Upload app.py

app.py

@@ -5,47 +5,74 @@ from pinecone import Pinecone
 from typing import List, Tuple
 import tempfile
 import shutil
+from dotenv import load_dotenv
+
+# Load environment variables
+load_dotenv()
+
+# Validate required environment variables
+required_env_vars = ["PINECONE_API_KEY"]
+missing_vars = [var for var in required_env_vars if not os.getenv(var)]
+
+if missing_vars:
+    raise ValueError(f"Missing required environment variables: {', '.join(missing_vars)}")
 
 # Initialize Pinecone
-pc = Pinecone(api_key="pcsk_4CboGg_BNMrddoKLGxfrzFLhequEQ7DmTCzT2BYXpiefUBHUKeLKXhbbmozifeVJiVWXrv")
+pinecone_api_key = os.getenv("PINECONE_API_KEY")
+pc = Pinecone(api_key=pinecone_api_key)
 
 # 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"""
+def process_files(files, *metadata_inputs):
+    """Process multiple files with individual metadata for each file"""
     if not files:
-        return "❌ Error: No files selected", "", ""
-    
-    if not sections.strip():
-        return "❌ Error: Sections field is required", "", ""
+        return "❌ Error: No files selected", ""
     
-    if not keywords.strip():
-        return "❌ Error: Keywords field is required", "", ""
-    
-    if not description.strip():
-        return "❌ Error: Description field is required", "", ""
+    if len(files) > 10:
+        return "❌ Error: Maximum 10 files allowed at a time", ""
     
     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")
+        assistant_name = os.getenv("PINECONE_ASSISTANT_NAME", "gstminutes")
+        assistant = pc.assistant.Assistant(assistant_name=assistant_name)
 
-        for file_path in files:
+        # Process each file with its individual metadata
+        for i, file_path in enumerate(files):
             try:
-                # Get filename from path
                 filename = os.path.basename(file_path)
                 
+                # Get metadata for this specific file (3 fields per file: sections, keywords, description)
+                sections_idx = i * 3
+                keywords_idx = i * 3 + 1
+                description_idx = i * 3 + 2
+                
+                if sections_idx < len(metadata_inputs):
+                    sections = metadata_inputs[sections_idx] or ""
+                    keywords = metadata_inputs[keywords_idx] or ""
+                    description = metadata_inputs[description_idx] or ""
+                else:
+                    sections = keywords = description = ""
+                
+                # Skip if no metadata provided for this file
+                if not sections.strip() and not keywords.strip() and not description.strip():
+                    errors.append({
+                        "filename": filename,
+                        "error": "❌ Error: No metadata provided"
+                    })
+                    continue
+                
+                # Prepare metadata for this file
+                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()
+                }
+                
                 # Copy to uploads directory
                 destination_path = os.path.join(UPLOAD_FOLDER, filename)
                 shutil.copy2(file_path, destination_path)
@@ -60,6 +87,7 @@ def process_files(files, sections, keywords, description):
                 results.append({
                     "filename": filename,
                     "status": "✅ Success",
+                    "metadata": metadata,
                     "response": str(response)
                 })
 
@@ -84,30 +112,51 @@ def process_files(files, sections, keywords, description):
         if results:
             detailed_results += "### ✅ **Successful Uploads:**\n"
             for result in results:
-                detailed_results += f"- **{result['filename']}** - {result['status']}\n"
-            detailed_results += "\n"
+                detailed_results += f"- **{result['filename']}**\n"
+                detailed_results += f"  - Sections: {', '.join(result['metadata']['sections'])}\n"
+                detailed_results += f"  - Keywords: {', '.join(result['metadata']['keywords'])}\n"
+                detailed_results += f"  - Description: {result['metadata']['description']}\n\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
+        return status_message, detailed_results
     
     except Exception as e:
         error_msg = f"❌ **Critical Error:** {str(e)}"
-        return error_msg, "", ""
+        return error_msg, ""
+
+def update_metadata_fields(files):
+    """Update metadata fields based on uploaded files"""
+    if not files:
+        return [gr.update(visible=False)] * 30  # Hide all fields
+    
+    if len(files) > 10:
+        # Show error and hide all fields
+        return [gr.update(visible=False)] * 30
+    
+    updates = []
+    for i in range(len(files)):
+        if i < len(files):
+            filename = os.path.basename(files[i])
+            # Show 3 fields per file (sections, keywords, description)
+            updates.extend([
+                gr.update(visible=True, label=f"📑 Sections for {filename}", placeholder="e.g., Introduction, Financial Data, Compliance"),
+                gr.update(visible=True, label=f"🔍 Keywords for {filename}", placeholder="e.g., GST, tax, compliance, revenue"),
+                gr.update(visible=True, label=f"📝 Description for {filename}", placeholder="Brief description of this document")
+            ])
+        
+    # Hide remaining fields
+    while len(updates) < 30:
+        updates.append(gr.update(visible=False))
+    
+    return updates[:30]
 
 def clear_form():
     """Clear all form fields"""
-    return None, "", "", "", "", "", ""
+    return [None] + [""] * 30 + ["", ""]
 
 # Create Gradio interface
 with gr.Blocks(
@@ -148,35 +197,28 @@ with gr.Blocks(
         with gr.Column(scale=1):
             gr.Markdown("### 📁 **File Upload**")
             files_input = gr.File(
-                label="Select Documents",
+                label="Select Documents (Max 10 files)",
                 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"
+            gr.Markdown("### 🏷️ **Document Metadata (Individual for Each File)**")
+            gr.Markdown("*Upload files first, then metadata fields will appear for each document*")
+    
+    # Dynamic metadata fields container
+    with gr.Column() as metadata_container:
+        # Create 30 text fields (enough for 10 files with 3 fields each)
+        metadata_fields = []
+        for i in range(30):
+            field = gr.Textbox(
+                label=f"Field {i}",
+                placeholder="",
+                visible=False,
+                lines=2
             )
+            metadata_fields.append(field)
     
     with gr.Row():
         with gr.Column(scale=1):
@@ -204,28 +246,30 @@ with gr.Blocks(
             )
     
     with gr.Row():
-        with gr.Column(scale=1):
+        with gr.Column():
             results_output = gr.Markdown(
                 label="📋 Detailed Results",
                 value=""
             )
-            
-        with gr.Column(scale=1):
-            metadata_output = gr.Markdown(
-                label="🏷️ Applied Metadata",
-                value=""
-            )
     
     # Event handlers
+    
+    # Update metadata fields when files are uploaded
+    files_input.change(
+        fn=update_metadata_fields,
+        inputs=[files_input],
+        outputs=metadata_fields
+    )
+    
     upload_btn.click(
         fn=process_files,
-        inputs=[files_input, sections_input, keywords_input, description_input],
-        outputs=[status_output, results_output, metadata_output]
+        inputs=[files_input] + metadata_fields,
+        outputs=[status_output, results_output]
     )
     
     clear_btn.click(
         fn=clear_form,
-        outputs=[files_input, sections_input, keywords_input, description_input, status_output, results_output, metadata_output]
+        outputs=[files_input] + metadata_fields + [status_output, results_output]
     )
     
     # Footer
@@ -234,9 +278,9 @@ with gr.Blocks(
         ---
         
         ### 💡 **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
+        - Select up to 10 PDF, DOC, DOCX, or TXT files at once
+        - Upload files first, then fill individual metadata for each document
+        - Each file gets its own sections, keywords, and description
         - Check the results section for upload status
         
         ### 📞 **Support:**