fix: use UTC consistently across all datetime operations to resolve timezone display issues

admin/server/auth.py

@@ -179,8 +179,8 @@ def login_admin(email: str, password: str):
     resp = user.to_json()
     user.access_token = get_uuid()
     login_user(user)
-    user.update_time = (current_timestamp(),)
-    user.update_date = (datetime_format(datetime.now()),)
+    user.update_time = current_timestamp()
+    user.update_date = datetime_format(datetime.utcnow())
     user.last_login_time = get_format_time()
     user.save()
     msg = "Welcome back!"

admin/server/routes.py

@@ -504,7 +504,7 @@ def generate_user_api_key(username: str) -> tuple[Response, int]:
             "token": key,
             "beta": generate_confirmation_token().replace("ragflow-", "")[:32],
             "create_time": current_timestamp(),
-            "create_date": datetime_format(datetime.now()),
+            "create_date": datetime_format(datetime.utcnow()),
             "update_time": None,
             "update_date": None,
         }

api/apps/api_app.py

@@ -31,12 +31,12 @@ def stats():
             tenants[0].tenant_id,
             request.args.get(
                 "from_date",
-                (datetime.now() -
+                (datetime.utcnow() -
                  timedelta(
                      days=7)).strftime("%Y-%m-%d 00:00:00")),
             request.args.get(
                 "to_date",
-                datetime.now().strftime("%Y-%m-%d %H:%M:%S")),
+                datetime.utcnow().strftime("%Y-%m-%d %H:%M:%S")),
             "agent" if "canvas_id" in request.args else None)
 
         res = {"pv": [], "uv": [], "speed": [], "tokens": [], "round": [], "thumb_up": []}

api/apps/chunk_app.py

@@ -325,8 +325,9 @@ async def create():
     if not isinstance(d["question_kwd"], list):
         return get_data_error_result(message="`question_kwd` is required to be a list")
     d["question_tks"] = rag_tokenizer.tokenize("\n".join(d["question_kwd"]))
-    d["create_time"] = str(datetime.datetime.now()).replace("T", " ")[:19]
-    d["create_timestamp_flt"] = datetime.datetime.now().timestamp()
+    now_utc = datetime.datetime.now(datetime.timezone.utc)
+    d["create_time"] = now_utc.strftime("%Y-%m-%d %H:%M:%S")
+    d["create_timestamp_flt"] = now_utc.timestamp()
     if "tag_kwd" in req:
         if not isinstance(req["tag_kwd"], list):
             return get_data_error_result(message="`tag_kwd` is required to be a list")

api/apps/restful_apis/system_api.py

@@ -143,7 +143,7 @@ def new_token():
             "token": generate_confirmation_token(),
             "beta": generate_confirmation_token().replace("ragflow-", "")[:32],
             "create_time": current_timestamp(),
-            "create_date": datetime_format(datetime.now()),
+            "create_date": datetime_format(datetime.utcnow()),
             "update_time": None,
             "update_date": None,
         }

api/apps/sdk/doc.py

@@ -952,8 +952,9 @@ async def add_chunk(tenant_id, dataset_id, document_id):
     d["important_tks"] = rag_tokenizer.tokenize(" ".join(req.get("important_keywords", [])))
     d["question_kwd"] = [str(q).strip() for q in req.get("questions", []) if str(q).strip()]
     d["question_tks"] = rag_tokenizer.tokenize("\n".join(req.get("questions", [])))
-    d["create_time"] = str(datetime.datetime.now()).replace("T", " ")[:19]
-    d["create_timestamp_flt"] = datetime.datetime.now().timestamp()
+    now_utc = datetime.datetime.now(datetime.timezone.utc)
+    d["create_time"] = now_utc.strftime("%Y-%m-%d %H:%M:%S")
+    d["create_timestamp_flt"] = now_utc.timestamp()
     d["kb_id"] = dataset_id
     d["docnm_kwd"] = doc.name
     d["doc_id"] = document_id

api/apps/system_app.py

@@ -14,7 +14,7 @@
 #  limitations under the License
 #
 import logging
-from datetime import datetime
+from datetime import datetime, timezone
 import json
 
 from api.apps import login_required
@@ -127,7 +127,7 @@ def status():
     task_executor_heartbeats = {}
     try:
         task_executors = REDIS_CONN.smembers("TASKEXE")
-        now = datetime.now().timestamp()
+        now = datetime.now(timezone.utc).timestamp()
         for task_executor_id in task_executors:
             heartbeats = REDIS_CONN.zrangebyscore(task_executor_id, now - 60 * 30, now)
             heartbeats = [json.loads(heartbeat) for heartbeat in heartbeats]

api/apps/user_app.py

@@ -127,7 +127,7 @@ async def login():
         user.access_token = get_uuid()
         login_user(user)
         user.update_time = current_timestamp()
-        user.update_date = datetime_format(datetime.now())
+        user.update_date = datetime_format(datetime.utcnow())
         user.save()
         msg = "Welcome back!"
 

api/db/services/api_service.py

@@ -30,7 +30,7 @@ class APITokenService(CommonService):
     def used(cls, token):
         return cls.model.update({
             "update_time": current_timestamp(),
-            "update_date": datetime_format(datetime.now()),
+            "update_date": datetime_format(datetime.utcnow()),
         }).where(
             cls.model.token == token
         )

api/db/services/common_service.py

@@ -207,7 +207,7 @@ def insert(cls, **kwargs):
         if "id" not in kwargs:
             kwargs["id"] = get_uuid()
         timestamp = current_timestamp()
-        cur_datetime = datetime_format(datetime.now())
+        cur_datetime = datetime_format(datetime.utcnow())
         kwargs["create_time"] = timestamp
         kwargs["create_date"] = cur_datetime
         kwargs["update_time"] = timestamp
@@ -228,7 +228,7 @@ def insert_many(cls, data_list, batch_size=100):
             batch_size (int, optional): Number of records to insert in each batch. Defaults to 100.
         """
         current_ts = current_timestamp()
-        current_datetime = datetime_format(datetime.now())
+        current_datetime = datetime_format(datetime.utcnow())
         with DB.atomic():
             for d in data_list:
                 d["create_time"] = current_ts
@@ -253,7 +253,7 @@ def update_many_by_id(cls, data_list):
         """
 
         timestamp = current_timestamp()
-        cur_datetime = datetime_format(datetime.now())
+        cur_datetime = datetime_format(datetime.utcnow())
         for data in data_list:
             data["update_time"] = timestamp
             data["update_date"] = cur_datetime
@@ -272,7 +272,7 @@ def update_by_id(cls, pid, data):
         # Returns:
         #     Number of records updated
         data["update_time"] = current_timestamp()
-        data["update_date"] = datetime_format(datetime.now())
+        data["update_date"] = datetime_format(datetime.utcnow())
         num = cls.model.update(data).where(cls.model.id == pid).execute()
         return num
 

api/db/services/connector_service.py

@@ -198,7 +198,7 @@ def list_sync_tasks(cls, connector_id=None, page_number=None, items_per_page=15)
 
     @classmethod
     def start(cls, id, connector_id):
-        cls.update_by_id(id, {"status": TaskStatus.RUNNING, "time_started": datetime.now().strftime('%Y-%m-%d %H:%M:%S') })
+        cls.update_by_id(id, {"status": TaskStatus.RUNNING, "time_started": datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S') })
         ConnectorService.update_by_id(connector_id, {"status": TaskStatus.RUNNING})
 
     @classmethod

api/db/services/dialog_service.py

@@ -103,7 +103,7 @@ def update_many_by_id(cls, data_list):
         with DB.atomic():
             for data in data_list:
                 data["update_time"] = current_timestamp()
-                data["update_date"] = datetime_format(datetime.now())
+                data["update_date"] = datetime_format(datetime.utcnow())
                 cls.model.update(data).where(cls.model.id == data["id"]).execute()
 
     @classmethod

api/db/services/document_service.py

@@ -881,13 +881,14 @@ def _sync_progress(cls, docs: list[dict]):
                 # only for special task and parsed docs and unfinished
                 freeze_progress = special_task_running and doc_progress >= 1 and not finished
                 msg = "\n".join(sorted(msg))
+                now_utc = datetime.utcnow()
                 begin_at = d.get("process_begin_at")
                 if not begin_at:
-                    begin_at = datetime.now()
+                    begin_at = now_utc
                     # fallback
                     cls.update_by_id(d["id"], {"process_begin_at": begin_at})
 
-                info = {"process_duration": max(datetime.timestamp(datetime.now()) - begin_at.timestamp(), 0), "run": status}
+                info = {"process_duration": max((now_utc - begin_at).total_seconds(), 0), "run": status}
                 if prg != 0 and not freeze_progress:
                     info["progress"] = prg
                 if msg:
@@ -997,14 +998,15 @@ def queue_raptor_o_graphrag_tasks(sample_doc, ty, priority, fake_doc_id="", doc_
         hasher.update(str(chunking_config[field]).encode("utf-8"))
 
     def new_task():
+        now_utc = datetime.utcnow()
         return {
             "id": get_uuid(),
             "doc_id": fake_doc_id,
             "from_page": 100000000,
             "to_page": 100000000,
             "task_type": ty,
-            "progress_msg": datetime.now().strftime("%H:%M:%S") + " created task " + ty,
-            "begin_at": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
+            "progress_msg": now_utc.strftime("%H:%M:%S") + " created task " + ty,
+            "begin_at": now_utc.strftime("%Y-%m-%d %H:%M:%S"),
         }
 
     task = new_task()
@@ -1079,8 +1081,9 @@ def dummy(prog=None, msg=""):
             d = deepcopy(doc)
             d.update(ck)
             d["id"] = xxhash.xxh64((ck["content_with_weight"] + str(d["doc_id"])).encode("utf-8")).hexdigest()
-            d["create_time"] = str(datetime.now()).replace("T", " ")[:19]
-            d["create_timestamp_flt"] = datetime.now().timestamp()
+            now_utc = datetime.utcnow()
+            d["create_time"] = now_utc.strftime("%Y-%m-%d %H:%M:%S")
+            d["create_timestamp_flt"] = now_utc.timestamp()
             if not d.get("image"):
                 docs.append(d)
                 continue

api/db/services/evaluation_service.py

@@ -284,7 +284,7 @@ def start_evaluation(cls, dataset_id: str, dialog_id: str,
             # Create evaluation run
             run_id = get_uuid()
             if not name:
-                name = f"Evaluation Run {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}"
+                name = f"Evaluation Run {datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S')} UTC"
 
             run = {
                 "id": run_id,

api/db/services/file2document_service.py

@@ -74,8 +74,8 @@ def delete_by_document_id(cls, doc_id):
     @DB.connection_context()
     def update_by_file_id(cls, file_id, obj):
         obj["update_time"] = current_timestamp()
-        obj["update_date"] = datetime_format(datetime.now())
-        cls.model.update(obj).where(cls.model.id == file_id).execute()
+        obj["update_date"] = datetime_format(datetime.utcnow())
+        cls.model.update(obj).where(cls.model.file_id == file_id).execute()
         return File2Document(**obj)
 
     @classmethod

api/db/services/knowledgebase_service.py

@@ -526,7 +526,7 @@ def get_kb_by_name(cls, kb_name, user_id):
     def atomic_increase_doc_num_by_id(cls, kb_id):
         data = {}
         data["update_time"] = current_timestamp()
-        data["update_date"] = datetime_format(datetime.now())
+        data["update_date"] = datetime_format(datetime.utcnow())
         data["doc_num"] = cls.model.doc_num + 1
         num = cls.model.update(data).where(cls.model.id == kb_id).execute()
         return num
@@ -568,7 +568,7 @@ def decrease_document_num_in_delete(cls, kb_id, doc_num_info: dict):
             'chunk_num': kb_row.chunk_num - doc_num_info['chunk_num'],
             'token_num': kb_row.token_num - doc_num_info['token_num'],
             'update_time': current_timestamp(),
-            'update_date': datetime_format(datetime.now())
+            'update_date': datetime_format(datetime.utcnow())
         }
         return cls.model.update(update_dict).where(cls.model.id == kb_id).execute()
 

api/db/services/langfuse_service.py

@@ -59,13 +59,13 @@ def delete_ty_tenant_id(cls, tenant_id):
     @classmethod
     def update_by_tenant(cls, tenant_id, langfuse_keys):
         langfuse_keys["update_time"] = current_timestamp()
-        langfuse_keys["update_date"] = datetime_format(datetime.now())
+        langfuse_keys["update_date"] = datetime_format(datetime.utcnow())
         return cls.model.update(**langfuse_keys).where(cls.model.tenant_id == tenant_id).execute()
 
     @classmethod
     def save(cls, **kwargs):
         current_ts = current_timestamp()
-        current_date = datetime_format(datetime.now())
+        current_date = datetime_format(datetime.utcnow())
 
         kwargs["create_time"] = current_ts
         kwargs["create_date"] = current_date

api/db/services/pipeline_operation_log_service.py

@@ -189,7 +189,7 @@ def create(cls, document_id, pipeline_id, task_type, task_id=None, referred_docu
             avatar=avatar,
         )
         timestamp = current_timestamp()
-        datetime_now = datetime_format(datetime.now())
+        datetime_now = datetime_format(datetime.utcnow())
         log["create_time"] = timestamp
         log["create_date"] = datetime_now
         log["update_time"] = timestamp

api/db/services/search_service.py

@@ -29,7 +29,7 @@ class SearchService(CommonService):
     @classmethod
     def save(cls, **kwargs):
         current_ts = current_timestamp()
-        current_date = datetime_format(datetime.now())
+        current_date = datetime_format(datetime.utcnow())
 
         kwargs["create_time"] = current_ts
         kwargs["create_date"] = current_date

api/db/services/system_settings_service.py

@@ -33,7 +33,7 @@ def get_by_name(cls, name):
     @DB.connection_context()
     def update_by_name(cls, name, obj):
         obj["update_time"] = current_timestamp()
-        obj["update_date"] = datetime_format(datetime.now())
+        obj["update_date"] = datetime_format(datetime.utcnow())
         cls.model.update(obj).where(cls.model.name == name).execute()
         return SystemSettings(**obj)
 

api/db/services/task_service.py

@@ -124,7 +124,7 @@ def get_task(cls, task_id, doc_ids=[]):
         if not docs:
             return None
 
-        msg = f"\n{datetime.now().strftime('%H:%M:%S')} Task has been received."
+        msg = f"\n{datetime.utcnow().strftime('%H:%M:%S')} Task has been received."
         prog = random.random() / 10.0
         if docs[0]["retry_count"] >= 3:
             msg = "\nERROR: Task is abandoned after 3 times attempts."
@@ -342,7 +342,10 @@ def update_progress(cls, id, info):
                         ((prog == -1) | (prog > cls.model.progress))))
                     ).execute()
 
-        process_duration = (datetime.now() - task.begin_at).total_seconds()
+        if task.begin_at:
+            process_duration = max((datetime.utcnow() - task.begin_at).total_seconds(), 0)
+        else:
+            process_duration = 0
         cls.model.update(process_duration=process_duration).where(cls.model.id == id).execute()
 
     @classmethod
@@ -380,7 +383,7 @@ def new_task():
             "progress": 0.0,
             "from_page": 0,
             "to_page": 100000000,
-            "begin_at": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
+            "begin_at": datetime.utcnow().strftime("%Y-%m-%d %H:%M:%S"),
         }
 
     parse_task_array = []
@@ -500,7 +503,7 @@ def reuse_prev_task_chunks(task: dict, prev_tasks: list[dict], chunking_config:
     else:
         task["progress_msg"] = ""
     task["progress_msg"] = " ".join(
-        [datetime.now().strftime("%H:%M:%S"), task["progress_msg"], "Reused previous task's chunks."])
+        [datetime.utcnow().strftime("%H:%M:%S"), task["progress_msg"], "Reused previous task's chunks."])
     prev_task["chunk_ids"] = ""
 
     return len(task["chunk_ids"].split())
@@ -533,7 +536,7 @@ def queue_dataflow(tenant_id:str, flow_id:str, task_id:str, doc_id:str=CANVAS_DE
         to_page=100000000,
         task_type="dataflow" if not rerun else "dataflow_rerun",
         priority=priority,
-        begin_at= datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
+        begin_at= datetime.utcnow().strftime("%Y-%m-%d %H:%M:%S"),
     )
     if doc_id not in [CANVAS_DEBUG_DOC_ID, GRAPH_RAPTOR_FAKE_DOC_ID]:
         TaskService.model.delete().where(TaskService.model.doc_id == doc_id).execute()

api/db/services/user_service.py

@@ -117,7 +117,7 @@ def save(cls, **kwargs):
                 str(kwargs["password"]))
 
         current_ts = current_timestamp()
-        current_date = datetime_format(datetime.now())
+        current_date = datetime_format(datetime.utcnow())
 
         kwargs["create_time"] = current_ts
         kwargs["create_date"] = current_date
@@ -139,7 +139,7 @@ def update_user(cls, user_id, user_dict):
         with DB.atomic():
             if user_dict:
                 user_dict["update_time"] = current_timestamp()
-                user_dict["update_date"] = datetime_format(datetime.now())
+                user_dict["update_date"] = datetime_format(datetime.utcnow())
                 cls.model.update(user_dict).where(
                     cls.model.id == user_id).execute()
 
@@ -150,7 +150,7 @@ def update_user_password(cls, user_id, new_password):
             update_dict = {
                 "password": generate_password_hash(str(new_password)),
                 "update_time": current_timestamp(),
-                "update_date": datetime_format(datetime.now())
+                "update_date": datetime_format(datetime.utcnow())
             }
             cls.model.update(update_dict).where(cls.model.id == user_id).execute()
 

api/utils/health_utils.py

@@ -13,7 +13,7 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-from datetime import datetime
+from datetime import datetime, timezone
 import json
 import os
 import requests
@@ -309,7 +309,7 @@ def check_task_executor_alive():
     task_executor_heartbeats = {}
     try:
         task_executors = REDIS_CONN.smembers("TASKEXE")
-        now = datetime.now().timestamp()
+        now = datetime.now(timezone.utc).timestamp()
         for task_executor_id in task_executors:
             heartbeats = REDIS_CONN.zrangebyscore(task_executor_id, now - 60 * 30, now)
             heartbeats = [json.loads(heartbeat) for heartbeat in heartbeats]