Update app.py

app.py

@@ -8,10 +8,16 @@ from typhoon_ocr import prepare_ocr_messages
 import gradio as gr
 from PIL import Image
 
+# โหลด environment variables จาก .env
 load_dotenv()
 
-openai = OpenAI(base_url=os.getenv("TYPHOON_BASE_URL"), api_key=os.getenv("TYPHOON_API_KEY"))
+# ตั้งค่า OpenAI (ใช้ API ของ Typhoon OCR)
+openai = OpenAI(
+    base_url=os.getenv("TYPHOON_BASE_URL"), 
+    api_key=os.getenv("TYPHOON_API_KEY")
+)
 
+# ตั้งค่า Theme (ใช้ของเดิม)
 theme = gr.themes.Soft(
     primary_hue=gr.themes.Color(
         c50="#f7f7fd",
@@ -30,6 +36,20 @@ theme = gr.themes.Soft(
     neutral_hue="stone",
 )
 
+# ตัวแปรสำหรับบันทึกผลลัพธ์
+OUTPUT_FILE = "ocr_results.txt"
+
+def save_ocr_result(text):
+    """บันทึกผลลัพธ์ OCR แบบต่อเนื่องในไฟล์เดียว พร้อมเว้น 2 บรรทัดระหว่างข้อมูล"""
+    with open(OUTPUT_FILE, "a", encoding="utf-8") as f:
+        f.write(text + "\n\n")
+    return OUTPUT_FILE
+
+def clear_output_file():
+    """ล้างไฟล์ผลลัพธ์เก่า (เรียกครั้งเดียวเมื่อเริ่มใช้งานใหม่)"""
+    with open(OUTPUT_FILE, "w", encoding="utf-8") as f:
+        f.write("")
+
 def process_pdf(pdf_or_image_file, task_type, page_number):
     if pdf_or_image_file is None:
         return None, "No file uploaded"
@@ -37,7 +57,7 @@ def process_pdf(pdf_or_image_file, task_type, page_number):
     orig_filename = pdf_or_image_file.name
     
     try:
-        # Use the new simplified function to prepare OCR messages with page number
+        # ใช้ prepare_ocr_messages ตามเดิม
         messages = prepare_ocr_messages(
             pdf_or_image_path=orig_filename,
             task_type=task_type,
@@ -46,12 +66,12 @@ def process_pdf(pdf_or_image_file, task_type, page_number):
             page_num=page_number if page_number else 1
         )
         
-        # Extract the image from the message content for display
+        # ดึงภาพจากผลลัพธ์
         image_url = messages[0]["content"][1]["image_url"]["url"]
         image_base64 = image_url.replace("data:image/png;base64,", "")
         image_pil = Image.open(BytesIO(base64.b64decode(image_base64)))
         
-        # Send messages to OpenAI compatible API
+        # ส่งไป API
         response = openai.chat.completions.create(
             model=os.getenv("TYPHOON_OCR_MODEL"),
             messages=messages,
@@ -64,20 +84,22 @@ def process_pdf(pdf_or_image_file, task_type, page_number):
         )
         text_output = response.choices[0].message.content
         
-        # Try to parse the output assuming it is a Python dictionary containing 'natural_text'
+        # ดึง natural_text
         try:
             json_data = json.loads(text_output)
             markdown_out = json_data.get('natural_text', "").replace("<figure>", "").replace("</figure>", "")
         except Exception as e:
             markdown_out = f"⚠️ Could not extract `natural_text` from output.\nError: {str(e)}"
         
-        return image_pil, markdown_out
+        # บันทึกผลลัพธ์ต่อเนื่องในไฟล์
+        save_ocr_result(markdown_out)
+        
+        return image_pil, markdown_out, gr.File.update(value=OUTPUT_FILE)
     
     except Exception as e:
-        return None, f"Error processing file: {str(e)}"
-
+        return None, f"Error processing file: {str(e)}", None
 
-# Build the Gradio UI.
+# สร้าง UI
 with gr.Blocks(theme=theme) as demo:
     title = gr.HTML("""
     <h1>Typhoon OCR</h1>
@@ -87,7 +109,7 @@ with gr.Blocks(theme=theme) as demo:
     <br />
     <details>
         <summary><strong>Disclaimer</strong></summary>
-        The responses generated by this Artificial Intelligence (AI) system are autonomously constructed and do not necessarily reflect the views or positions of the developing organizations, their affiliates, or any of their employees. These AI-generated responses do not represent those of the organizations. The organizations do not endorse, support, sanction, encourage, verify, or agree with the comments, opinions, or statements generated by this AI. The information produced by this AI is not intended to malign any religion, ethnic group, club, organization, company, individual, anyone, or anything. It is not the intent of the organizations to malign any group or individual. The AI operates based on its programming and training data and its responses should not be interpreted as the explicit intent or opinion of the organizations.
+        The responses generated by this AI system are autonomously constructed and do not necessarily reflect the views or positions of the developing organizations, their affiliates, or any of their employees. These AI-generated responses do not represent those of the organizations. The organizations do not endorse, support, sanction, encourage, verify, or agree with the comments, opinions, or statements generated by this AI. The information produced by this AI is not intended to malign any religion, ethnic group, club, organization, company, individual, anyone, or anything. It is not the intent of the organizations to malign any group or individual. The AI operates based on its programming and training data and its responses should not be interpreted as the explicit intent or opinion of the organizations.
     </details>
     <br />
     <details>
@@ -99,12 +121,14 @@ with gr.Blocks(theme=theme) as demo:
         <summary><strong>License</strong></summary>
         This project utilizes certain datasets and checkpoints that are subject to their respective original licenses. Users must comply with all terms and conditions of these original licenses. The content of this project itself is licensed under the Apache license 2.0.
     </details>
-""")
+    """)
+    
     with gr.Row():
         with gr.Column(scale=1):
-            # Update file_types to accept PDF as well as common image formats.
+            # อัปโหลดไฟล์
             pdf_input = gr.File(label="📄 Upload Image file or PDF file", file_types=[".pdf", ".png", ".jpg", ".jpeg"])
             
+            # เลือก Task
             with gr.Group(elem_classes=["task-background"]):
                 task_dropdown = gr.Radio(["default", "structure"], label="🎯 Select Task", value="default")
                 gr.HTML("""
@@ -115,7 +139,6 @@ with gr.Blocks(theme=theme) as demo:
                 demo.css = """
                 .task-background {
                     background: var(--block-background-fill) !important;
-                    
                 }
                 .task-background > * {
                     background: var(--block-background-fill) !important;
@@ -125,19 +148,31 @@ with gr.Blocks(theme=theme) as demo:
                     font-size: 12px;
                 }
                 """
+            # เลือกเพจ
             page_number = gr.Number(label="📄 Page Number (for PDFs only)", value=1, minimum=1, step=1)
+            
+            # ปุ่มรัน
             run_button = gr.Button("🚀 Run")
+            
+            # แสดงภาพ
             image_output = gr.Image(label="📸 Preview Image", type="pil")
+        
         with gr.Column(scale=2):
+            # แสดงผลลัพธ์ Markdown
             markdown_output = gr.Markdown(label='Markdown Result', show_label=True)
+            
+            # ปุ่มดาวน์โหลดไฟล์
+            download_button = gr.File(label="📥 ดาวน์โหลดผลลัพธ์ทั้งหมด (Text File)", interactive=False)
 
-    
-    # Connect the UI inputs to the processing function.
+    # เชื่อมต่อ UI กับฟังก์ชัน
     run_button.click(
         fn=process_pdf,
         inputs=[pdf_input, task_dropdown, page_number],
-        outputs=[image_output, markdown_output]
+        outputs=[image_output, markdown_output, download_button]
     )
 
-# Launch the Gradio demo (temporary public share for 72 hours)
-demo.launch(share=False)
+# เรียกครั้งเดียวเมื่อเริ่มเพื่อล้างไฟล์เก่า
+clear_output_file()
+
+# รันแอป
+demo.launch(share=False)