DataBeak

#!/usr/bin/env python3 """Example demonstrating CSV Editor's null value support and Claude Code compatibility. This example shows how CSV Editor handles the exact scenario that was failing before: - JSON null values in row insertion - Claude Code's JSON string serialization - Comprehensive null value support """ import asyncio import uuid import pandas as pd from databeak.models.csv_session import get_session_manager # type: ignore[import-not-found] from databeak.tools.transformations import ( # type: ignore[import-not-found] get_row_data, insert_row, ) async def main() -> None: """Demonstrate null value handling and Claude Code compatibility.""" print("CSV Editor - Null Value Support & Claude Code Compatibility Demo") print("=" * 65) # Create a session with job application tracking data session_manager = get_session_manager() session_id = str(uuid.uuid4()) session = session_manager.get_or_create_session(session_id) # Create initial data structure initial_data = pd.DataFrame( { "Company": ["TechCorp"], "Position": ["Software Engineer"], "Location": ["San Francisco"], "Application Date": ["2025-01-15"], "Status": ["Applied"], "Job Posting URL": ["https://techcorp.com/jobs/123"], "Source": ["LinkedIn"], "Contact Person": ["Jane Smith"], "Contact Email": ["jane@techcorp.com"], "Follow-up Date": ["2025-01-20"], "Interview Date": [None], # Initially null "Response Date": [None], # Initially null "Outcome": [None], # Initially null "Strategic Advantage": ["Python expertise"], "Key Requirements Match": ["High"], "Salary Range": ["120-150k"], "Notes": ["Applied through referral"], }, ) session.load_data(initial_data) print(f"✅ Created session: {session_id}") print(f"📊 Initial data shape: {initial_data.shape}") # Example 1: The exact case that was failing before print("\n🎯 Example 1: Insert row with null values (Claude Code format)") # This is exactly what the user was trying to do: job_application_data = { "Company": "Finergo", "Position": "TBD", "Location": "TBD", "Application Date": "2025-09-03", "Status": "Screening Interview Completed", "Job Posting URL": None, # NULL VALUE "Source": "TBD", "Contact Person": None, # NULL VALUE "Contact Email": None, # NULL VALUE "Follow-up Date": "2025-09-06", "Interview Date": "2025-09-03", "Response Date": None, # NULL VALUE "Outcome": None, # NULL VALUE "Strategic Advantage": "TBD", "Key Requirements Match": "TBD", "Salary Range": "TBD", "Notes": "Screening interview completed. Follow-up on Friday regarding salary conversations.", } # Method 1: Direct dictionary (works) result = await insert_row(session_id, -1, job_application_data) print(f"✅ Dictionary insert: {result['success']}") # Method 2: JSON string (as Claude Code sends) - NOW WORKS! import json json_string = json.dumps(job_application_data) result = await insert_row(session_id, -1, json_string) print(f"✅ JSON string insert: {result['success']}") # Verify the null values were preserved correctly row_data = await get_row_data(session_id, 2) # Last inserted row print("\n📋 Null value verification:") null_fields = [ "Job Posting URL", "Contact Person", "Contact Email", "Response Date", "Outcome", ] for field in null_fields: value = row_data["data"][field] is_null = value is None print(f" {field}: {value} (null: {is_null})") print(f"\n📈 Final data shape: {session.df.shape}") print("🎉 All null values preserved correctly!") print("\n" + "=" * 65) print("✅ CSV Editor now fully supports:") print(" • JSON null values in all operations") print(" • Claude Code's JSON string serialization") print(" • Type-safe data operations") print(" • Modular, maintainable architecture") if __name__ == "__main__": asyncio.run(main())