update pdf_loader.py

loader/pdf_loader.py

@@ -15,13 +15,11 @@ class UnstructuredPaddlePDFLoader(UnstructuredFileLoader):
             full_dir_path = os.path.join(os.path.dirname(filepath), dir_path)
             if not os.path.exists(full_dir_path):
                 os.makedirs(full_dir_path)
-            filename = os.path.split(filepath)[-1]
             ocr = PaddleOCR(lang="ch", use_gpu=False, show_log=False)
             doc = fitz.open(filepath)
-            txt_file_path = os.path.join(full_dir_path, "%s.txt" % (filename))
+            txt_file_path = os.path.join(full_dir_path, f"{os.path.split(filepath)[-1]}.txt")
             img_name = os.path.join(full_dir_path, 'tmp.png')
             with open(txt_file_path, 'w', encoding='utf-8') as fout:
-
                 for i in range(doc.page_count):
                     page = doc[i]
                     text = page.get_text("")