes.Buffer{}
 	switch u.Scheme {
 	case "contact":
-
 		list, err := s.db.GetContacts(u.Host, 0, 0)
 		if err != nil {
 			return fmt.Errorf("无法获取联系人列表: %v", err)
@@ -262,12 +276,15 @@ func (s *Service) resourcesRead(session *mcp.Session, req *mcp.Request) error {
 		}
 		limit := util.MustAnyToInt(u.Query().Get("limit"))
 		offset := util.MustAnyToInt(u.Query().Get("offset"))
-		messages, err := s.db.GetMessages(start, end, u.Host, limit, offset)
+		messages, err := s.db.GetMessages(start, end, u.Host, "", "", limit, offset)
 		if err != nil {
 			return fmt.Errorf("无法获取聊天记录: %v", err)
 		}
+		if len(messages) == 0 {
+			buf.WriteString("未找到符合查询条件的聊天记录")
+		}
 		for _, m := range messages {
-			buf.WriteString(m.PlainText(len(u.Host) == 0, ""))
+			buf.WriteString(m.PlainText(strings.Contains(u.Host, ","), util.PerfectTimeFormat(start, end), ""))
 			buf.WriteString("\n")
 		}
 	default:
--- a/internal/errors/errors.go
+++ b/internal/errors/errors.go
@@ -65,7 +65,7 @@ func Newf(cause error, code int, format string, args ...interface{}) *Error {
 	return &Error{
 		Message: fmt.Sprintf(format, args...),
 		Cause:   cause,
-		Code:    http.StatusInternalServerError,
+		Code:    code,
 	}
 }

--- a/internal/errors/middleware.go
+++ b/internal/errors/middleware.go
@@ -2,6 +2,7 @@ package errors

 import (
 	"net/http"
+	"runtime/debug"

 	"github.com/gin-gonic/gin"
 	"github.com/google/uuid"
@@ -50,7 +51,7 @@ func RecoveryMiddleware() gin.HandlerFunc {
 				}

 				// 记录错误日志
-				log.Err(err).Msg("PANIC RECOVERED")
+				log.Err(err).Msgf("PANIC RECOVERED\n%s", string(debug.Stack()))

 				// 返回 500 错误
 				c.JSON(http.StatusInternalServerError, err)
--- a/internal/model/message.go
+++ b/internal/model/message.go
@@ -183,7 +183,11 @@ func (m *Message) SetContent(key string, value interface{}) {
 	m.Contents[key] = value
 }

-func (m *Message) PlainText(showChatRoom bool, host string) string {
+func (m *Message) PlainText(showChatRoom bool, timeFormat string, host string) string {
+
+	if timeFormat == "" {
+		timeFormat = "01-02 15:04:05"
+	}

 	m.SetContent("host", host)

@@ -216,7 +220,7 @@ func (m *Message) PlainText(showChatRoom bool, host string) string {
 		buf.WriteString("] ")
 	}

-	buf.WriteString(m.Time.Format("2006-01-02 15:04:05"))
+	buf.WriteString(m.Time.Format(timeFormat))
 	buf.WriteString("\n")

 	buf.WriteString(m.PlainTextContent())
@@ -262,7 +266,11 @@ func (m *Message) PlainTextContent() string {
 			if !ok {
 				return "[合并转发]"
 			}
-			return recordInfo.String("", m.Contents["host"].(string))
+			host := ""
+			if m.Contents["host"] != nil {
+				host = m.Contents["host"].(string)
+			}
+			return recordInfo.String("", host)
 		case 33, 36:
 			if m.Contents["title"] == "" {
 				return "[小程序]"
@@ -290,7 +298,11 @@ func (m *Message) PlainTextContent() string {
 				return "> [引用]\n" + m.Content
 			}
 			buf := strings.Builder{}
-			referContent := refer.PlainText(false, m.Contents["host"].(string))
+			host := ""
+			if m.Contents["host"] != nil {
+				host = m.Contents["host"].(string)
+			}
+			referContent := refer.PlainText(false, "", host)
 			for _, line := range strings.Split(referContent, "\n") {
 				if line == "" {
 					continue
--- a/internal/wechatdb/datasource/darwinv3/datasource.go
+++ b/internal/wechatdb/datasource/darwinv3/datasource.go
@@ -5,6 +5,8 @@ import (
 	"crypto/md5"
 	"encoding/hex"
 	"fmt"
+	"regexp"
+	"sort"
 	"strings"
 	"time"

@@ -15,6 +17,7 @@ import (
 	"github.com/sjzar/chatlog/internal/errors"
 	"github.com/sjzar/chatlog/internal/model"
 	"github.com/sjzar/chatlog/internal/wechatdb/datasource/dbm"
+	"github.com/sjzar/chatlog/pkg/util"
 )

 const (
@@ -188,70 +191,162 @@ func (ds *DataSource) initChatRoomDb() error {
 	return nil
 }

-// GetMessages 实现获取消息的方法
-func (ds *DataSource) GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, limit, offset int) ([]*model.Message, error) {
-	// 在 darwinv3 中，每个联系人/群聊的消息存储在单独的表中，表名为 Chat_md5(talker)
-	// 首先需要找到对应的表名
+func (ds *DataSource) GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, sender string, keyword string, limit, offset int) ([]*model.Message, error) {
 	if talker == "" {
 		return nil, errors.ErrTalkerEmpty
 	}

-	_talkerMd5Bytes := md5.Sum([]byte(talker))
-	talkerMd5 := hex.EncodeToString(_talkerMd5Bytes[:])
-	dbPath, ok := ds.talkerDBMap[talkerMd5]
-	if !ok {
-		return nil, errors.TalkerNotFound(talker)
+	// 解析talker参数，支持多个talker（以英文逗号分隔）
+	talkers := util.Str2List(talker, ",")
+	if len(talkers) == 0 {
+		return nil, errors.ErrTalkerEmpty
 	}
-	db, err := ds.dbm.OpenDB(dbPath)
-	if err != nil {
-		return nil, err
-	}
-	tableName := fmt.Sprintf("Chat_%s", talkerMd5)

-	// 构建查询条件
-	query := fmt.Sprintf(`
-		SELECT msgCreateTime, msgContent, messageType, mesDes
-		FROM %s 
-		WHERE msgCreateTime >= ? AND msgCreateTime <= ? 
-		ORDER BY msgCreateTime ASC
-	`, tableName)
+	// 解析sender参数，支持多个发送者（以英文逗号分隔）
+	senders := util.Str2List(sender, ",")

-	if limit > 0 {
-		query += fmt.Sprintf(" LIMIT %d", limit)
-
-		if offset > 0 {
-			query += fmt.Sprintf(" OFFSET %d", offset)
+	// 预编译正则表达式（如果有keyword）
+	var regex *regexp.Regexp
+	if keyword != "" {
+		var err error
+		regex, err = regexp.Compile(keyword)
+		if err != nil {
+			return nil, errors.QueryFailed("invalid regex pattern", err)
 		}
 	}

-	// 执行查询
-	rows, err := db.QueryContext(ctx, query, startTime.Unix(), endTime.Unix())
-	if err != nil {
-		return nil, errors.QueryFailed(query, err)
-	}
-	defer rows.Close()
+	// 从每个相关数据库中查询消息，并在读取时进行过滤
+	filteredMessages := []*model.Message{}

-	// 处理查询结果
-	messages := []*model.Message{}
-	for rows.Next() {
-		var msg model.MessageDarwinV3
-		err := rows.Scan(
-			&msg.MsgCreateTime,
-			&msg.MsgContent,
-			&msg.MessageType,
-			&msg.MesDes,
-		)
-		if err != nil {
-			log.Err(err).Msgf("扫描消息行失败")
+	// 对每个talker进行查询
+	for _, talkerItem := range talkers {
+		// 检查上下文是否已取消
+		if err := ctx.Err(); err != nil {
+			return nil, err
+		}
+
+		// 在 darwinv3 中，需要先找到对应的数据库
+		_talkerMd5Bytes := md5.Sum([]byte(talkerItem))
+		talkerMd5 := hex.EncodeToString(_talkerMd5Bytes[:])
+		dbPath, ok := ds.talkerDBMap[talkerMd5]
+		if !ok {
+			// 如果找不到对应的数据库，跳过此talker
 			continue
 		}

-		// 将消息包装为通用模型
-		message := msg.Wrap(talker)
-		messages = append(messages, message)
+		db, err := ds.dbm.OpenDB(dbPath)
+		if err != nil {
+			log.Error().Msgf("数据库 %s 未打开", dbPath)
+			continue
+		}
+
+		tableName := fmt.Sprintf("Chat_%s", talkerMd5)
+
+		// 构建查询条件
+		query := fmt.Sprintf(`
+			SELECT msgCreateTime, msgContent, messageType, mesDes
+			FROM %s 
+			WHERE msgCreateTime >= ? AND msgCreateTime <= ? 
+			ORDER BY msgCreateTime ASC
+		`, tableName)
+
+		// 执行查询
+		rows, err := db.QueryContext(ctx, query, startTime.Unix(), endTime.Unix())
+		if err != nil {
+			// 如果表不存在，跳过此talker
+			if strings.Contains(err.Error(), "no such table") {
+				continue
+			}
+			log.Err(err).Msgf("从数据库 %s 查询消息失败", dbPath)
+			continue
+		}
+
+		// 处理查询结果，在读取时进行过滤
+		for rows.Next() {
+			var msg model.MessageDarwinV3
+			err := rows.Scan(
+				&msg.MsgCreateTime,
+				&msg.MsgContent,
+				&msg.MessageType,
+				&msg.MesDes,
+			)
+			if err != nil {
+				rows.Close()
+				log.Err(err).Msgf("扫描消息行失败")
+				continue
+			}
+
+			// 将消息包装为通用模型
+			message := msg.Wrap(talkerItem)
+
+			// 应用sender过滤
+			if len(senders) > 0 {
+				senderMatch := false
+				for _, s := range senders {
+					if message.Sender == s {
+						senderMatch = true
+						break
+					}
+				}
+				if !senderMatch {
+					continue // 不匹配sender，跳过此消息
+				}
+			}
+
+			// 应用keyword过滤
+			if regex != nil {
+				plainText := message.PlainTextContent()
+				if !regex.MatchString(plainText) {
+					continue // 不匹配keyword，跳过此消息
+				}
+			}
+
+			// 通过所有过滤条件，保留此消息
+			filteredMessages = append(filteredMessages, message)
+
+			// 检查是否已经满足分页处理数量
+			if limit > 0 && len(filteredMessages) >= offset+limit {
+				// 已经获取了足够的消息，可以提前返回
+				rows.Close()
+
+				// 对所有消息按时间排序
+				sort.Slice(filteredMessages, func(i, j int) bool {
+					return filteredMessages[i].Seq < filteredMessages[j].Seq
+				})
+
+				// 处理分页
+				if offset >= len(filteredMessages) {
+					return []*model.Message{}, nil
+				}
+				end := offset + limit
+				if end > len(filteredMessages) {
+					end = len(filteredMessages)
+				}
+				return filteredMessages[offset:end], nil
+			}
+		}
+		rows.Close()
 	}

-	return messages, nil
+	// 对所有消息按时间排序
+	// FIXME 不同 talker 需要使用 Time 排序
+	sort.Slice(filteredMessages, func(i, j int) bool {
+		return filteredMessages[i].Time.Before(filteredMessages[j].Time)
+	})
+
+	// 处理分页
+	if limit > 0 {
+		if offset >= len(filteredMessages) {
+			return []*model.Message{}, nil
+		}
+		end := offset + limit
+		if end > len(filteredMessages) {
+			end = len(filteredMessages)
+		}
+		return filteredMessages[offset:end], nil
+	}
+
+	return filteredMessages, nil
 }

 // 从表名中提取 talker
--- a/internal/wechatdb/datasource/datasource.go
+++ b/internal/wechatdb/datasource/datasource.go
@@ -16,7 +16,7 @@ import (
 type DataSource interface {

 	// 消息
-	GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, limit, offset int) ([]*model.Message, error)
+	GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, sender string, keyword string, limit, offset int) ([]*model.Message, error)

 	// 联系人
 	GetContacts(ctx context.Context, key string, limit, offset int) ([]*model.Contact, error)
--- a/internal/wechatdb/datasource/v4/datasource.go
+++ b/internal/wechatdb/datasource/v4/datasource.go
@@ -6,6 +6,7 @@ import (
 	"database/sql"
 	"encoding/hex"
 	"fmt"
+	"regexp"
 	"sort"
 	"strings"
 	"time"
@@ -17,6 +18,7 @@ import (
 	"github.com/sjzar/chatlog/internal/errors"
 	"github.com/sjzar/chatlog/internal/model"
 	"github.com/sjzar/chatlog/internal/wechatdb/datasource/dbm"
+	"github.com/sjzar/chatlog/pkg/util"
 )

 const (
@@ -175,11 +177,16 @@ func (ds *DataSource) getDBInfosForTimeRange(startTime, endTime time.Time) []Mes
 	return dbs
 }

-func (ds *DataSource) GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, limit, offset int) ([]*model.Message, error) {
+func (ds *DataSource) GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, sender string, keyword string, limit, offset int) ([]*model.Message, error) {
 	if talker == "" {
 		return nil, errors.ErrTalkerEmpty
 	}
-	log.Debug().Msg(talker)
+
+	// 解析talker参数，支持多个talker（以英文逗号分隔）
+	talkers := util.Str2List(talker, ",")
+	if len(talkers) == 0 {
+		return nil, errors.ErrTalkerEmpty
+	}

 	// 找到时间范围内的数据库文件
 	dbInfos := ds.getDBInfosForTimeRange(startTime, endTime)
@@ -187,13 +194,21 @@ func (ds *DataSource) GetMessages(ctx context.Context, startTime, endTime time.T
 		return nil, errors.TimeRangeNotFound(startTime, endTime)
 	}

-	if len(dbInfos) == 1 {
-		// LIMIT 和 OFFSET 逻辑在单文件情况下可以直接在 SQL 里处理
-		return ds.getMessagesSingleFile(ctx, dbInfos[0], startTime, endTime, talker, limit, offset)
+	// 解析sender参数，支持多个发送者（以英文逗号分隔）
+	senders := util.Str2List(sender, ",")
+
+	// 预编译正则表达式（如果有keyword）
+	var regex *regexp.Regexp
+	if keyword != "" {
+		var err error
+		regex, err = regexp.Compile(keyword)
+		if err != nil {
+			return nil, errors.QueryFailed("invalid regex pattern", err)
+		}
 	}

-	// 从每个相关数据库中查询消息
-	totalMessages := []*model.Message{}
+	// 从每个相关数据库中查询消息，并在读取时进行过滤
+	filteredMessages := []*model.Message{}

 	for _, dbInfo := range dbInfos {
 		// 检查上下文是否已取消
@@ -207,183 +222,141 @@ func (ds *DataSource) GetMessages(ctx context.Context, startTime, endTime time.T
 			continue
 		}

-		messages, err := ds.getMessagesFromDB(ctx, db, startTime, endTime, talker)
-		if err != nil {
-			log.Err(err).Msgf("从数据库 %s 获取消息失败", dbInfo.FilePath)
-			continue
-		}
+		// 对每个talker进行查询
+		for _, talkerItem := range talkers {
+			// 构建表名
+			_talkerMd5Bytes := md5.Sum([]byte(talkerItem))
+			talkerMd5 := hex.EncodeToString(_talkerMd5Bytes[:])
+			tableName := "Msg_" + talkerMd5

-		totalMessages = append(totalMessages, messages...)
+			// 检查表是否存在
+			var exists bool
+			err = db.QueryRowContext(ctx,
+				"SELECT 1 FROM sqlite_master WHERE type='table' AND name=?",
+				tableName).Scan(&exists)

-		if limit+offset > 0 && len(totalMessages) >= limit+offset {
-			break
+			if err != nil {
+				if err == sql.ErrNoRows {
+					// 表不存在，继续下一个talker
+					continue
+				}
+				return nil, errors.QueryFailed("", err)
+			}
+
+			// 构建查询条件
+			conditions := []string{"create_time >= ? AND create_time <= ?"}
+			args := []interface{}{startTime.Unix(), endTime.Unix()}
+			log.Debug().Msgf("Table name: %s", tableName)
+			log.Debug().Msgf("Start time: %d, End time: %d", startTime.Unix(), endTime.Unix())
+
+			query := fmt.Sprintf(`
+				SELECT m.sort_seq, m.server_id, m.local_type, n.user_name, m.create_time, m.message_content, m.packed_info_data, m.status
+				FROM %s m
+				LEFT JOIN Name2Id n ON m.real_sender_id = n.rowid
+				WHERE %s 
+				ORDER BY m.sort_seq ASC
+			`, tableName, strings.Join(conditions, " AND "))
+
+			// 执行查询
+			rows, err := db.QueryContext(ctx, query, args...)
+			if err != nil {
+				// 如果表不存在，SQLite 会返回错误
+				if strings.Contains(err.Error(), "no such table") {
+					continue
+				}
+				log.Err(err).Msgf("从数据库 %s 查询消息失败", dbInfo.FilePath)
+				continue
+			}
+
+			// 处理查询结果，在读取时进行过滤
+			for rows.Next() {
+				var msg model.MessageV4
+				err := rows.Scan(
+					&msg.SortSeq,
+					&msg.ServerID,
+					&msg.LocalType,
+					&msg.UserName,
+					&msg.CreateTime,
+					&msg.MessageContent,
+					&msg.PackedInfoData,
+					&msg.Status,
+				)
+				if err != nil {
+					rows.Close()
+					return nil, errors.ScanRowFailed(err)
+				}
+
+				// 将消息转换为标准格式
+				message := msg.Wrap(talkerItem)
+
+				// 应用sender过滤
+				if len(senders) > 0 {
+					senderMatch := false
+					for _, s := range senders {
+						if message.Sender == s {
+							senderMatch = true
+							break
+						}
+					}
+					if !senderMatch {
+						continue // 不匹配sender，跳过此消息
+					}
+				}
+
+				// 应用keyword过滤
+				if regex != nil {
+					plainText := message.PlainTextContent()
+					if !regex.MatchString(plainText) {
+						continue // 不匹配keyword，跳过此消息
+					}
+				}
+
+				// 通过所有过滤条件，保留此消息
+				filteredMessages = append(filteredMessages, message)
+
+				// 检查是否已经满足分页处理数量
+				if limit > 0 && len(filteredMessages) >= offset+limit {
+					// 已经获取了足够的消息，可以提前返回
+					rows.Close()
+
+					// 对所有消息按时间排序
+					sort.Slice(filteredMessages, func(i, j int) bool {
+						return filteredMessages[i].Seq < filteredMessages[j].Seq
+					})
+
+					// 处理分页
+					if offset >= len(filteredMessages) {
+						return []*model.Message{}, nil
+					}
+					end := offset + limit
+					if end > len(filteredMessages) {
+						end = len(filteredMessages)
+					}
+					return filteredMessages[offset:end], nil
+				}
+			}
+			rows.Close()
 		}
 	}

 	// 对所有消息按时间排序
-	sort.Slice(totalMessages, func(i, j int) bool {
-		return totalMessages[i].Seq < totalMessages[j].Seq
+	sort.Slice(filteredMessages, func(i, j int) bool {
+		return filteredMessages[i].Seq < filteredMessages[j].Seq
 	})

 	// 处理分页
 	if limit > 0 {
-		if offset >= len(totalMessages) {
+		if offset >= len(filteredMessages) {
 			return []*model.Message{}, nil
 		}
 		end := offset + limit
-		if end > len(totalMessages) {
-			end = len(totalMessages)
+		if end > len(filteredMessages) {
+			end = len(filteredMessages)
 		}
-		return totalMessages[offset:end], nil
+		return filteredMessages[offset:end], nil
 	}

-	return totalMessages, nil
-}
-
-// getMessagesSingleFile 从单个数据库文件获取消息
-func (ds *DataSource) getMessagesSingleFile(ctx context.Context, dbInfo MessageDBInfo, startTime, endTime time.Time, talker string, limit, offset int) ([]*model.Message, error) {
-	db, err := ds.dbm.OpenDB(dbInfo.FilePath)
-	if err != nil {
-		return nil, errors.DBConnectFailed(dbInfo.FilePath, nil)
-	}
-
-	// 构建表名
-	_talkerMd5Bytes := md5.Sum([]byte(talker))
-	talkerMd5 := hex.EncodeToString(_talkerMd5Bytes[:])
-	tableName := "Msg_" + talkerMd5
-
-	// 检查表是否存在
-	var exists bool
-	err = db.QueryRowContext(ctx,
-		"SELECT 1 FROM sqlite_master WHERE type='table' AND name=?",
-		tableName).Scan(&exists)
-
-	if err != nil {
-		if err == sql.ErrNoRows {
-			// 表不存在，返回空结果
-			return []*model.Message{}, nil
-		}
-		return nil, errors.QueryFailed("", err)
-	}
-
-	// 构建查询条件
-	conditions := []string{"create_time >= ? AND create_time <= ?"}
-	args := []interface{}{startTime.Unix(), endTime.Unix()}
-
-	query := fmt.Sprintf(`
-		SELECT m.sort_seq, m.server_id, m.local_type, n.user_name, m.create_time, m.message_content, m.packed_info_data, m.status
-		FROM %s m
-		LEFT JOIN Name2Id n ON m.real_sender_id = n.rowid
-		WHERE %s 
-		ORDER BY m.sort_seq ASC
-	`, tableName, strings.Join(conditions, " AND "))
-
-	if limit > 0 {
-		query += fmt.Sprintf(" LIMIT %d", limit)
-		if offset > 0 {
-			query += fmt.Sprintf(" OFFSET %d", offset)
-		}
-	}
-
-	// 执行查询
-	rows, err := db.QueryContext(ctx, query, args...)
-	if err != nil {
-		return nil, errors.QueryFailed(query, err)
-	}
-	defer rows.Close()
-
-	// 处理查询结果
-	messages := []*model.Message{}
-
-	for rows.Next() {
-		var msg model.MessageV4
-		err := rows.Scan(
-			&msg.SortSeq,
-			&msg.ServerID,
-			&msg.LocalType,
-			&msg.UserName,
-			&msg.CreateTime,
-			&msg.MessageContent,
-			&msg.PackedInfoData,
-			&msg.Status,
-		)
-		if err != nil {
-			return nil, errors.ScanRowFailed(err)
-		}
-
-		messages = append(messages, msg.Wrap(talker))
-	}
-
-	return messages, nil
-}
-
-// getMessagesFromDB 从数据库获取消息
-func (ds *DataSource) getMessagesFromDB(ctx context.Context, db *sql.DB, startTime, endTime time.Time, talker string) ([]*model.Message, error) {
-	// 构建表名
-	_talkerMd5Bytes := md5.Sum([]byte(talker))
-	talkerMd5 := hex.EncodeToString(_talkerMd5Bytes[:])
-	tableName := "Msg_" + talkerMd5
-
-	// 检查表是否存在
-	var exists bool
-	err := db.QueryRowContext(ctx,
-		"SELECT 1 FROM sqlite_master WHERE type='table' AND name=?",
-		tableName).Scan(&exists)
-
-	if err != nil {
-		if err == sql.ErrNoRows {
-			// 表不存在，返回空结果
-			return []*model.Message{}, nil
-		}
-		return nil, errors.QueryFailed("", err)
-	}
-
-	// 构建查询条件
-	conditions := []string{"create_time >= ? AND create_time <= ?"}
-	args := []interface{}{startTime.Unix(), endTime.Unix()}
-
-	query := fmt.Sprintf(`
-		SELECT m.sort_seq, m.server_id, m.local_type, n.user_name, m.create_time, m.message_content, m.packed_info_data, m.status
-		FROM %s m
-		LEFT JOIN Name2Id n ON m.real_sender_id = n.rowid
-		WHERE %s 
-		ORDER BY m.sort_seq ASC
-	`, tableName, strings.Join(conditions, " AND "))
-
-	// 执行查询
-	rows, err := db.QueryContext(ctx, query, args...)
-	if err != nil {
-		// 如果表不存在，SQLite 会返回错误
-		if strings.Contains(err.Error(), "no such table") {
-			return []*model.Message{}, nil
-		}
-		return nil, errors.QueryFailed(query, err)
-	}
-	defer rows.Close()
-
-	// 处理查询结果
-	messages := []*model.Message{}
-
-	for rows.Next() {
-		var msg model.MessageV4
-		err := rows.Scan(
-			&msg.SortSeq,
-			&msg.ServerID,
-			&msg.LocalType,
-			&msg.UserName,
-			&msg.CreateTime,
-			&msg.MessageContent,
-			&msg.PackedInfoData,
-			&msg.Status,
-		)
-		if err != nil {
-			return nil, errors.ScanRowFailed(err)
-		}
-
-		messages = append(messages, msg.Wrap(talker))
-	}
-
-	return messages, nil
+	return filteredMessages, nil
 }

 // 联系人
--- a/internal/wechatdb/datasource/windowsv3/datasource.go
+++ b/internal/wechatdb/datasource/windowsv3/datasource.go
@@ -2,9 +2,9 @@ package windowsv3

 import (
 	"context"
-	"database/sql"
 	"encoding/hex"
 	"fmt"
+	"regexp"
 	"sort"
 	"strings"
 	"time"
@@ -16,6 +16,7 @@ import (
 	"github.com/sjzar/chatlog/internal/errors"
 	"github.com/sjzar/chatlog/internal/model"
 	"github.com/sjzar/chatlog/internal/wechatdb/datasource/dbm"
+	"github.com/sjzar/chatlog/pkg/util"
 )

 const (
@@ -221,21 +222,38 @@ func (ds *DataSource) getDBInfosForTimeRange(startTime, endTime time.Time) []Mes
 	return dbs
 }

-// GetMessages 实现 DataSource 接口的 GetMessages 方法
-func (ds *DataSource) GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, limit, offset int) ([]*model.Message, error) {
+func (ds *DataSource) GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, sender string, keyword string, limit, offset int) ([]*model.Message, error) {
+	if talker == "" {
+		return nil, errors.ErrTalkerEmpty
+	}
+
+	// 解析talker参数，支持多个talker（以英文逗号分隔）
+	talkers := util.Str2List(talker, ",")
+	if len(talkers) == 0 {
+		return nil, errors.ErrTalkerEmpty
+	}
+
 	// 找到时间范围内的数据库文件
 	dbInfos := ds.getDBInfosForTimeRange(startTime, endTime)
 	if len(dbInfos) == 0 {
 		return nil, errors.TimeRangeNotFound(startTime, endTime)
 	}

-	if len(dbInfos) == 1 {
-		// LIMIT 和 OFFSET 逻辑在单文件情况下可以直接在 SQL 里处理
-		return ds.getMessagesSingleFile(ctx, dbInfos[0], startTime, endTime, talker, limit, offset)
+	// 解析sender参数，支持多个发送者（以英文逗号分隔）
+	senders := util.Str2List(sender, ",")
+
+	// 预编译正则表达式（如果有keyword）
+	var regex *regexp.Regexp
+	if keyword != "" {
+		var err error
+		regex, err = regexp.Compile(keyword)
+		if err != nil {
+			return nil, errors.QueryFailed("invalid regex pattern", err)
+		}
 	}

 	// 从每个相关数据库中查询消息
-	totalMessages := []*model.Message{}
+	filteredMessages := []*model.Message{}

 	for _, dbInfo := range dbInfos {
 		// 检查上下文是否已取消
@@ -249,172 +267,137 @@ func (ds *DataSource) GetMessages(ctx context.Context, startTime, endTime time.T
 			continue
 		}

-		messages, err := ds.getMessagesFromDB(ctx, db, dbInfo, startTime, endTime, talker)
-		if err != nil {
-			log.Err(err).Msgf("从数据库 %s 获取消息失败", dbInfo.FilePath)
-			continue
-		}
+		// 对每个talker进行查询
+		for _, talkerItem := range talkers {
+			// 构建查询条件
+			conditions := []string{"Sequence >= ? AND Sequence <= ?"}
+			args := []interface{}{startTime.Unix() * 1000, endTime.Unix() * 1000}

-		totalMessages = append(totalMessages, messages...)
+			// 添加talker条件
+			talkerID, ok := dbInfo.TalkerMap[talkerItem]
+			if ok {
+				conditions = append(conditions, "TalkerId = ?")
+				args = append(args, talkerID)
+			} else {
+				conditions = append(conditions, "StrTalker = ?")
+				args = append(args, talkerItem)
+			}

-		if limit+offset > 0 && len(totalMessages) >= limit+offset {
-			break
+			query := fmt.Sprintf(`
+				SELECT MsgSvrID, Sequence, CreateTime, StrTalker, IsSender, 
+					Type, SubType, StrContent, CompressContent, BytesExtra
+				FROM MSG 
+				WHERE %s 
+				ORDER BY Sequence ASC
+			`, strings.Join(conditions, " AND "))
+
+			// 执行查询
+			rows, err := db.QueryContext(ctx, query, args...)
+			if err != nil {
+				// 如果表不存在，跳过此talker
+				if strings.Contains(err.Error(), "no such table") {
+					continue
+				}
+				log.Err(err).Msgf("从数据库 %s 查询消息失败", dbInfo.FilePath)
+				continue
+			}
+
+			// 处理查询结果，在读取时进行过滤
+			for rows.Next() {
+				var msg model.MessageV3
+				var compressContent []byte
+				var bytesExtra []byte
+
+				err := rows.Scan(
+					&msg.MsgSvrID,
+					&msg.Sequence,
+					&msg.CreateTime,
+					&msg.StrTalker,
+					&msg.IsSender,
+					&msg.Type,
+					&msg.SubType,
+					&msg.StrContent,
+					&compressContent,
+					&bytesExtra,
+				)
+				if err != nil {
+					rows.Close()
+					return nil, errors.ScanRowFailed(err)
+				}
+				msg.CompressContent = compressContent
+				msg.BytesExtra = bytesExtra
+
+				// 将消息转换为标准格式
+				message := msg.Wrap()
+
+				// 应用sender过滤
+				if len(senders) > 0 {
+					senderMatch := false
+					for _, s := range senders {
+						if message.Sender == s {
+							senderMatch = true
+							break
+						}
+					}
+					if !senderMatch {
+						continue // 不匹配sender，跳过此消息
+					}
+				}
+
+				// 应用keyword过滤
+				if regex != nil {
+					plainText := message.PlainTextContent()
+					if !regex.MatchString(plainText) {
+						continue // 不匹配keyword，跳过此消息
+					}
+				}
+
+				// 通过所有过滤条件，保留此消息
+				filteredMessages = append(filteredMessages, message)
+
+				// 检查是否已经满足分页处理数量
+				if limit > 0 && len(filteredMessages) >= offset+limit {
+					// 已经获取了足够的消息，可以提前返回
+					rows.Close()
+
+					// 对所有消息按时间排序
+					sort.Slice(filteredMessages, func(i, j int) bool {
+						return filteredMessages[i].Seq < filteredMessages[j].Seq
+					})
+
+					// 处理分页
+					if offset >= len(filteredMessages) {
+						return []*model.Message{}, nil
+					}
+					end := offset + limit
+					if end > len(filteredMessages) {
+						end = len(filteredMessages)
+					}
+					return filteredMessages[offset:end], nil
+				}
+			}
+			rows.Close()
 		}
 	}

 	// 对所有消息按时间排序
-	sort.Slice(totalMessages, func(i, j int) bool {
-		return totalMessages[i].Seq < totalMessages[j].Seq
+	sort.Slice(filteredMessages, func(i, j int) bool {
+		return filteredMessages[i].Seq < filteredMessages[j].Seq
 	})

 	// 处理分页
 	if limit > 0 {
-		if offset >= len(totalMessages) {
+		if offset >= len(filteredMessages) {
 			return []*model.Message{}, nil
 		}
 		end := offset + limit
-		if end > len(totalMessages) {
-			end = len(totalMessages)
+		if end > len(filteredMessages) {
+			end = len(filteredMessages)
 		}
-		return totalMessages[offset:end], nil
+		return filteredMessages[offset:end], nil
 	}

-	return totalMessages, nil
-}
-
-// getMessagesSingleFile 从单个数据库文件获取消息
-func (ds *DataSource) getMessagesSingleFile(ctx context.Context, dbInfo MessageDBInfo, startTime, endTime time.Time, talker string, limit, offset int) ([]*model.Message, error) {
-	db, err := ds.dbm.OpenDB(dbInfo.FilePath)
-	if err != nil {
-		return nil, errors.DBConnectFailed(dbInfo.FilePath, nil)
-	}
-
-	// 构建查询条件
-	conditions := []string{"Sequence >= ? AND Sequence <= ?"}
-	args := []interface{}{startTime.Unix() * 1000, endTime.Unix() * 1000}
-	if len(talker) > 0 {
-		// TalkerId 有索引，优先使用
-		talkerID, ok := dbInfo.TalkerMap[talker]
-		if ok {
-			conditions = append(conditions, "TalkerId = ?")
-			args = append(args, talkerID)
-		} else {
-			conditions = append(conditions, "StrTalker = ?")
-			args = append(args, talker)
-		}
-	}
-	query := fmt.Sprintf(`
-        SELECT MsgSvrID, Sequence, CreateTime, StrTalker, IsSender, 
-            Type, SubType, StrContent, CompressContent, BytesExtra
-        FROM MSG 
-        WHERE %s 
-        ORDER BY Sequence ASC
-    `, strings.Join(conditions, " AND "))
-
-	if limit > 0 {
-		query += fmt.Sprintf(" LIMIT %d", limit)
-
-		if offset > 0 {
-			query += fmt.Sprintf(" OFFSET %d", offset)
-		}
-	}
-
-	// 执行查询
-	rows, err := db.QueryContext(ctx, query, args...)
-	if err != nil {
-		return nil, errors.QueryFailed(query, err)
-	}
-	defer rows.Close()
-
-	// 处理查询结果
-	totalMessages := []*model.Message{}
-	for rows.Next() {
-		var msg model.MessageV3
-		var compressContent []byte
-		var bytesExtra []byte
-		err := rows.Scan(
-			&msg.MsgSvrID,
-			&msg.Sequence,
-			&msg.CreateTime,
-			&msg.StrTalker,
-			&msg.IsSender,
-			&msg.Type,
-			&msg.SubType,
-			&msg.StrContent,
-			&compressContent,
-			&bytesExtra,
-		)
-		if err != nil {
-			return nil, errors.ScanRowFailed(err)
-		}
-		msg.CompressContent = compressContent
-		msg.BytesExtra = bytesExtra
-		totalMessages = append(totalMessages, msg.Wrap())
-	}
-	return totalMessages, nil
-}
-
-// getMessagesFromDB 从数据库获取消息
-func (ds *DataSource) getMessagesFromDB(ctx context.Context, db *sql.DB, dbInfo MessageDBInfo, startTime, endTime time.Time, talker string) ([]*model.Message, error) {
-	// 构建查询条件
-	conditions := []string{"Sequence >= ? AND Sequence <= ?"}
-	args := []interface{}{startTime.Unix() * 1000, endTime.Unix() * 1000}
-
-	if len(talker) > 0 {
-		talkerID, ok := dbInfo.TalkerMap[talker]
-		if ok {
-			conditions = append(conditions, "TalkerId = ?")
-			args = append(args, talkerID)
-		} else {
-			conditions = append(conditions, "StrTalker = ?")
-			args = append(args, talker)
-		}
-	}
-
-	query := fmt.Sprintf(`
-		SELECT MsgSvrID, Sequence, CreateTime, StrTalker, IsSender, 
-			Type, SubType, StrContent, CompressContent, BytesExtra
-		FROM MSG 
-		WHERE %s 
-		ORDER BY Sequence ASC
-	`, strings.Join(conditions, " AND "))
-
-	// 执行查询
-	rows, err := db.QueryContext(ctx, query, args...)
-	if err != nil {
-		return nil, errors.QueryFailed(query, err)
-	}
-	defer rows.Close()
-
-	// 处理查询结果
-	messages := []*model.Message{}
-	for rows.Next() {
-		var msg model.MessageV3
-		var compressContent []byte
-		var bytesExtra []byte
-
-		err := rows.Scan(
-			&msg.MsgSvrID,
-			&msg.Sequence,
-			&msg.CreateTime,
-			&msg.StrTalker,
-			&msg.IsSender,
-			&msg.Type,
-			&msg.SubType,
-			&msg.StrContent,
-			&compressContent,
-			&bytesExtra,
-		)
-		if err != nil {
-			return nil, errors.ScanRowFailed(err)
-		}
-		msg.CompressContent = compressContent
-		msg.BytesExtra = bytesExtra
-
-		messages = append(messages, msg.Wrap())
-	}
-
-	return messages, nil
+	return filteredMessages, nil
 }

 // GetContacts 实现获取联系人信息的方法
--- a/internal/wechatdb/repository/chatroom.go
+++ b/internal/wechatdb/repository/chatroom.go
@@ -18,8 +18,8 @@ func (r *Repository) initChatRoomCache(ctx context.Context) error {
 	}

 	chatRoomMap := make(map[string]*model.ChatRoom)
-	remarkToChatRoom := make(map[string]*model.ChatRoom)
-	nickNameToChatRoom := make(map[string]*model.ChatRoom)
+	remarkToChatRoom := make(map[string][]*model.ChatRoom)
+	nickNameToChatRoom := make(map[string][]*model.ChatRoom)
 	chatRoomList := make([]string, 0)
 	chatRoomRemark := make([]string, 0)
 	chatRoomNickName := make([]string, 0)
@@ -30,11 +30,21 @@ func (r *Repository) initChatRoomCache(ctx context.Context) error {
 		chatRoomMap[chatRoom.Name] = chatRoom
 		chatRoomList = append(chatRoomList, chatRoom.Name)
 		if chatRoom.Remark != "" {
-			remarkToChatRoom[chatRoom.Remark] = chatRoom
+			remark, ok := remarkToChatRoom[chatRoom.Remark]
+			if !ok {
+				remark = make([]*model.ChatRoom, 0)
+			}
+			remark = append(remark, chatRoom)
+			remarkToChatRoom[chatRoom.Remark] = remark
 			chatRoomRemark = append(chatRoomRemark, chatRoom.Remark)
 		}
 		if chatRoom.NickName != "" {
-			nickNameToChatRoom[chatRoom.NickName] = chatRoom
+			nickName, ok := nickNameToChatRoom[chatRoom.NickName]
+			if !ok {
+				nickName = make([]*model.ChatRoom, 0)
+			}
+			nickName = append(nickName, chatRoom)
+			nickNameToChatRoom[chatRoom.NickName] = nickName
 			chatRoomNickName = append(chatRoomNickName, chatRoom.NickName)
 		}
 	}
@@ -49,11 +59,21 @@ func (r *Repository) initChatRoomCache(ctx context.Context) error {
 			chatRoomMap[contact.UserName] = chatRoom
 			chatRoomList = append(chatRoomList, contact.UserName)
 			if contact.Remark != "" {
-				remarkToChatRoom[contact.Remark] = chatRoom
+				remark, ok := remarkToChatRoom[chatRoom.Remark]
+				if !ok {
+					remark = make([]*model.ChatRoom, 0)
+				}
+				remark = append(remark, chatRoom)
+				remarkToChatRoom[chatRoom.Remark] = remark
 				chatRoomRemark = append(chatRoomRemark, contact.Remark)
 			}
 			if contact.NickName != "" {
-				nickNameToChatRoom[contact.NickName] = chatRoom
+				nickName, ok := nickNameToChatRoom[chatRoom.NickName]
+				if !ok {
+					nickName = make([]*model.ChatRoom, 0)
+				}
+				nickName = append(nickName, chatRoom)
+				nickNameToChatRoom[chatRoom.NickName] = nickName
 				chatRoomNickName = append(chatRoomNickName, contact.NickName)
 			}
 		}
@@ -63,9 +83,12 @@ func (r *Repository) initChatRoomCache(ctx context.Context) error {
 	sort.Strings(chatRoomNickName)

 	r.chatRoomCache = chatRoomMap
-	r.chatRoomList = chatRoomList
 	r.remarkToChatRoom = remarkToChatRoom
 	r.nickNameToChatRoom = nickNameToChatRoom
+	r.chatRoomList = chatRoomList
+	r.chatRoomRemark = chatRoomRemark
+	r.chatRoomNickName = chatRoomNickName
+
 	return nil
 }

@@ -75,7 +98,7 @@ func (r *Repository) GetChatRooms(ctx context.Context, key string, limit, offset
 	if key != "" {
 		ret = r.findChatRooms(key)
 		if len(ret) == 0 {
-			return nil, errors.ChatRoomNotFound(key)
+			return []*model.ChatRoom{}, nil
 		}

 		if limit > 0 {
@@ -129,21 +152,21 @@ func (r *Repository) findChatRoom(key string) *model.ChatRoom {
 		return chatRoom
 	}
 	if chatRoom, ok := r.remarkToChatRoom[key]; ok {
-		return chatRoom
+		return chatRoom[0]
 	}
 	if chatRoom, ok := r.nickNameToChatRoom[key]; ok {
-		return chatRoom
+		return chatRoom[0]
 	}

 	// Contain
 	for _, remark := range r.chatRoomRemark {
 		if strings.Contains(remark, key) {
-			return r.remarkToChatRoom[remark]
+			return r.remarkToChatRoom[remark][0]
 		}
 	}
 	for _, nickName := range r.chatRoomNickName {
 		if strings.Contains(nickName, key) {
-			return r.nickNameToChatRoom[nickName]
+			return r.nickNameToChatRoom[nickName][0]
 		}
 	}

@@ -157,26 +180,42 @@ func (r *Repository) findChatRooms(key string) []*model.ChatRoom {
 		ret = append(ret, chatRoom)
 		distinct[chatRoom.Name] = true
 	}
-	if chatRoom, ok := r.remarkToChatRoom[key]; ok && !distinct[chatRoom.Name] {
-		ret = append(ret, chatRoom)
-		distinct[chatRoom.Name] = true
+	if chatRooms, ok := r.remarkToChatRoom[key]; ok {
+		for _, chatRoom := range chatRooms {
+			if !distinct[chatRoom.Name] {
+				ret = append(ret, chatRoom)
+				distinct[chatRoom.Name] = true
+			}
+		}
 	}
-	if chatRoom, ok := r.nickNameToChatRoom[key]; ok && !distinct[chatRoom.Name] {
-		ret = append(ret, chatRoom)
-		distinct[chatRoom.Name] = true
+	if chatRooms, ok := r.nickNameToChatRoom[key]; ok {
+		for _, chatRoom := range chatRooms {
+			if !distinct[chatRoom.Name] {
+				ret = append(ret, chatRoom)
+				distinct[chatRoom.Name] = true
+			}
+		}
 	}

 	// Contain
 	for _, remark := range r.chatRoomRemark {
-		if strings.Contains(remark, key) && !distinct[r.remarkToChatRoom[remark].Name] {
-			ret = append(ret, r.remarkToChatRoom[remark])
-			distinct[r.remarkToChatRoom[remark].Name] = true
+		if strings.Contains(remark, key) {
+			for _, chatRoom := range r.remarkToChatRoom[remark] {
+				if !distinct[chatRoom.Name] {
+					ret = append(ret, chatRoom)
+					distinct[chatRoom.Name] = true
+				}
+			}
 		}
 	}
 	for _, nickName := range r.chatRoomNickName {
-		if strings.Contains(nickName, key) && !distinct[r.nickNameToChatRoom[nickName].Name] {
-			ret = append(ret, r.nickNameToChatRoom[nickName])
-			distinct[r.nickNameToChatRoom[nickName].Name] = true
+		if strings.Contains(nickName, key) {
+			for _, chatRoom := range r.nickNameToChatRoom[nickName] {
+				if !distinct[chatRoom.Name] {
+					ret = append(ret, chatRoom)
+					distinct[chatRoom.Name] = true
+				}
+			}
 		}
 	}

--- a/internal/wechatdb/repository/contact.go
+++ b/internal/wechatdb/repository/contact.go
@@ -18,9 +18,9 @@ func (r *Repository) initContactCache(ctx context.Context) error {
 	}

 	contactMap := make(map[string]*model.Contact)
-	aliasMap := make(map[string]*model.Contact)
-	remarkMap := make(map[string]*model.Contact)
-	nickNameMap := make(map[string]*model.Contact)
+	aliasMap := make(map[string][]*model.Contact)
+	remarkMap := make(map[string][]*model.Contact)
+	nickNameMap := make(map[string][]*model.Contact)
 	chatRoomUserMap := make(map[string]*model.Contact)
 	chatRoomInContactMap := make(map[string]*model.Contact)
 	contactList := make([]string, 0)
@@ -34,15 +34,30 @@ func (r *Repository) initContactCache(ctx context.Context) error {

 		// 建立快速查找索引
 		if contact.Alias != "" {
-			aliasMap[contact.Alias] = contact
+			alias, ok := aliasMap[contact.Alias]
+			if !ok {
+				alias = make([]*model.Contact, 0)
+			}
+			alias = append(alias, contact)
+			aliasMap[contact.Alias] = alias
 			aliasList = append(aliasList, contact.Alias)
 		}
 		if contact.Remark != "" {
-			remarkMap[contact.Remark] = contact
+			remark, ok := remarkMap[contact.Remark]
+			if !ok {
+				remark = make([]*model.Contact, 0)
+			}
+			remark = append(remark, contact)
+			remarkMap[contact.Remark] = remark
 			remarkList = append(remarkList, contact.Remark)
 		}
 		if contact.NickName != "" {
-			nickNameMap[contact.NickName] = contact
+			nickName, ok := nickNameMap[contact.NickName]
+			if !ok {
+				nickName = make([]*model.Contact, 0)
+			}
+			nickName = append(nickName, contact)
+			nickNameMap[contact.NickName] = nickName
 			nickNameList = append(nickNameList, contact.NickName)
 		}

@@ -88,7 +103,7 @@ func (r *Repository) GetContacts(ctx context.Context, key string, limit, offset
 	if key != "" {
 		ret = r.findContacts(key)
 		if len(ret) == 0 {
-			return nil, errors.ContactNotFound(key)
+			return []*model.Contact{}, nil
 		}
 		if limit > 0 {
 			end := offset + limit
@@ -124,29 +139,29 @@ func (r *Repository) findContact(key string) *model.Contact {
 		return contact
 	}
 	if contact, ok := r.aliasToContact[key]; ok {
-		return contact
+		return contact[0]
 	}
 	if contact, ok := r.remarkToContact[key]; ok {
-		return contact
+		return contact[0]
 	}
 	if contact, ok := r.nickNameToContact[key]; ok {
-		return contact
+		return contact[0]
 	}

 	// Contain
 	for _, alias := range r.aliasList {
 		if strings.Contains(alias, key) {
-			return r.aliasToContact[alias]
+			return r.aliasToContact[alias][0]
 		}
 	}
 	for _, remark := range r.remarkList {
 		if strings.Contains(remark, key) {
-			return r.remarkToContact[remark]
+			return r.remarkToContact[remark][0]
 		}
 	}
 	for _, nickName := range r.nickNameList {
 		if strings.Contains(nickName, key) {
-			return r.nickNameToContact[nickName]
+			return r.nickNameToContact[nickName][0]
 		}
 	}
 	return nil
@@ -159,37 +174,62 @@ func (r *Repository) findContacts(key string) []*model.Contact {
 		ret = append(ret, contact)
 		distinct[contact.UserName] = true
 	}
-	if contact, ok := r.aliasToContact[key]; ok && !distinct[contact.UserName] {
-		ret = append(ret, contact)
-		distinct[contact.UserName] = true
+	if contacts, ok := r.aliasToContact[key]; ok {
+		for _, contact := range contacts {
+			if !distinct[contact.UserName] {
+				ret = append(ret, contact)
+				distinct[contact.UserName] = true
+			}
+		}
 	}
-	if contact, ok := r.remarkToContact[key]; ok && !distinct[contact.UserName] {
-		ret = append(ret, contact)
-		distinct[contact.UserName] = true
+	if contacts, ok := r.remarkToContact[key]; ok {
+		for _, contact := range contacts {
+			if !distinct[contact.UserName] {
+				ret = append(ret, contact)
+				distinct[contact.UserName] = true
+			}
+		}
 	}
-	if contact, ok := r.nickNameToContact[key]; ok && !distinct[contact.UserName] {
-		ret = append(ret, contact)
-		distinct[contact.UserName] = true
+	if contacts, ok := r.nickNameToContact[key]; ok {
+		for _, contact := range contacts {
+			if !distinct[contact.UserName] {
+				ret = append(ret, contact)
+				distinct[contact.UserName] = true
+			}
+		}
 	}
 	// Contain
 	for _, alias := range r.aliasList {
-		if strings.Contains(alias, key) && !distinct[r.aliasToContact[alias].UserName] {
-			ret = append(ret, r.aliasToContact[alias])
-			distinct[r.aliasToContact[alias].UserName] = true
+		if strings.Contains(alias, key) {
+			for _, contact := range r.aliasToContact[alias] {
+				if !distinct[contact.UserName] {
+					ret = append(ret, contact)
+					distinct[contact.UserName] = true
+				}
+			}
 		}
 	}
 	for _, remark := range r.remarkList {
-		if strings.Contains(remark, key) && !distinct[r.remarkToContact[remark].UserName] {
-			ret = append(ret, r.remarkToContact[remark])
-			distinct[r.remarkToContact[remark].UserName] = true
+		if strings.Contains(remark, key) {
+			for _, contact := range r.remarkToContact[remark] {
+				if !distinct[contact.UserName] {
+					ret = append(ret, contact)
+					distinct[contact.UserName] = true
+				}
+			}
 		}
 	}
 	for _, nickName := range r.nickNameList {
-		if strings.Contains(nickName, key) && !distinct[r.nickNameToContact[nickName].UserName] {
-			ret = append(ret, r.nickNameToContact[nickName])
-			distinct[r.nickNameToContact[nickName].UserName] = true
+		if strings.Contains(nickName, key) {
+			for _, contact := range r.nickNameToContact[nickName] {
+				if !distinct[contact.UserName] {
+					ret = append(ret, contact)
+					distinct[contact.UserName] = true
+				}
+			}
 		}
 	}
+
 	return ret
 }

--- a/internal/wechatdb/repository/message.go
+++ b/internal/wechatdb/repository/message.go
@@ -2,23 +2,20 @@ package repository

 import (
 	"context"
+	"strings"
 	"time"

 	"github.com/sjzar/chatlog/internal/model"
+	"github.com/sjzar/chatlog/pkg/util"

 	"github.com/rs/zerolog/log"
 )

 // GetMessages 实现 Repository 接口的 GetMessages 方法
-func (r *Repository) GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, limit, offset int) ([]*model.Message, error) {
+func (r *Repository) GetMessages(ctx context.Context, startTime, endTime time.Time, talker string, sender string, keyword string, limit, offset int) ([]*model.Message, error) {

-	if contact, _ := r.GetContact(ctx, talker); contact != nil {
-		talker = contact.UserName
-	} else if chatRoom, _ := r.GetChatRoom(ctx, talker); chatRoom != nil {
-		talker = chatRoom.Name
-	}
-
-	messages, err := r.ds.GetMessages(ctx, startTime, endTime, talker, limit, offset)
+	talker, sender = r.parseTalkerAndSender(ctx, talker, sender)
+	messages, err := r.ds.GetMessages(ctx, startTime, endTime, talker, sender, keyword, limit, offset)
 	if err != nil {
 		return nil, err
 	}
@@ -62,3 +59,53 @@ func (r *Repository) enrichMessage(msg *model.Message) {
 		}
 	}
 }
+
+func (r *Repository) parseTalkerAndSender(ctx context.Context, talker, sender string) (string, string) {
+	displayName2User := make(map[string]string)
+	users := make(map[string]bool)
+
+	talkers := util.Str2List(talker, ",")
+	if len(talkers) > 0 {
+		for i := 0; i < len(talkers); i++ {
+			if contact, _ := r.GetContact(ctx, talkers[i]); contact != nil {
+				talkers[i] = contact.UserName
+			} else if chatRoom, _ := r.GetChatRoom(ctx, talker); chatRoom != nil {
+				talkers[i] = chatRoom.Name
+			}
+		}
+		// 获取群聊的用户列表
+		for i := 0; i < len(talkers); i++ {
+			if chatRoom, _ := r.GetChatRoom(ctx, talkers[i]); chatRoom != nil {
+				for user, displayName := range chatRoom.User2DisplayName {
+					displayName2User[displayName] = user
+				}
+				for _, user := range chatRoom.Users {
+					users[user.UserName] = true
+				}
+			}
+		}
+		talker = strings.Join(talkers, ",")
+	}
+
+	senders := util.Str2List(sender, ",")
+	if len(senders) > 0 {
+		for i := 0; i < len(senders); i++ {
+			if user, ok := displayName2User[senders[i]]; ok {
+				senders[i] = user
+			} else {
+				// FIXME 大量群聊用户名称重复，无法直接通过 GetContact 获取 ID，后续再优化
+				for user := range users {
+					if contact := r.getFullContact(user); contact != nil {
+						if contact.DisplayName() == senders[i] {
+							senders[i] = user
+							break
+						}
+					}
+				}
+			}
+		}
+		sender = strings.Join(senders, ",")
+	}
+
+	return talker, sender
+}
--- a/internal/wechatdb/repository/repository.go
+++ b/internal/wechatdb/repository/repository.go
@@ -17,9 +17,9 @@ type Repository struct {

 	// Cache for contact
 	contactCache      map[string]*model.Contact
-	aliasToContact    map[string]*model.Contact
-	remarkToContact   map[string]*model.Contact
-	nickNameToContact map[string]*model.Contact
+	aliasToContact    map[string][]*model.Contact
+	remarkToContact   map[string][]*model.Contact
+	nickNameToContact map[string][]*model.Contact
 	chatRoomInContact map[string]*model.Contact
 	contactList       []string
 	aliasList         []string
@@ -28,8 +28,8 @@ type Repository struct {

 	// Cache for chat room
 	chatRoomCache      map[string]*model.ChatRoom
-	remarkToChatRoom   map[string]*model.ChatRoom
-	nickNameToChatRoom map[string]*model.ChatRoom
+	remarkToChatRoom   map[string][]*model.ChatRoom
+	nickNameToChatRoom map[string][]*model.ChatRoom
 	chatRoomList       []string
 	chatRoomRemark     []string
 	chatRoomNickName   []string
@@ -43,17 +43,17 @@ func New(ds datasource.DataSource) (*Repository, error) {
 	r := &Repository{
 		ds:                 ds,
 		contactCache:       make(map[string]*model.Contact),
-		aliasToContact:     make(map[string]*model.Contact),
-		remarkToContact:    make(map[string]*model.Contact),
-		nickNameToContact:  make(map[string]*model.Contact),
+		aliasToContact:     make(map[string][]*model.Contact),
+		remarkToContact:    make(map[string][]*model.Contact),
+		nickNameToContact:  make(map[string][]*model.Contact),
 		chatRoomUserToInfo: make(map[string]*model.Contact),
 		contactList:        make([]string, 0),
 		aliasList:          make([]string, 0),
 		remarkList:         make([]string, 0),
 		nickNameList:       make([]string, 0),
 		chatRoomCache:      make(map[string]*model.ChatRoom),
-		remarkToChatRoom:   make(map[string]*model.ChatRoom),
-		nickNameToChatRoom: make(map[string]*model.ChatRoom),
+		remarkToChatRoom:   make(map[string][]*model.ChatRoom),
+		nickNameToChatRoom: make(map[string][]*model.ChatRoom),
 		chatRoomList:       make([]string, 0),
 		chatRoomRemark:     make([]string, 0),
 		chatRoomNickName:   make([]string, 0),
--- a/internal/wechatdb/wechatdb.go
+++ b/internal/wechatdb/wechatdb.go
@@ -57,11 +57,11 @@ func (w *DB) Initialize() error {
 	return nil
 }

-func (w *DB) GetMessages(start, end time.Time, talker string, limit, offset int) ([]*model.Message, error) {
+func (w *DB) GetMessages(start, end time.Time, talker string, sender string, keyword string, limit, offset int) ([]*model.Message, error) {
 	ctx := context.Background()

 	// 使用 repository 获取消息
-	messages, err := w.repo.GetMessages(ctx, start, end, talker, limit, offset)
+	messages, err := w.repo.GetMessages(ctx, start, end, talker, sender, keyword, limit, offset)
 	if err != nil {
 		return nil, err
 	}
--- a/pkg/util/strings.go
+++ b/pkg/util/strings.go
@@ -3,6 +3,7 @@ package util
 import (
 	"fmt"
 	"strconv"
+	"strings"
 	"unicode"
 	"unicode/utf8"
 )
@@ -45,3 +46,26 @@ func IsNumeric(s string) bool {
 func SplitInt64ToTwoInt32(input int64) (int64, int64) {
 	return input & 0xFFFFFFFF, input >> 32
 }
+
+func Str2List(str string, sep string) []string {
+	list := make([]string, 0)
+
+	if str == "" {
+		return list
+	}
+
+	listMap := make(map[string]bool)
+	for _, elem := range strings.Split(str, sep) {
+		elem = strings.TrimSpace(elem)
+		if len(elem) == 0 {
+			continue
+		}
+		if _, ok := listMap[elem]; ok {
+			continue
+		}
+		listMap[elem] = true
+		list = append(list, elem)
+	}
+
+	return list
+}
--- a/pkg/util/time.go
+++ b/pkg/util/time.go
@@ -582,8 +582,8 @@ func adjustStartTime(t time.Time, g TimeGranularity) time.Time {
 func adjustEndTime(t time.Time, g TimeGranularity) time.Time {
 	switch g {
 	case GranularitySecond, GranularityMinute, GranularityHour:
-		// 对于精确到秒/分钟/小时的时间，设置为当天结束
-		return time.Date(t.Year(), t.Month(), t.Day(), 23, 59, 59, 999999999, t.Location())
+		// 对于精确到秒/分钟/小时的时间，保持原样
+		return t
 	case GranularityDay:
 		// 精确到天，设置为当天结束
 		return time.Date(t.Year(), t.Month(), t.Day(), 23, 59, 59, 999999999, t.Location())
@@ -634,3 +634,25 @@ func isValidDate(year, month, day int) bool {

 	return day <= daysInMonth
 }
+
+func PerfectTimeFormat(start time.Time, end time.Time) string {
+	endTime := end
+
+	// 如果结束时间是某一天的 0 点整，将其减去 1 秒，视为前一天的结束
+	if endTime.Hour() == 0 && endTime.Minute() == 0 && endTime.Second() == 0 && endTime.Nanosecond() == 0 {
+		endTime = endTime.Add(-time.Second) // 减去 1 秒
+	}
+
+	// 判断是否跨年
+	if start.Year() != endTime.Year() {
+		return "2006-01-02 15:04:05" // 完整格式，包含年月日时分秒
+	}
+
+	// 判断是否跨天（但在同一年内）
+	if start.YearDay() != endTime.YearDay() {
+		return "01-02 15:04:05" // 月日时分秒格式
+	}
+
+	// 在同一天内
+	return "15:04:05" // 只显示时分秒
+}