diff --git a/pilot/embedding_engine/docx_loader.py b/pilot/embedding_engine/docx_loader.py index 50d1a6b45..458254b66 100644 --- a/pilot/embedding_engine/docx_loader.py +++ b/pilot/embedding_engine/docx_loader.py @@ -22,5 +22,7 @@ class DocxLoader(BaseLoader): para = doc.paragraphs[i] text = para.text content.append(text) - docs.append(Document(page_content=''.join(content), metadata={"source": self.file_path})) + docs.append( + Document(page_content="".join(content), metadata={"source": self.file_path}) + ) return docs diff --git a/pilot/embedding_engine/ppt_loader.py b/pilot/embedding_engine/ppt_loader.py index b3d970291..04f5e5c93 100644 --- a/pilot/embedding_engine/ppt_loader.py +++ b/pilot/embedding_engine/ppt_loader.py @@ -20,5 +20,9 @@ class PPTLoader(BaseLoader): for slide in pr.slides: for shape in slide.shapes: if hasattr(shape, "text") and shape.text is not "": - docs.append(Document(page_content=shape.text, metadata={"source": slide.slide_id})) + docs.append( + Document( + page_content=shape.text, metadata={"source": slide.slide_id} + ) + ) return docs