FIX:1598 Use PageTextSplitter for DatasourceKnowledge (#1599)

Co-authored-by: shenk-b <[email protected]>
Co-authored-by: aries_ckt <[email protected]>

dbgpt/rag/knowledge/datasource.py

@@ -52,6 +52,7 @@ def support_chunk_strategy(cls) -> List[ChunkStrategy]:
         return [
             ChunkStrategy.CHUNK_BY_SIZE,
             ChunkStrategy.CHUNK_BY_SEPARATOR,
+            ChunkStrategy.CHUNK_BY_PAGE,
         ]
 
     @classmethod
@@ -63,3 +64,12 @@ def type(cls) -> KnowledgeType:
     def document_type(cls) -> DocumentType:
         """Return document type."""
         return DocumentType.DATASOURCE
+
+    @classmethod
+    def default_chunk_strategy(cls) -> ChunkStrategy:
+        """Return default chunk strategy.
+
+        Returns:
+            ChunkStrategy: default chunk strategy
+        """
+        return ChunkStrategy.CHUNK_BY_PAGE

dbgpt/rag/summary/db_summary_client.py

@@ -101,8 +101,10 @@ def init_db_profile(self, db_summary_client, dbname):
             from dbgpt.rag.assembler.db_schema import DBSchemaAssembler
 
             db_assembler = DBSchemaAssembler.load_from_connection(
-                connector=db_summary_client.db, vector_store_connector=vector_connector
+                connector=db_summary_client.db,
+                vector_store_connector=vector_connector,
             )
+
             if len(db_assembler.get_chunks()) > 0:
                 db_assembler.persist()
         else: