superset-toolkit 0.2.2

Professional toolkit for Superset API automation - datasets, charts, and dashboards

🚀 Superset Toolkit

Production-grade SDK for Apache Superset API automation with professional patterns.

✨ Key Features

• 🎯 Client-Centric Design: No more repetitive (session, base_url) parameters
• 👤 Username-Aware Operations: Work with usernames directly, no manual ID resolution
• 🔧 JWT-Based Authentication: Robust user ID extraction from tokens (works for any user)
• 🚀 Composite Workflows: Complete operations in single function calls
• 📦 Batch Operations: Efficient bulk chart/dashboard management
• 🧹 Resource Lifecycle: Context managers with automatic cleanup
• 🛡️ Professional Error Handling: Graceful fallbacks and detailed logging
• 📊 Full Chart Support: Table, pie, histogram, area charts with username support

Installation

Basic Installation

pip install -e .

With CLI Support

pip install -e ".[cli]"

Development Installation

pip install -e ".[dev]"

🚀 Quick Start

Professional Client-Centric Usage (Recommended)

from superset_toolkit import SupersetClient
from superset_toolkit.config import Config

# Configure connection
config = Config(
    superset_url="https://your-superset-instance.com",
    username="your-username",
    password="your-secure-password",
    schema="your-schema",
    database_name="your-database"
)

# Context manager with automatic cleanup
with SupersetClient(config=config) as client:
    
    # Create chart (all complexity hidden)
    chart_id = client.create_table_chart(
        name="Sales Report",
        table="sales_data",
        owner="analyst"  # Just username - no ID resolution needed!
    )
    
    # Create dashboard with automatic chart linking
    dashboard_id = client.create_dashboard(
        title="Sales Dashboard", 
        slug="sales-dashboard",
        owner="analyst",
        charts=["Sales Report"]  # Auto-links existing charts
    )
    
    # Query resources by owner
    user_charts = client.get_charts(owner="analyst")
    user_dashboards = client.get_dashboards(owner="analyst")
    
    # Get comprehensive user summary
    summary = client.get_user_summary("analyst")
    print(f"User has: {summary['summary']}")
    
    # Clean up everything for a user
    client.cleanup_user("temp_user", dry_run=False)

Batch & Composite Operations

# Create dashboard with multiple charts in one operation
result = client.create_dashboard_with_charts(
    dashboard_title="Analytics Dashboard",
    slug="analytics-dash",
    chart_configs=[
        {"name": "Sales Chart", "table": "sales", "columns": ["region", "amount"]},
        {"name": "Revenue Chart", "table": "revenue", "columns": ["month", "total"]}
    ],
    owner="analytics_team"
)

# Create multiple charts efficiently  
chart_ids = client.create_charts_batch([
    {"name": "Chart 1", "table": "data1"},
    {"name": "Chart 2", "table": "data2"},
    {"name": "Chart 3", "table": "data3"}
], owner="data_team")

Enhanced Standalone Functions (For Advanced Use Cases)

from superset_toolkit.charts import create_table_chart
from superset_toolkit.queries import get_charts_by_username

client = SupersetClient()

# Enhanced standalone functions now support username parameter
chart_id = create_table_chart(
    client.session, client.base_url, 
    "Advanced Chart", dataset_id,
    username="data_analyst"  # No manual user ID resolution!
)

# Direct function calls for specific operations
charts = get_charts_by_username(client.session, client.base_url, "data_analyst")

Environment Variables (Optional)

export SUPERSET_URL="https://your-superset-instance.com"
export SUPERSET_USERNAME="your-username" 
export SUPERSET_PASSWORD="your-password"
export SUPERSET_SCHEMA="your_schema"  # Optional, defaults to 'reports'
export SUPERSET_DATABASE_NAME="YourDatabase"  # Optional, defaults to 'Trino'

Module Organization:

• client.py: Enhanced SupersetClient with professional methods
• auth.py: JWT-based authentication with permission-aware fallbacks
• charts.py: Username-aware chart creation (table, pie, histogram, area)
• dashboard.py: Dashboard creation with automatic chart linking
• queries.py: Resource filtering and querying by owner/dataset
• datasets.py: Dataset management with permission handling

📊 Advanced Usage Examples

Multiple Chart Types with Username Support

with SupersetClient() as client:
    # Table chart
    table_chart = client.create_chart_from_table(
        chart_name="Sales Data",
        table="sales",
        owner="analyst",
        chart_type="table",
        columns=["region", "amount", "date"]
    )
    
    # Pie chart  
    pie_chart = client.create_chart_from_table(
        chart_name="Sales by Region", 
        table="sales",
        owner="analyst",
        chart_type="pie",
        metric={"aggregate": "SUM", "column": {"column_name": "amount"}},
        groupby=["region"]
    )
    
    # Histogram
    hist_chart = client.create_chart_from_table(
        chart_name="Amount Distribution",
        table="sales", 
        owner="analyst",
        chart_type="histogram",
        all_columns_x=["amount"],
        bins=10
    )

Resource Management & Migration

with SupersetClient() as client:
    # Get comprehensive user summary
    summary = client.get_user_summary("data_analyst")
    print(f"User has: {summary['summary']}")
    
    # Migrate resources between users
    result = client.migrate_user_resources(
        from_user="old_analyst", 
        to_user="new_analyst",
        dry_run=False
    )
    
    # Clean up user resources
    cleanup = client.cleanup_user("temp_user", dry_run=False)
    print(f"Deleted: {len(cleanup['chart_ids'])} charts, {len(cleanup['dashboard_ids'])} dashboards")

Dataset Ownership Management

from superset_toolkit import SupersetClient
from superset_toolkit.datasets import add_dataset_owner, refresh_dataset_metadata
from superset_toolkit.ensure import get_dataset_id

# Login as admin (with privileges to modify ownership)
client = SupersetClient()

# Get dataset ID
dataset_id = get_dataset_id(client.session, client.base_url, "sales_data", "public")

# Refresh dataset metadata (update columns from database)
refresh_dataset_metadata(client.session, client.base_url, dataset_id)

# Add an owner to the dataset without removing existing owners
# Admin logs in, but adds other users as owners for collaboration
add_dataset_owner(
    client.session, 
    client.base_url, 
    dataset_id, 
    username="data_analyst"  # Add this user as owner
)

# Add multiple owners
for username in ["analyst1", "analyst2", "analyst3"]:
    add_dataset_owner(client.session, client.base_url, dataset_id, username)

Key Features:

• ✅ Preserves existing owners - doesn't remove anyone
• ✅ Admin authentication - login as admin, add others as owners
• ✅ Idempotent - won't duplicate if user is already an owner
• ✅ Collaboration-friendly - enable team access to datasets

🔧 Installation & Setup

# Install the toolkit
pip install -e .

# Optional: Install with CLI support  
pip install -e ".[cli]"

🎯 Why Choose This SDK?

Before (Traditional Approach):

# Manual user ID resolution, parameter repetition, fragmented operations
user_id = get_user_id_by_username(session, base_url, "john")
dataset_id = ensure_dataset(session, base_url, db_id, schema, table)  
chart_id = create_table_chart(session, base_url, name, dataset_id, user_id)
dashboard_id = ensure_dashboard(session, base_url, title, slug)
link_chart_to_dashboard(session, base_url, chart_id, dashboard_id)

After (Professional SDK):

# Clean, username-first, composite operations
with SupersetClient() as client:
    chart_id = client.create_table_chart("Report", table="sales", owner="analyst")
    dashboard_id = client.create_dashboard("Dashboard", "dashboard", charts=["Report"])

📚 Documentation

• 📖 Full Documentation - Comprehensive guides and API reference
• 🎯 Examples - Ready-to-run examples for common patterns
• 🔧 Configuration Guide - Setup and customization
• 👤 User Management - Username-aware operations

🎯 Supported Chart Types

Chart Type	Client Method	Standalone Function	Username Support
Table	client.create_table_chart()	create_table_chart()	✅
Pie	client.create_chart_from_table(type="pie")	create_pie_chart()	✅
Histogram	client.create_chart_from_table(type="histogram")	create_histogram_chart()	✅
Area	client.create_chart_from_table(type="area")	create_area_chart()	✅
Pivot	Available via standalone function	create_pivot_table_chart()	✅

🛡️ Error Handling & Permissions

The SDK gracefully handles permission restrictions:

• JWT Token Extraction: Gets user ID without requiring admin permissions
• 403 Fallback Logic: Smart fallbacks for non-admin users
• Professional Exceptions: Clear error messages with context

# Robust error handling
try:
    chart_id = client.create_table_chart("Report", table="sales", owner="user") 
except AuthenticationError as e:
    print(f"Auth issue: {e}")
except SupersetToolkitError as e:
    print(f"Operation failed: {e}")

📁 Project Structure

superset_toolkit/
├── 📖 docs/              # Comprehensive documentation
├── 🎯 examples/          # Ready-to-run examples  
├── 🔧 src/superset_toolkit/
│   ├── client.py         # Professional SupersetClient class
│   ├── auth.py          # JWT + permission-aware authentication
│   ├── charts.py        # Username-aware chart creation
│   ├── dashboard.py     # Dashboard management 
│   ├── queries.py       # Resource filtering and queries
│   └── utils/           # Utilities (metrics, etc.)
└── 🧪 src/superset-api-test/  # Test scripts

🚀 Getting Started

1. Install: pip install -e .
2. Configure: Set up credentials (Config class or env vars)
3. Explore: Check examples/ for common patterns
4. Read Docs: Review docs/ for comprehensive guides

📝 License & Contributing

MIT License - Open source project for the Superset community.