update pdf_loader.py

2023-05-31 18:03:37 +08:00 · 2023-05-31 18:03:37 +08:00 · 5c0c1eed93
parent 99ee2e9fd8
commit 5c0c1eed93
1 changed files with 2 additions and 1 deletions
--- a/loader/pdf_loader.py
+++ b/loader/pdf_loader.py
@ -39,7 +39,8 @@ class UnstructuredPaddlePDFLoader(UnstructuredFileLoader):
                        result = ocr.ocr(img_name)
                        ocr_result = [i[1][0] for line in result for i in line]
                        fout.write("\n".join(ocr_result))
-            os.remove(img_name)
+            if os.path.exists(img_name):
+                os.remove(img_name)
            return txt_file_path

        txt_file_path = pdf_ocr_txt(self.file_path)