feat: Add Gradio progress bar updates to the `run_inference` function for better user feedback.

app.py

@@ -171,7 +171,7 @@ def _escape_inequalities_in_math(md: str) -> str:
 
 # --- Inference Logic ---
 
-def run_inference(img_path, task_type="ocr"):
+def run_inference(img_path, task_type="ocr", progress=gr.Progress()):
     if not PADDLE_AVAILABLE:
         return "❌ Paddle backend not installed.", "", "", ""
     
@@ -183,6 +183,7 @@ def run_inference(img_path, task_type="ocr"):
 
     try:
         logger.info(f"--- Inference Start: {task_type} ---")
+        progress(0, desc="📦 Initializing inference engine...")
         output = pipeline.predict(input=img_path)
         logger.info(f"Output object type: {type(output)}")
         logger.info(f"Output object: {output}")
@@ -196,26 +197,25 @@ def run_inference(img_path, task_type="ocr"):
         os.makedirs(run_output_dir, exist_ok=True)
 
         logger.info(f"will iterate")
-        for res in output:
-            logger.info(f"Output object: {res}")
-            res.print() 
+        progress(0.2, desc="🔍 Parsing document structure...")
     
         for i, res in enumerate(output):
             logger.info(f"Processing segment {i+1}...")
+            progress((i + 1) / 5, desc=f"✍️ Recognizing content (segment {i+1})...")
+            
             # Save results
             res.save_to_json(save_path=run_output_dir)
             res.save_to_markdown(save_path=run_output_dir)
             res.print()
             
-            # Read back generated files from this segment's save
-            # Paddle naming: res_{i}.md, res_{i}.json, etc.
+            # Read back generated files
             fnames = os.listdir(run_output_dir)
             for fname in fnames:
                 fpath = os.path.join(run_output_dir, fname)
                 if fname.endswith(".md"):
                     with open(fpath, 'r', encoding='utf-8') as f:
                         content = f.read()
-                        if content not in md_content: # Avoid duplicates if listdir is messy
+                        if content not in md_content:
                             md_content += content + "\n\n"
                 elif fname.endswith(".json"):
                     with open(fpath, 'r', encoding='utf-8') as f:
@@ -227,12 +227,15 @@ def run_inference(img_path, task_type="ocr"):
                     vis_html += f'<div style="margin-bottom:20px; border: 2px solid #10b981; border-radius: 12px; overflow: hidden; background:white;">'
                     vis_html += f'<img src="{vis_src}" alt="Vis {i+1}" style="width:100%;">'
                     vis_html += f'</div>'
+            
+            logger.info(f"Finished processing segment {i+1}")
 
         if not md_content:
             md_content = "⚠️ Finished but no content was recognized."
         
         md_preview = _escape_inequalities_in_math(md_content)
         logger.info("--- Inference Finished Successfully ---")
+        progress(1.0, desc="✅ Recovery complete")
         return md_preview, md_content, vis_html, json_content
 
     except Exception as e: