fix: process pdf once

app.py

@@ -28,7 +28,7 @@ def ocr_process(image, got_mode, ocr_color="", ocr_box=""):
                 res = model.chat_crop(tokenizer, image_path, ocr_type="ocr")
             else:
                 res = model.chat(tokenizer, image_path, ocr_type="ocr", ocr_box=ocr_box, ocr_color=ocr_color)
-            return gr.Markdown(res)
+            return res
         elif "format" in got_mode:
             if "multi-crop" in got_mode:
                 res = model.chat_crop(tokenizer, image_path, ocr_type="format", render=True, save_render_file=result_path)
@@ -36,17 +36,17 @@ def ocr_process(image, got_mode, ocr_color="", ocr_box=""):
                 res = model.chat(tokenizer, image_path, ocr_type="format", ocr_box=ocr_box, ocr_color=ocr_color, render=True, save_render_file=result_path)
 
             if os.path.exists(result_path):
-                with open(result_path, "r") as f:
+                with open(result_path, "r", encoding="utf-8") as f:
                     html_content = f.read()
                 encoded_html = base64.b64encode(html_content.encode("utf-8")).decode("utf-8")
-                iframe_src = f"data:text/html;base64,{encoded_html}"
-                iframe = f'<iframe src="{iframe_src}" width="100%" height="600px"></iframe>'
-                download_link = f'<a href="data:text/html;base64,{encoded_html}" download="result.html">下载完整结果</a>'
-                return gr.Markdown(f"{download_link}\n\n{iframe}")
+                data_uri = f"data:text/html;charset=utf-8;base64,{encoded_html}"
+                preview = f'<iframe src="{data_uri}" width="100%" height="600px"></iframe>'
+                download_link = f'<a href="{data_uri}" download="result.html">下载完整结果</a>'
+                return f"{download_link}\n\n{preview}\n\n识别结果：\n{res}"
 
-        return gr.Markdown("错误: 未知的OCR模式")
+        return "错误: 未知的OCR模式"
     except Exception as e:
-        return gr.Markdown(f"错误: {str(e)}")
+        return f"错误: {str(e)}"
 
 
 with gr.Blocks() as demo:
@@ -67,7 +67,7 @@ with gr.Blocks() as demo:
 
     submit_button = gr.Button("开始OCR识别")
 
-    output = gr.Markdown(label="识别结果")
+    output = gr.HTML(label="识别结果")
 
     submit_button.click(ocr_process, inputs=[image_input, got_mode, ocr_color, ocr_box], outputs=output)