update

src/demo/asg_loader.py

@@ -161,6 +161,13 @@ class DocumentLoading:
         return extracted_data['abstract'] + extracted_data['introduction'] + extracted_data['main_content']
 
     def load_pdf(self, pdf_file, survey_id, mode):
+        # 确保 mode 合法
+        valid_modes = ['auto', 'txt', 'ocr']
+        mineru_mode = mode if mode in valid_modes else 'auto'
+        if mode not in valid_modes:
+            print(f"Warning: unsupported mineru method '{mode}', defaulting to 'auto'.")
+            # 同时修正用于生成路径的 mode
+            mode = mineru_mode
         base_name = os.path.splitext(os.path.basename(pdf_file))[0]
         target_dir = os.path.join(get_path('md', survey_id), base_name)
         md_file_path = os.path.join(target_dir, mode, f"{base_name}.md")
@@ -170,7 +177,7 @@ class DocumentLoading:
             print(f"Markdown file for {pdf_file} already exists at {md_file_path}. Skipping conversion.", flush=True)
             return self.process_md_file(md_file_path, survey_id)
             
-        command = ["mineru", "-p", pdf_file, "-o", get_path('md', survey_id), "-m", mode]
+        command = ["mineru", "-p", pdf_file, "-o", get_path('md', survey_id), "-m", mineru_mode]
         try:
             subprocess.run(command, check=True)
             # 检查是否生成了 Markdown 文件

src/demo/views.py

@@ -313,6 +313,11 @@ def PosRank_get_top5_ngrams(input_pd):
 
 def process_file(file_name, survey_id, mode):
     global embedder
+    # 校验 mode，仅允许 auto/txt/ocr
+    valid_modes = ['auto', 'txt', 'ocr']
+    if mode not in valid_modes:
+        print(f"Warning: unsupported mode '{mode}', defaulting to 'auto'.")
+        mode = 'auto'
     # 如果 file_name 不是绝对路径，则拼接 MEDIA_ROOT
     abs_file_path = file_name if os.path.isabs(file_name) else os.path.join(settings.MEDIA_ROOT, file_name)
     result = process_pdf(abs_file_path, survey_id, embedder, mode)