ignore more

api/core/indexing_runner.py

@@ -672,8 +672,7 @@ class IndexingRunner:
 
         if extra_update_params:
             update_params.update(extra_update_params)
-        # type: ignore
+        db.session.query(DatasetDocument).filter_by(id=document_id).update(update_params)  # type: ignore
-        db.session.query(DatasetDocument).filter_by(id=document_id).update(update_params)
         db.session.commit()
 
     @staticmethod

api/core/rag/extractor/notion_extractor.py

@@ -331,10 +331,10 @@ class NotionExtractor(BaseExtractor):
         last_edited_time = self.get_notion_last_edited_time()
         data_source_info = document_model.data_source_info_dict
         data_source_info["last_edited_time"] = last_edited_time
-        # type: ignore
+
         db.session.query(DocumentModel).filter_by(id=document_model.id).update(
             {DocumentModel.data_source_info: json.dumps(data_source_info)}
-        )
+        )  # type: ignore
         db.session.commit()
 
     def get_notion_last_edited_time(self) -> str:

api/models/dataset.py

@@ -660,10 +660,10 @@ class DocumentSegment(Base):
     dataset_id = mapped_column(StringUUID, nullable=False)
     document_id = mapped_column(StringUUID, nullable=False)
     position: Mapped[int]
-    content = mapped_column(db.Text, nullable=False)
+    content: Mapped[str]
-    answer = mapped_column(db.Text, nullable=True)
+    answer: Mapped[Optional[str]]
-    word_count = mapped_column(db.Integer, nullable=False)
+    word_count: Mapped[int]
-    tokens = mapped_column(db.Integer, nullable=False)
+    tokens: Mapped[int]
 
     # indexing fields
     keywords = mapped_column(db.JSON, nullable=True)

api/services/dataset_service.py

@@ -214,11 +214,9 @@ class DatasetService:
         dataset.created_by = account.id
         dataset.updated_by = account.id
         dataset.tenant_id = tenant_id
-        # type: ignore
+        dataset.embedding_model_provider = embedding_model.provider if embedding_model else None  # type: ignore
-        dataset.embedding_model_provider = embedding_model.provider if embedding_model else None
+        dataset.embedding_model = embedding_model.model if embedding_model else None  # type: ignore
-        # type: ignore
+        dataset.retrieval_model = retrieval_model.model_dump() if retrieval_model else None  # type: ignore
-        dataset.embedding_model = embedding_model.model if embedding_model else None
-        dataset.retrieval_model = retrieval_model.model_dump() if retrieval_model else None
         dataset.permission = permission or DatasetPermissionEnum.ONLY_ME
         dataset.provider = provider
         db.session.add(dataset)
@@ -1541,9 +1539,10 @@ class DocumentService:
         db.session.add(document)
         db.session.commit()
         # update document segment
-        # type: ignore
+
-        update_params = {DocumentSegment.status: "re_segment"}
+        db.session.query(DocumentSegment).filter_by(document_id=document.id).update(
-        db.session.query(DocumentSegment).filter_by(document_id=document.id).update(update_params)
+            {DocumentSegment.status: "re_segment"}
+        )  # type: ignore
         db.session.commit()
         # trigger async task
         document_indexing_update_task.delay(document.dataset_id, document.id)
@@ -2228,7 +2227,7 @@ class SegmentService:
                     # calc embedding use tokens
                     if document.doc_form == "qa_model":
                         segment.answer = args.answer
-                        tokens = embedding_model.get_text_embedding_num_tokens(texts=[content + segment.answer])[0]
+                        tokens = embedding_model.get_text_embedding_num_tokens(texts=[content + segment.answer])[0]  # type: ignore
                     else:
                         tokens = embedding_model.get_text_embedding_num_tokens(texts=[content])[0]
                 segment.content = content