Update modules/index_func.py

modules/index_func.py

@@ -36,9 +36,12 @@ def get_documents(file_src):
                 except:
                     pdftext = ""
                     with open(filepath, "rb") as pdfFileObj:
-                        pdfReader = PyPDF2.PdfReader(pdfFileObj)
+                        pdfReader = PyPDF2.PdfReader(pdfFileObj)    #pdfReader.pages 有多少頁
                         for page in tqdm(pdfReader.pages):
-                            pdftext += page.extract_text()
+                            pdftmp = page.extract_text()          #每頁的文字加起來
+                            logging.info(f"pdftmp：{pdftmp}")
+                            pdftext += pdftmp
+                            
                 texts = [Document(page_content=pdftext,
                                   metadata={"source": filepath})]
             elif file_type == ".docx":