diff --git a/backend/MediaCacheManager.py b/backend/MediaCacheManager.py
index 1e9f1a0..f7e64f5 100644
--- a/backend/MediaCacheManager.py
+++ b/backend/MediaCacheManager.py
@@ -101,9 +101,9 @@ class MediaChunkHolder(object):
         self.mem = self.mem + mem
         self.length = len(self.mem)
         if self.length > self.target_len:
-            logger.warning(RuntimeWarning(
-                f"MeidaChunk Overflow:start:{self.start},len:{self.length},tlen:{self.target_len}"))
+            logger.warning(RuntimeWarning(f"MeidaChunk Overflow:start:{self.start},len:{self.length},tlen:{self.target_len}"))
         self.notify_waiters()
+        self.try_clear_waiter_and_requester()
 
     def add_chunk_requester(self, req: Request) -> None:
         if self.is_completed():
@@ -144,7 +144,7 @@ class MediaChunkHolder(object):
         # clear all waiter and requester
         self.notify_waiters()
         self.requesters.clear()
-        return True 
+        return True
 
 
 class MediaChunkHolderManager(object):
diff --git a/backend/TgFileSystemClient.py b/backend/TgFileSystemClient.py
index 802a9b6..a93b587 100644
--- a/backend/TgFileSystemClient.py
+++ b/backend/TgFileSystemClient.py
@@ -7,7 +7,7 @@ import os
 import functools
 import traceback
 import logging
-from typing import Union, Optional, Literal
+from typing import Union, Optional, Literal, Callable
 
 from telethon import TelegramClient, types, hints, events
 from telethon.custom import QRLogin
@@ -32,8 +32,7 @@ class TgFileSystemClient(object):
     client: TelegramClient
     media_chunk_manager: MediaChunkHolderManager
     dialogs_cache: Optional[hints.TotalList] = None
-    msg_cache: list[types.Message] = []
-    worker_routines: list[asyncio.Task] = []
+    worker_routines: list[asyncio.Task]
     qr_login: QRLogin | None = None
     login_task: asyncio.Task | None = None
     # rsa key
@@ -80,6 +79,7 @@ class TgFileSystemClient(object):
         )
         self.media_chunk_manager = MediaChunkHolderManager()
         self.db = db
+        self.worker_routines = []
 
     def __del__(self) -> None:
         if self.client.loop.is_running():
@@ -164,7 +164,6 @@ class TgFileSystemClient(object):
         if len(self.client_param.whitelist_chat) > 0:
             self._register_update_event(from_users=self.client_param.whitelist_chat)
             await self.task_queue.put((self._get_unique_task_id(), self._cache_whitelist_chat()))
-            # await self.task_queue.put((self._get_unique_task_id(), self._cache_whitelist_chat2()))
 
     async def stop(self) -> None:
         await self.client.loop.create_task(self._cancel_tasks())
@@ -181,35 +180,50 @@ class TgFileSystemClient(object):
                 logger.error(f"{err=}")
                 logger.error(traceback.format_exc())
 
-    async def _cache_whitelist_chat2(self):
-        for chat_id in self.client_param.whitelist_chat:
+    async def _cache_whitelist_chat_full_policy(self, chat_id: int, callback: Callable = None):
+        async for msg in self.client.iter_messages(chat_id):
+            if len(self.db.get_msg_by_unique_id(UserManager.generate_unique_id_by_msg(self.me, msg))) != 0:
+                continue
+            self.db.insert_by_message(self.me, msg)
+        if callback is not None:
+            callback()
+        logger.info(f"{chat_id} quit cache task.")
+
+    async def _cache_whitelist_chat_lazy_policy(self, chat_id: int, callback: Callable = None):
+        # update newest msg
+        newest_msg = self.db.get_newest_msg_by_chat_id(chat_id)
+        if len(newest_msg) > 0:
+            newest_msg = newest_msg[0]
             async for msg in self.client.iter_messages(chat_id):
-                if len(self.db.get_msg_by_unique_id(UserManager.generate_unique_id_by_msg(self.me, msg))) != 0:
-                    continue
+                if msg.id <= self.db.get_column_msg_id(newest_msg):
+                    break
                 self.db.insert_by_message(self.me, msg)
-            logger.info(f"{chat_id} quit cache task.")
+        # update oldest msg
+        oldest_msg = self.db.get_oldest_msg_by_chat_id(chat_id)
+        if len(oldest_msg) > 0:
+            oldest_msg = oldest_msg[0]
+            offset = self.db.get_column_msg_id(oldest_msg)
+            async for msg in self.client.iter_messages(chat_id, offset_id=offset):
+                self.db.insert_by_message(self.me, msg)
+        else:
+            async for msg in self.client.iter_messages(chat_id):
+                self.db.insert_by_message(self.me, msg)
+        if callback is not None:
+            callback()
+        logger.info(f"{chat_id} quit cache task.")
 
     async def _cache_whitelist_chat(self):
+        max_cache_tasks_num = TgFileSystemClient.MAX_WORKER_ROUTINE // 2
+        tasks_sem = asyncio.Semaphore(value=max_cache_tasks_num)
+
+        def _sem_release_callback():
+            tasks_sem.release()
+
         for chat_id in self.client_param.whitelist_chat:
-            # update newest msg
-            newest_msg = self.db.get_newest_msg_by_chat_id(chat_id)
-            if len(newest_msg) > 0:
-                newest_msg = newest_msg[0]
-                async for msg in self.client.iter_messages(chat_id):
-                    if msg.id <= self.db.get_column_msg_id(newest_msg):
-                        break
-                    self.db.insert_by_message(self.me, msg)
-            # update oldest msg
-            oldest_msg = self.db.get_oldest_msg_by_chat_id(chat_id)
-            if len(oldest_msg) > 0:
-                oldest_msg = oldest_msg[0]
-                offset = self.db.get_column_msg_id(oldest_msg)
-                async for msg in self.client.iter_messages(chat_id, offset_id=offset):
-                    self.db.insert_by_message(self.me, msg)
-            else:
-                async for msg in self.client.iter_messages(chat_id):
-                    self.db.insert_by_message(self.me, msg)
-            logger.info(f"{chat_id} quit cache task.")
+            await tasks_sem.acquire()
+            await self.task_queue.put(
+                (self._get_unique_task_id(), self._cache_whitelist_chat_lazy_policy(chat_id, callback=_sem_release_callback))
+            )
 
     @_acheck_before_call
     async def get_message(self, chat_id: int | str, msg_id: int) -> types.Message:
@@ -287,17 +301,15 @@ class TgFileSystemClient(object):
 
     async def get_messages_by_search_db(
         self,
-        chat_id: int,
+        chat_ids: list[int],
         search_word: str,
         limit: int = 10,
         offset: int = 0,
         inc: bool = False,
         ignore_case: bool = False,
     ) -> list[any]:
-        if chat_id not in self.client_param.whitelist_chat:
-            return []
         res = self.db.get_msg_by_chat_id_and_keyword(
-            chat_id,
+            chat_ids,
             search_word,
             limit=limit,
             offset=offset,
@@ -333,8 +345,6 @@ class TgFileSystemClient(object):
                 f"_download_media_chunk err:{err=},{offset=},{target_size=},{media_holder},\r\n{err=}\r\n{traceback.format_exc()}"
             )
         else:
-            if not media_holder.try_clear_waiter_and_requester():
-                logger.error("I think never run here.")
             if not self.media_chunk_manager.move_media_chunk_to_disk(media_holder):
                 logger.warning(f"move to disk failed, {media_holder=}")
             logger.debug(f"downloaded chunk:{offset=},{target_size=},{media_holder}")
diff --git a/backend/UserManager.py b/backend/UserManager.py
index d61239d..6807223 100644
--- a/backend/UserManager.py
+++ b/backend/UserManager.py
@@ -67,18 +67,26 @@ class UserManager(object):
 
     def get_msg_by_chat_id_and_keyword(
         self,
-        chat_id: int,
+        chat_ids: list[int],
         keyword: str,
         limit: int = 10,
         offset: int = 0,
         inc: bool = False,
         ignore_case: bool = False,
     ) -> list[any]:
+        if not chat_ids:
+            logger.warning(f"chat_ids is empty.")
+            return []
         keyword_condition = "msg_ctx LIKE '%{key}%' OR file_name LIKE '%{key}%'"
         if ignore_case:
             keyword_condition = "LOWER(msg_ctx) LIKE LOWER('%{key}%') OR LOWER(file_name) LIKE LOWER('%{key}%')"
         keyword_condition = keyword_condition.format(key=keyword)
-        execute_script = f"SELECT * FROM message WHERE chat_id = {chat_id} AND ({keyword_condition}) ORDER BY date_time {'' if inc else 'DESC '}LIMIT {limit} OFFSET {offset}"
+        chat_ids_str = ""
+        if len(chat_ids) > 1:
+            chat_ids_str = f"{tuple(chat_ids)}"
+        else:
+            chat_ids_str = f"({chat_ids[0]})"
+        execute_script = f"SELECT * FROM message WHERE chat_id in {chat_ids_str} AND ({keyword_condition}) ORDER BY date_time {'' if inc else 'DESC '}LIMIT {limit} OFFSET {offset}"
         logger.info(f"{execute_script=}")
         res = self.cur.execute(execute_script)
         return res
diff --git a/backend/api.py b/backend/api.py
index b17c879..f31a993 100644
--- a/backend/api.py
+++ b/backend/api.py
@@ -39,16 +39,18 @@ app.add_middleware(
     allow_headers=["*"],
 )
 
+
 class TgToFileListRequestBody(BaseModel):
     token: str
     search: str = ""
-    chat_id: int = 0
+    chat_ids: list[int] = []
     index: int = 0
     length: int = 10
     refresh: bool = False
     inner: bool = False
     inc: bool = False
 
+
 @app.post("/tg/api/v1/file/search")
 @apiutils.atimeit
 async def search_tg_file_list(body: TgToFileListRequestBody):
@@ -59,19 +61,21 @@ async def search_tg_file_list(body: TgToFileListRequestBody):
         res_type = "msg"
         client = await clients_mgr.get_client_force(body.token)
         res_dict = []
-        res = await client.get_messages_by_search_db(body.chat_id, body.search, limit=body.length, inc=body.inc, offset=body.index)
+        res = await client.get_messages_by_search_db(
+            body.chat_ids, body.search, limit=body.length, inc=body.inc, offset=body.index
+        )
         for item in res:
             msg_info = json.loads(item)
             file_name = apiutils.get_message_media_name_from_dict(msg_info)
             chat_id = apiutils.get_message_chat_id_from_dict(msg_info)
             msg_id = apiutils.get_message_msg_id_from_dict(msg_info)
-            msg_info['file_name'] = file_name
-            msg_info['download_url'] = f"{param.base.exposed_url}/tg/api/v1/file/get/{chat_id}/{msg_id}/{file_name}"
-            msg_info['src_tg_link'] = f"https://t.me/c/{chat_id}/{msg_id}"
+            msg_info["file_name"] = file_name
+            msg_info["download_url"] = f"{param.base.exposed_url}/tg/api/v1/file/get/{chat_id}/{msg_id}/{file_name}"
+            msg_info["src_tg_link"] = f"https://t.me/c/{chat_id}/{msg_id}"
             res_dict.append(msg_info)
 
         client_dict = json.loads(client.to_json())
-        client_dict['sign'] = body.token
+        client_dict["sign"] = body.token
 
         response_dict = {
             "client": client_dict,
@@ -95,7 +99,9 @@ async def get_tg_file_list(body: TgToFileListRequestBody):
         client = await clients_mgr.get_client_force(body.token)
         res_dict = []
         if body.search != "":
-            res = await client.get_messages_by_search(body.chat_id, search_word=body.search, limit=body.length, offset=body.index, inner_search=body.inner)
+            res = await client.get_messages_by_search(
+                body.chat_id, search_word=body.search, limit=body.length, offset=body.index, inner_search=body.inner
+            )
         else:
             res = await client.get_messages(body.chat_id, limit=body.length, offset=body.index)
         res_type = "msg"
@@ -104,8 +110,10 @@ async def get_tg_file_list(body: TgToFileListRequestBody):
             if file_name == "":
                 file_name = "unknown.tmp"
             msg_info = json.loads(item.to_json())
-            msg_info['file_name'] = file_name
-            msg_info['download_url'] = f"{param.base.exposed_url}/tg/api/v1/file/get/{body.chat_id}/{item.id}/{file_name}?sign={body.token}"
+            msg_info["file_name"] = file_name
+            msg_info["download_url"] = (
+                f"{param.base.exposed_url}/tg/api/v1/file/get/{body.chat_id}/{item.id}/{file_name}?sign={body.token}"
+            )
             res_dict.append(msg_info)
 
         response_dict = {
@@ -130,10 +138,7 @@ async def get_tg_file_media_stream(token: str, cid: int, mid: int, request: Requ
         "accept-ranges": "bytes",
         "content-encoding": "identity",
         # "content-length": stream_file_size,
-        "access-control-expose-headers": (
-            "content-type, accept-ranges, content-length, "
-            "content-range, content-encoding"
-        ),
+        "access-control-expose-headers": ("content-type, accept-ranges, content-length, " "content-range, content-encoding"),
     }
     range_header = request.headers.get("range")
     try:
@@ -151,8 +156,7 @@ async def get_tg_file_media_stream(token: str, cid: int, mid: int, request: Requ
         if file_name == "":
             maybe_file_type = mime_type.split("/")[-1]
             file_name = f"{chat_id}.{msg_id}.{maybe_file_type}"
-        headers[
-            "Content-Disposition"] = f"inline; filename*=utf-8'{quote(file_name)}'"
+        headers["Content-Disposition"] = f"inline; filename*=utf-8'{quote(file_name)}'"
 
         if range_header is not None:
             start, end = apiutils.get_range_header(range_header, file_size)
@@ -176,7 +180,7 @@ async def get_tg_file_media_stream(token: str, cid: int, mid: int, request: Requ
 
 @app.get("/tg/api/v1/file/get/{chat_id}/{msg_id}/{file_name}")
 @apiutils.atimeit
-async def get_tg_file_media(chat_id: int|str, msg_id: int, file_name: str, sign: str, req: Request):
+async def get_tg_file_media(chat_id: int | str, msg_id: int, file_name: str, sign: str, req: Request):
     try:
         if isinstance(chat_id, str):
             chat_id = int(chat_id)
@@ -224,6 +228,7 @@ class TgToChatListRequestBody(BaseModel):
     length: int = 0
     refresh: bool = False
 
+
 @app.post("/tg/api/v1/client/chat")
 @apiutils.atimeit
 async def get_tg_client_chat_list(body: TgToChatListRequestBody, request: Request):
@@ -235,8 +240,16 @@ async def get_tg_client_chat_list(body: TgToChatListRequestBody, request: Reques
         res_dict = {}
 
         res = await client.get_dialogs(limit=body.length, offset=body.index, refresh=body.refresh)
-        res_dict = [{"id": item.id, "is_channel": item.is_channel,
-                        "is_group": item.is_group, "is_user": item.is_user, "name": item.name, } for item in res]
+        res_dict = [
+            {
+                "id": item.id,
+                "is_channel": item.is_channel,
+                "is_group": item.is_group,
+                "is_user": item.is_user,
+                "name": item.name,
+            }
+            for item in res
+        ]
 
         response_dict = {
             "client": json.loads(client.to_json()),
@@ -249,6 +262,7 @@ async def get_tg_client_chat_list(body: TgToChatListRequestBody, request: Reques
         logger.error(f"{err=}")
         return Response(json.dumps({"detail": f"{err=}"}), status_code=status.HTTP_404_NOT_FOUND)
 
+
 if __name__ == "__main__":
     param = configParse.get_TgToFileSystemParameter()
     uvicorn.run(app, host="0.0.0.0", port=param.base.port)
diff --git a/backend/api_implement.py b/backend/api_implement.py
index 37cdcf2..4b50d7c 100644
--- a/backend/api_implement.py
+++ b/backend/api_implement.py
@@ -40,16 +40,20 @@ async def link_convert(link: str) -> str:
     return url
 
 
-async def get_clients_manager_status(detail) -> dict[str, any]:
-    clients_mgr = TgFileSystemClientManager.get_instance()
-    ret = await clients_mgr.get_status()
-    if not detail:
-        return ret
+async def get_chat_details(mgr: TgFileSystemClientManager) -> dict[int, any]:
     chat_details = {}
-    for _, client in clients_mgr.clients.items():
+    for _, client in mgr.clients.items():
         chat_list = client.client_param.whitelist_chat
         for chat_id in chat_list:
             chat_entity = await client.get_entity(chat_id)
             chat_details[chat_id] = json.loads(chat_entity.to_json())
-    ret["info"] = chat_details
+    return chat_details
+
+
+async def get_clients_manager_status(detail: bool) -> dict[str, any]:
+    clients_mgr = TgFileSystemClientManager.get_instance()
+    ret = await clients_mgr.get_status()
+    if not detail:
+        return ret
+    ret["clist"] = await get_chat_details(clients_mgr)
     return ret
diff --git a/configParse.py b/configParse.py
index 1d81ce8..e3d77aa 100644
--- a/configParse.py
+++ b/configParse.py
@@ -30,12 +30,11 @@ class TgToFileSystemParameter(BaseModel):
         addr: str = ""
         port: int = ""
     proxy: TgProxyParameter
-    
+
     class TgWebParameter(BaseModel):
         enable: bool = False
         token: str = ""
         port: int = 2000
-        chat_id: list[int] = []
     web: TgWebParameter
 
 @functools.lru_cache
diff --git a/frontend/remote_api.py b/frontend/remote_api.py
index cd55381..9fdc565 100644
--- a/frontend/remote_api.py
+++ b/frontend/remote_api.py
@@ -2,6 +2,7 @@ import sys
 import os
 import json
 import logging
+import traceback
 from urllib.parse import quote
 
 import requests
@@ -9,8 +10,6 @@ import requests
 sys.path.append(os.getcwd() + "/../")
 import configParse
 
-logger = logging.getLogger(__file__.split("/")[-1])
-
 param = configParse.get_TgToFileSystemParameter()
 
 background_server_url = f"{param.base.exposed_url}"
@@ -23,7 +22,7 @@ def login_client_by_qr_code_url() -> str:
     request_url = background_server_url + login_api_route
     response = requests.get(request_url)
     if response.status_code != 200:
-        logger.warning(f"Could not login, err:{response.status_code}, {response.content.decode('utf-8')}")
+        logging.warning(f"Could not login, err:{response.status_code}, {response.content.decode('utf-8')}")
         return None
     url_info = json.loads(response.content.decode("utf-8"))
     return url_info.get("url")
@@ -32,24 +31,33 @@ def login_client_by_qr_code_url() -> str:
 status_api_route = "/tg/api/v1/client/status"
 
 
-def get_backend_client_status() -> dict[str, any]:
-    request_url = background_server_url + status_api_route
+def get_backend_client_status(flag: bool = False) -> dict[str, any]:
+    request_url = f"{background_server_url}{status_api_route}?flag={flag}"
     response = requests.get(request_url)
     if response.status_code != 200:
-        logger.warning(f"get_status, backend is running? err:{response.status_code}, {response.content.decode('utf-8')}")
+        logging.warning(f"get_status, backend is running? err:{response.status_code}, {response.content.decode('utf-8')}")
         return None
     return json.loads(response.content.decode("utf-8"))
 
 
+def get_white_list_chat_dict() -> dict[str, any]:
+    backend_status = get_backend_client_status(True)
+    try:
+        return backend_status["clist"]
+    except Exception as err:
+        logging.warning(f"{err=},{traceback.format_exc()}")
+    return {}
+
+
 search_api_route = "/tg/api/v1/file/search"
 
 
-def search_database_by_keyword(keyword: str, offset: int, limit: int, is_order: bool) -> list[any] | None:
+def search_database_by_keyword(keyword: str, chat_list: list[int], offset: int, limit: int, is_order: bool) -> list[any] | None:
     request_url = background_server_url + search_api_route
     req_body = {
         "token": param.web.token,
         "search": keyword,
-        "chat_id": param.web.chat_id[0],
+        "chat_ids": chat_list,
         "index": offset,
         "length": limit,
         "refresh": False,
@@ -59,7 +67,7 @@ def search_database_by_keyword(keyword: str, offset: int, limit: int, is_order:
 
     response = requests.post(request_url, data=json.dumps(req_body))
     if response.status_code != 200:
-        logger.warning(f"search_database_by_keyword err:{response.status_code}, {response.content.decode('utf-8')}")
+        logging.warning(f"search_database_by_keyword err:{response.status_code}, {response.content.decode('utf-8')}")
         return None
     search_res = json.loads(response.content.decode("utf-8"))
     return search_res
@@ -73,7 +81,7 @@ def convert_tg_link_to_proxy_link(link: str) -> str:
     request_url = background_server_url + link_convert_api_route + f"?link={link}"
     response = requests.get(request_url)
     if response.status_code != 200:
-        logger.warning(f"link convert fail: {response.status_code}, {response.content.decode('utf-8')}")
+        logging.warning(f"link convert fail: {response.status_code}, {response.content.decode('utf-8')}")
         return ""
     response_js = json.loads(response.content.decode("utf-8"))
     return response_js["url"]
diff --git a/frontend/search.py b/frontend/search.py
index 275bae7..8acfc1b 100644
--- a/frontend/search.py
+++ b/frontend/search.py
@@ -1,40 +1,58 @@
 import sys
 import os
+import logging
+import traceback
 
 import streamlit as st
 
 import utils
 import remote_api as api
 
+
 @st.experimental_fragment
 def loop():
-    if 'page_index' not in st.session_state:
+    if "page_index" not in st.session_state:
         st.session_state.page_index = 1
-    if 'force_skip' not in st.session_state:
+    if "force_skip" not in st.session_state:
         st.session_state.force_skip = False
+    if "chat_select_list" not in st.session_state:
+        st.session_state.chat_select_list = []
 
-    if 'search_key' not in st.query_params:
+    if "search_key" not in st.query_params:
         st.query_params.search_key = ""
-    if 'is_order' not in st.query_params:
+    if "is_order" not in st.query_params:
         st.query_params.is_order = False
-    if 'search_res_limit' not in st.query_params:
+    if "search_res_limit" not in st.query_params:
         st.query_params.search_res_limit = "10"
 
     @st.experimental_fragment
     def search_container():
-        st.query_params.search_key = st.text_input("**搜索🔎**", value=st.query_params.search_key)
+        if "chat_dict" not in st.session_state:
+            st.session_state.chat_dict = api.get_white_list_chat_dict()
+        columns = st.columns([1, 1])
+        st.query_params.search_key = columns[0].text_input("**搜索🔎**", value=st.query_params.search_key)
+        chat_list = []
+        for _, chat_info in st.session_state.chat_dict.items():
+            chat_list.append(chat_info["title"])
+        st.session_state.chat_select_list = columns[1].multiselect("**Search in**", chat_list, default=chat_list)
+
         columns = st.columns([7, 1])
         with columns[0]:
-            st.query_params.search_res_limit = str(st.number_input(
-                "**每页结果**", min_value=1, max_value=100, value=int(st.query_params.search_res_limit), format="%d"))
+            st.query_params.search_res_limit = str(
+                st.number_input(
+                    "**每页结果**", min_value=1, max_value=100, value=int(st.query_params.search_res_limit), format="%d"
+                )
+            )
         with columns[1]:
             st.text("排序")
             st.query_params.is_order = st.toggle("顺序", value=utils.strtobool(st.query_params.is_order))
 
     search_container()
 
-    search_clicked = st.button('Search', type='primary', use_container_width=True)
-    if not st.session_state.force_skip and (not search_clicked or st.query_params.search_key == "" or st.query_params.search_key is None):
+    search_clicked = st.button("Search", type="primary", use_container_width=True)
+    if not st.session_state.force_skip and (
+        not search_clicked or st.query_params.search_key == "" or st.query_params.search_key is None
+    ):
         return
 
     if not st.session_state.force_skip:
@@ -48,56 +66,62 @@ def loop():
         offset_index = (st.session_state.page_index - 1) * search_limit
         is_order = utils.strtobool(st.query_params.is_order)
 
-        search_res = api.search_database_by_keyword(st.query_params.search_key, offset_index, search_limit, is_order)
+        status_bar = st.empty()
+        status_bar.status("Searching......")
+        search_chat_id_list = []
+        for chat_id, chat_info in st.session_state.chat_dict.items():
+            try:
+                if chat_info["title"] in st.session_state.chat_select_list:
+                    search_chat_id_list.append(int(chat_id))
+            except Exception as err:
+                logging.warning(f"{err=},{traceback.format_exc()}")
+        search_res = api.search_database_by_keyword(
+            st.query_params.search_key, search_chat_id_list, offset_index, search_limit, is_order
+        )
+        status_bar.empty()
         if search_res is None:
             return
 
         def page_switch_render():
-            columns = st.columns(3)
-            with columns[0]:
-                if st.button("Prev", use_container_width=True):
-                    st.session_state.page_index = st.session_state.page_index - 1
-                    st.session_state.page_index = max(
-                        st.session_state.page_index, 1)
-                    st.session_state.force_skip = True
-                    st.rerun()
-            with columns[1]:
-                # st.text(f"{st.session_state.page_index}")
-                st.markdown(
-                    f"<p style='text-align: center;'>{st.session_state.page_index}</p>", unsafe_allow_html=True)
-                # st.markdown(f"<input type='number' style='text-align: center;' value={st.session_state.page_index}>", unsafe_allow_html=True)
-            with columns[2]:
-                if st.button("Next", use_container_width=True):
-                    st.session_state.page_index = st.session_state.page_index + 1
-                    st.session_state.force_skip = True
-                    st.rerun()
-            st.session_state.page_index = st.number_input(
-                "page_index", key="page_index_input", min_value=1, max_value=100, value=st.session_state.page_index, format="%d", label_visibility="hidden")
+            page_index = st.number_input(
+                "Page number:",
+                key="page_index_input",
+                min_value=1,
+                max_value=100,
+                value=st.session_state.page_index,
+                format="%d",
+            )
+            if page_index != st.session_state.page_index:
+                st.session_state.page_index = page_index
+                st.session_state.force_skip = True
+                st.rerun()
 
-        def media_file_res_container(index: int, msg_ctx: str, file_name: str, file_size: int, url: str, src_link: str):
+        def media_file_res_container(
+            index: int, msg_ctx: str, file_name: str, file_size: int, url: str, src_link: str, mime_type: str
+        ):
             file_size_str = f"{file_size/1024/1024:.2f}MB"
             container = st.container()
             container_columns = container.columns([1, 99])
 
             st.session_state.search_res_select_list[index] = container_columns[0].checkbox(
-                "search_res_checkbox_" + str(index), label_visibility='collapsed')
+                "search_res_checkbox_" + str(index), label_visibility="collapsed"
+            )
 
             expender_title = f"{(msg_ctx if len(msg_ctx) < 103 else msg_ctx[:100] + '...')} &mdash; *{file_size_str}*"
             popover = container_columns[1].popover(expender_title, use_container_width=True)
             # media_file_popover_container(popover, url, msg_ctx, file_name, file_size_str, src_link)
             popover_columns = popover.columns([1, 3, 1])
             video_holder = popover_columns[0].empty()
-            if video_holder.button("Preview", key=f"videoBtn{url}", use_container_width=True):
+            if video_holder.button("Preview", key=f"videoBtn{url}{index}", use_container_width=True):
                 video_holder.empty()
-                if url:
-                    video_holder.video(url)
-                else:
-                    video_holder.video('./static/404.webm', format="video/webm")
-            popover_columns[1].markdown(f'{msg_ctx}')
-            popover_columns[1].markdown(f'**{file_name}**')
-            popover_columns[1].markdown(f'文件大小：*{file_size_str}*')
-            popover_columns[2].link_button('⬇️Download Link', url, use_container_width=True)
-            popover_columns[2].link_button('🔗Telegram Link', src_link, use_container_width=True)
+                p_url = url if url else "./static/404.webm"
+                mime_type = mime_type if mime_type else "video/webm"
+                video_holder.video(p_url, autoplay=True, format=mime_type)
+            popover_columns[1].markdown(f"{msg_ctx}")
+            popover_columns[1].markdown(f"**{file_name}**")
+            popover_columns[1].markdown(f"File Size: *{file_size_str}*")
+            popover_columns[2].link_button("⬇️Download Link", url, use_container_width=True)
+            popover_columns[2].link_button("🔗Telegram Link", src_link, use_container_width=True)
 
         @st.experimental_fragment
         def show_search_res(res: dict[str, any]):
@@ -108,48 +132,49 @@ def loop():
                 return
             sign_token = ""
             try:
-                sign_token = res['client']['sign']
+                sign_token = res["client"]["sign"]
             except Exception as err:
                 pass
             st.session_state.search_res_select_list = [False] * len(search_res_list)
             url_list = []
             for i in range(len(search_res_list)):
                 v = search_res_list[i]
-                msg_ctx= ""
+                msg_ctx = ""
                 file_name = None
                 file_size = 0
                 download_url = ""
                 src_link = ""
+                mime_type = ""
                 try:
-                    src_link = v['src_tg_link']
-                    msg_ctx = v['message']
-                    msg_id = str(v['id'])
-                    doc = v['media']['document']
-                    file_size = doc['size']
+                    src_link = v["src_tg_link"]
+                    msg_ctx = v["message"]
+                    msg_id = str(v["id"])
+                    doc = v["media"]["document"]
+                    mime_type = doc["mime_type"]
+                    file_size = doc["size"]
                     if doc is not None:
-                        for attr in doc['attributes']:
-                            file_name = attr.get('file_name')
+                        for attr in doc["attributes"]:
+                            file_name = attr.get("file_name")
                             if file_name != "" and file_name is not None:
                                 break
                     if file_name == "" or file_name is None:
                         file_name = "Can not get file name"
-                    download_url = v['download_url']
-                    download_url += f'?sign={sign_token}'
-                    url_list.append(download_url)
+                    download_url = v["download_url"]
+                    download_url += f"?sign={sign_token}"
                 except Exception as err:
                     msg_ctx = f"{err=}\r\n\r\n" + msg_ctx
-                media_file_res_container(
-                    i, msg_ctx, file_name, file_size, download_url, src_link)
+                    logging.warning(f"{err=},{traceback.format_exc()}")
+                url_list.append(download_url)
+                media_file_res_container(i, msg_ctx, file_name, file_size, download_url, src_link, mime_type)
             page_switch_render()
 
             show_text = ""
             select_list = st.session_state.search_res_select_list
             for i in range(len(select_list)):
                 if select_list[i]:
-                    show_text = show_text + url_list[i] + '\n'
-            st.text_area("链接", value=show_text)
+                    show_text = show_text + url_list[i] + "\n"
+            st.text_area("Links", value=show_text)
 
         show_search_res(search_res)
 
-
     do_search_req()