Upload 2 files

app.py

@@ -0,0 +1,254 @@
+import gradio as gr
+import os
+from pathlib import Path
+from pinecone import Pinecone
+from typing import List, Tuple
+import tempfile
+import shutil
+
+# Initialize Pinecone
+pc = Pinecone(api_key="pcsk_4CboGg_BNMrddoKLGxfrzFLhequEQ7DmTCzT2BYXpiefUBHUKeLKXhbbmozifeVJiVWXrv")
+
+# Create uploads directory
+UPLOAD_FOLDER = "uploads"
+os.makedirs(UPLOAD_FOLDER, exist_ok=True)
+
+def process_files(files, sections, keywords, description):
+    """Process multiple files and upload to Pinecone Assistant"""
+    if not files:
+        return "❌ Error: No files selected", "", ""
+    
+    if not sections.strip():
+        return "❌ Error: Sections field is required", "", ""
+    
+    if not keywords.strip():
+        return "❌ Error: Keywords field is required", "", ""
+    
+    if not description.strip():
+        return "❌ Error: Description field is required", "", ""
+    
+    try:
+        results = []
+        errors = []
+        
+        # Prepare metadata (same for all files)
+        metadata = {
+            "sections": [s.strip() for s in sections.split(",") if s.strip()],
+            "keywords": [k.strip() for k in keywords.split(",") if k.strip()],
+            "description": description.strip()
+        }
+
+        # Initialize Pinecone Assistant
+        assistant = pc.assistant.Assistant(assistant_name="gstminutes")
+
+        for file_path in files:
+            try:
+                # Get filename from path
+                filename = os.path.basename(file_path)
+                
+                # Copy to uploads directory
+                destination_path = os.path.join(UPLOAD_FOLDER, filename)
+                shutil.copy2(file_path, destination_path)
+
+                # Upload to Pinecone Assistant
+                response = assistant.upload_file(
+                    file_path=destination_path, 
+                    metadata=metadata, 
+                    timeout=None
+                )
+
+                results.append({
+                    "filename": filename,
+                    "status": "✅ Success",
+                    "response": str(response)
+                })
+
+            except Exception as file_error:
+                errors.append({
+                    "filename": os.path.basename(file_path),
+                    "error": f"❌ Error: {str(file_error)}"
+                })
+
+        # Format results for display
+        success_count = len(results)
+        error_count = len(errors)
+        
+        status_message = f"📊 **Processing Complete**\n\n"
+        status_message += f"✅ **Successful uploads:** {success_count}\n"
+        status_message += f"❌ **Failed uploads:** {error_count}\n"
+        status_message += f"📁 **Total files processed:** {len(files)}\n\n"
+        
+        # Detailed results
+        detailed_results = "## 📋 **Detailed Results**\n\n"
+        
+        if results:
+            detailed_results += "### ✅ **Successful Uploads:**\n"
+            for result in results:
+                detailed_results += f"- **{result['filename']}** - {result['status']}\n"
+            detailed_results += "\n"
+        
+        if errors:
+            detailed_results += "### ❌ **Failed Uploads:**\n"
+            for error in errors:
+                detailed_results += f"- **{error['filename']}** - {error['error']}\n"
+            detailed_results += "\n"
+        
+        # Metadata info
+        metadata_info = "## 🏷️ **Applied Metadata**\n\n"
+        metadata_info += f"**Sections:** {', '.join(metadata['sections'])}\n\n"
+        metadata_info += f"**Keywords:** {', '.join(metadata['keywords'])}\n\n"
+        metadata_info += f"**Description:** {metadata['description']}\n"
+        
+        return status_message, detailed_results, metadata_info
+    
+    except Exception as e:
+        error_msg = f"❌ **Critical Error:** {str(e)}"
+        return error_msg, "", ""
+
+def clear_form():
+    """Clear all form fields"""
+    return None, "", "", "", "", "", ""
+
+# Create Gradio interface
+with gr.Blocks(
+    title="📄 Tax Document Ingestion System",
+    theme=gr.themes.Soft(),
+    css="""
+    .gradio-container {
+        max-width: 1200px !important;
+        margin: auto;
+    }
+    .upload-container {
+        border: 2px dashed #4CAF50;
+        border-radius: 10px;
+        padding: 20px;
+        text-align: center;
+        background-color: #f8f9fa;
+    }
+    """
+) as app:
+    
+    gr.Markdown(
+        """
+        # 📄 Tax Document Ingestion System
+        
+        Upload multiple documents with metadata to the Pinecone Assistant for GST Minutes processing.
+        
+        ## 🚀 Features:
+        - ✅ **Multiple file upload** - Select and upload multiple documents at once
+        - 🏷️ **Metadata tagging** - Add sections, keywords, and descriptions
+        - 🔄 **Batch processing** - All files processed with the same metadata
+        - 📊 **Detailed reporting** - See success/failure status for each file
+        
+        ---
+        """
+    )
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            gr.Markdown("### 📁 **File Upload**")
+            files_input = gr.File(
+                label="Select Documents",
+                file_count="multiple",
+                file_types=[".pdf", ".doc", ".docx", ".txt"],
+                elem_classes=["upload-container"]
+            )
+            
+        with gr.Column(scale=1):
+            gr.Markdown("### 🏷️ **Document Metadata**")
+            
+            sections_input = gr.Textbox(
+                label="📑 Sections (comma-separated)",
+                placeholder="e.g., Introduction, Financial Data, Compliance, Summary",
+                lines=2,
+                info="Enter document sections separated by commas"
+            )
+            
+            keywords_input = gr.Textbox(
+                label="🔍 Keywords (comma-separated)",
+                placeholder="e.g., GST, tax, compliance, revenue, audit",
+                lines=2,
+                info="Enter relevant keywords separated by commas"
+            )
+            
+            description_input = gr.Textbox(
+                label="📝 Description",
+                placeholder="Brief description of the document(s) purpose and content",
+                lines=3,
+                info="Provide a clear description of the document content"
+            )
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            upload_btn = gr.Button(
+                "🚀 Upload Documents to Pinecone Assistant",
+                variant="primary",
+                size="lg"
+            )
+            
+        with gr.Column(scale=1):
+            clear_btn = gr.Button(
+                "🗑️ Clear Form",
+                variant="secondary",
+                size="lg"
+            )
+    
+    gr.Markdown("---")
+    
+    # Results section
+    with gr.Row():
+        with gr.Column():
+            status_output = gr.Markdown(
+                label="📊 Upload Status",
+                value="*Ready to upload documents...*"
+            )
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            results_output = gr.Markdown(
+                label="📋 Detailed Results",
+                value=""
+            )
+            
+        with gr.Column(scale=1):
+            metadata_output = gr.Markdown(
+                label="🏷️ Applied Metadata",
+                value=""
+            )
+    
+    # Event handlers
+    upload_btn.click(
+        fn=process_files,
+        inputs=[files_input, sections_input, keywords_input, description_input],
+        outputs=[status_output, results_output, metadata_output]
+    )
+    
+    clear_btn.click(
+        fn=clear_form,
+        outputs=[files_input, sections_input, keywords_input, description_input, status_output, results_output, metadata_output]
+    )
+    
+    # Footer
+    gr.Markdown(
+        """
+        ---
+        
+        ### 💡 **Usage Tips:**
+        - Select multiple PDF, DOC, DOCX, or TXT files
+        - Use descriptive sections and keywords for better organization
+        - All selected files will use the same metadata
+        - Check the results section for upload status
+        
+        ### 📞 **Support:**
+        For issues or questions, contact the development team.
+        """
+    )
+
+if __name__ == "__main__":
+    app.launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+        share=False,
+        debug=True,
+        show_error=True
+    )