fix

app.py

@@ -11,6 +11,7 @@ import base64
 import logging
 import sys
 import shutil
+from huggingface_hub import HfApi
 
 # Configure logging
 # Simplified logging for cloud deployment
@@ -35,13 +36,27 @@ if not gemini_api_key:
 genai.configure(api_key=gemini_api_key)
 logger.info("Gemini API configured successfully")
 
+# Configure Hugging Face Hub API for file persistence
+space_repo_id = "polygraf-ai/business_card_extractor"  # Your Space repo
+hf_token = os.getenv("HF_TOKEN")  # Optional - only needed for Hub uploads
+
+if hf_token:
+    hf_api = HfApi(token=hf_token)
+    logger.info("Hugging Face Hub API configured for file uploads")
+    logger.info(f"Space repo: {space_repo_id}")
+else:
+    hf_api = None
+    logger.info("HF_TOKEN not found - files will be saved to persistent storage only")
+    logger.info("To enable Hub uploads: Set HF_TOKEN in Space secrets (optional)")
+
 # Create output directories
 logger.info("Setting up output directories")
-# Use absolute paths to avoid ambiguity in different environments
-output_dir = Path("business_card_exports").resolve()
-images_dir = Path("business_cards").resolve()
-output_dir.mkdir(exist_ok=True)
-images_dir.mkdir(exist_ok=True)
+# Use /data for persistent storage in Hugging Face Spaces
+BASE_PATH = Path("/data")
+output_dir = BASE_PATH / "business_card_exports"
+images_dir = BASE_PATH / "business_cards"
+output_dir.mkdir(parents=True, exist_ok=True)
+images_dir.mkdir(parents=True, exist_ok=True)
 logger.info(f"Export directory created/verified: {output_dir}")
 logger.info(f"Images directory created/verified: {images_dir}")
 
@@ -51,6 +66,25 @@ logger.info(f"Working directory: {os.getcwd()}")
 logger.info(f"Export directory: {output_dir.absolute()}")
 logger.info(f"Images directory: {images_dir.absolute()}")
 
+def upload_to_hub(file_path, path_in_repo, commit_message="Update files"):
+    """Upload a file to the Hugging Face Hub repository"""
+    if hf_api:  # Only need API token, repo ID is hardcoded
+        try:
+            logger.info(f"Uploading {file_path} to Hub at {path_in_repo}")
+            hf_api.upload_file(
+                path_or_fileobj=str(file_path),
+                path_in_repo=path_in_repo,
+                repo_id=space_repo_id,
+                repo_type="space",
+                commit_message=commit_message
+            )
+            logger.info(f"Successfully uploaded {file_path} to Hub")
+            return True
+        except Exception as e:
+            logger.error(f"Failed to upload {file_path} to Hub: {e}")
+            return False
+    return False
+
 def extract_business_card_data_batch(images, filenames, model_name="gemini-2.5-flash"):
     """Extract data from multiple business card images in a single API call"""
     
@@ -276,6 +310,13 @@ def process_business_cards(images, model_name="gemini-2.5-flash", save_images=Tr
                 saved_image_paths.append(str(image_path))
                 logger.debug(f"Saved image {i+1}: {unique_filename}")
                 
+                # Upload image to Hub
+                upload_to_hub(
+                    image_path,
+                    f"images/{unique_filename}",
+                    f"Add business card image {unique_filename}"
+                )
+                
             except Exception as e:
                 logger.error(f"Failed to save image {filename}: {e}")
         
@@ -450,6 +491,13 @@ def process_business_cards(images, model_name="gemini-2.5-flash", save_images=Tr
         
         logger.info(f"Current run Excel file saved successfully: {current_filename}")
         
+        # Upload current run file to Hub
+        upload_to_hub(
+            current_filename, 
+            f"exports/{current_filename.name}",
+            f"Add current run export {current_filename.name}"
+        )
+        
         # Create a manifest file for the current run directory
         manifest_filename = output_dir / f"current_run_{timestamp}_manifest.txt"
         try:
@@ -465,7 +513,7 @@ def process_business_cards(images, model_name="gemini-2.5-flash", save_images=Tr
                 f.write(f"Save Images: {save_images}\n")
                 f.write(f"Processing Date: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}\n")
                 f.write(f"Session ID: {timestamp}\n")
-                f.write(f"Session Summary: {summary}\n") # Use the summary variable from the previous run
+                f.write(f"Errors: {len(errors) if 'errors' in locals() else 0}\n")
             logger.info(f"Manifest file created: {manifest_filename}")
         except Exception as e:
             logger.error(f"Failed to create manifest file: {e}")
@@ -496,6 +544,13 @@ def process_business_cards(images, model_name="gemini-2.5-flash", save_images=Tr
         
         logger.info(f"Cumulative Excel file saved successfully: {cumulative_filename}")
         
+        # Upload cumulative file to Hub
+        upload_to_hub(
+            cumulative_filename, 
+            f"exports/{cumulative_filename.name}",
+            f"Update cumulative database - {len(cumulative_df)} total cards"
+        )
+        
         # Create a manifest file for the cumulative directory
         manifest_filename = output_dir / f"all_business_cards_total_{timestamp}_manifest.txt"
         try:
@@ -511,7 +566,7 @@ def process_business_cards(images, model_name="gemini-2.5-flash", save_images=Tr
                 f.write(f"Save Images: {save_images}\n")
                 f.write(f"Processing Date: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}\n")
                 f.write(f"Session ID: {timestamp}\n")
-                f.write(f"Session Summary: {summary}\n") # Use the summary variable from the previous run
+                f.write(f"Errors: {len(errors) if 'errors' in locals() else 0}\n")
             logger.info(f"Manifest file created: {manifest_filename}")
         except Exception as e:
             logger.error(f"Failed to create manifest file: {e}")
@@ -645,15 +700,20 @@ with gr.Blocks(title="Business Card Data Extractor") as demo:
         - 📁 **Current Run**: Contains only the cards you just processed
         - 📊 **Total Database**: Contains ALL cards ever processed (cumulative)
         
-        **Image Storage:**
-        - 💾 **Optional**: Save uploaded images to business_cards folder
-        - 📝 **Tracking**: Image file paths included in Excel database
+        **🔄 Persistent Storage & Hub Integration:**
+        - 💾 Files saved to `/data` directory (persistent across restarts)
+        - 🌐 Automatically uploaded to Hub repository (if configured)
+        - 📂 Excel files: `/data/business_card_exports/`
+        - 🖼️ Images: `/data/business_cards/` (if save option enabled)
+        
+        **📌 File Access:**
+        - Download directly from interface or check Space's Files tab
+        - Hub uploads available at: `exports/` and `images/` folders in repo
+        - Check `FOLDER_CONTENTS.txt` for complete file listing
         
-        **📌 Finding Your Files in Hugging Face Spaces:**
-        - All files are saved in the `business_card_exports` and `business_cards` folders
-        - Check `FOLDER_CONTENTS.txt` in the Files tab to see a complete list of all files
-        - Excel files are in `business_card_exports/`
-        - Images are in `business_cards/` (if save option enabled)
+        **⚙️ Optional Hub Integration:**
+        - Set `HF_TOKEN` in Space secrets to enable file uploads to repository
+        - Files will be uploaded to `polygraf-ai/business_card_extractor`
         """
     )
     
@@ -712,7 +772,8 @@ with gr.Blocks(title="Business Card Data Extractor") as demo:
         - 📄 **Data Extraction**: Names, emails, phone numbers, addresses, and more
         - 📞 **Smart Combination**: Multiple emails/phones combined with commas
         - 🏠 **Address Merging**: All phone types and address fields combined
-        - 💾 **Image Storage**: Optionally save images to business_cards folder
+        - 💾 **Persistent Storage**: Files saved to `/data` directory (survives restarts)
+        - 🌐 **Hub Integration**: Auto-upload to repository (optional)
         - 📊 **Dual Output**: Current run + cumulative database files
         - 📝 **Full Tracking**: Processing date, filename, image path, and AI model used
         - 🎯 **One Row Per Card**: Each business card becomes one spreadsheet row

env.example

@@ -6,5 +6,11 @@
 # For Hugging Face Spaces: Add this as a Repository Secret named "Gemini_API"
 Gemini_API=your_gemini_api_key_here
 
-# Example:
-# Gemini_API=AIzaSyBxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 
+# Hugging Face Hub Integration (Optional - for file uploads to repository)
+# Get your token from: https://huggingface.co/settings/tokens
+# For Hugging Face Spaces: Add this as a Repository Secret
+HF_TOKEN=your_huggingface_token_here
+
+# Examples:
+# Gemini_API=AIzaSyBxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+# HF_TOKEN=hf_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 

requirements.txt

@@ -3,4 +3,5 @@ pydantic==2.10.6
 google-generativeai==0.8.0
 pandas==2.1.4
 openpyxl==3.1.2
-Pillow==10.2.0
+Pillow==10.2.0
+huggingface_hub>=0.19.0