@dotdo/postgres 0.1.0

package/src/playground/sample-datasets.ts

@@ -0,0 +1,674 @@
+/**
+ * Sample Datasets for Playground
+ *
+ * Provides pre-built sample datasets for demo and learning purposes.
+ * Each dataset includes schema creation and realistic data.
+ *
+ * @module playground/sample-datasets
+ */
+
+/**
+ * A sample dataset with metadata
+ */
+export interface SampleDataset {
+  /** Unique identifier */
+  id: string
+  /** Display name */
+  name: string
+  /** Brief description */
+  description: string
+  /** Categories/tags for filtering */
+  tags: string[]
+  /** SQL to create the schema */
+  schemaSql: string
+  /** SQL to insert data */
+  dataSql: string
+  /** Tables created by this dataset */
+  tables: string[]
+  /** Approximate row counts per table */
+  rowCounts: Record<string, number>
+  /** Complexity level */
+  complexity: 'simple' | 'moderate' | 'complex'
+}
+
+/**
+ * Blog/Social Media Dataset
+ * A classic users/posts/comments schema for learning SQL
+ */
+export const BLOG_DATASET: SampleDataset = {
+  id: 'blog',
+  name: 'Blog Platform',
+  description: 'Users, posts, and comments - perfect for learning joins and aggregations',
+  tags: ['social', 'blog', 'users', 'posts', 'comments'],
+  tables: ['users', 'posts', 'comments', 'tags', 'post_tags'],
+  rowCounts: {
+    users: 5,
+    posts: 15,
+    comments: 20,
+    tags: 5,
+    post_tags: 20,
+  },
+  complexity: 'simple',
+  schemaSql: `
+-- Users table
+CREATE TABLE IF NOT EXISTS users (
+  id SERIAL PRIMARY KEY,
+  name TEXT NOT NULL,
+  email TEXT UNIQUE NOT NULL,
+  bio TEXT,
+  created_at TIMESTAMP DEFAULT NOW()
+);
+
+-- Posts table
+CREATE TABLE IF NOT EXISTS posts (
+  id SERIAL PRIMARY KEY,
+  user_id INTEGER NOT NULL REFERENCES users(id) ON DELETE CASCADE,
+  title TEXT NOT NULL,
+  body TEXT,
+  published BOOLEAN DEFAULT false,
+  view_count INTEGER DEFAULT 0,
+  created_at TIMESTAMP DEFAULT NOW(),
+  updated_at TIMESTAMP DEFAULT NOW()
+);
+
+-- Comments table
+CREATE TABLE IF NOT EXISTS comments (
+  id SERIAL PRIMARY KEY,
+  post_id INTEGER NOT NULL REFERENCES posts(id) ON DELETE CASCADE,
+  user_id INTEGER NOT NULL REFERENCES users(id) ON DELETE CASCADE,
+  content TEXT NOT NULL,
+  created_at TIMESTAMP DEFAULT NOW()
+);
+
+-- Tags table
+CREATE TABLE IF NOT EXISTS tags (
+  id SERIAL PRIMARY KEY,
+  name TEXT UNIQUE NOT NULL,
+  description TEXT
+);
+
+-- Post-Tags junction table (many-to-many)
+CREATE TABLE IF NOT EXISTS post_tags (
+  post_id INTEGER REFERENCES posts(id) ON DELETE CASCADE,
+  tag_id INTEGER REFERENCES tags(id) ON DELETE CASCADE,
+  PRIMARY KEY (post_id, tag_id)
+);
+
+-- Indexes for performance
+CREATE INDEX IF NOT EXISTS idx_posts_user_id ON posts(user_id);
+CREATE INDEX IF NOT EXISTS idx_posts_published ON posts(published) WHERE published = true;
+CREATE INDEX IF NOT EXISTS idx_comments_post_id ON comments(post_id);
+CREATE INDEX IF NOT EXISTS idx_comments_user_id ON comments(user_id);
+`,
+  dataSql: `
+-- Insert users
+INSERT INTO users (name, email, bio) VALUES
+  ('Alice Johnson', 'alice@example.com', 'Full-stack developer and tech blogger'),
+  ('Bob Smith', 'bob@example.com', 'Database enthusiast and SQL guru'),
+  ('Charlie Brown', 'charlie@example.com', 'Open source contributor'),
+  ('Diana Ross', 'diana@example.com', 'DevOps engineer and cloud architect'),
+  ('Eve Wilson', 'eve@example.com', 'Security researcher and writer')
+ON CONFLICT (email) DO NOTHING;
+
+-- Insert tags
+INSERT INTO tags (name, description) VALUES
+  ('postgresql', 'PostgreSQL database topics'),
+  ('javascript', 'JavaScript and frontend development'),
+  ('devops', 'DevOps, CI/CD, and infrastructure'),
+  ('security', 'Security best practices'),
+  ('tutorial', 'Step-by-step tutorials')
+ON CONFLICT (name) DO NOTHING;
+
+-- Insert posts
+INSERT INTO posts (user_id, title, body, published, view_count, created_at)
+SELECT
+  u.id,
+  CASE (row_number() OVER ()) % 5
+    WHEN 0 THEN 'Getting Started with PostgreSQL'
+    WHEN 1 THEN 'Understanding SQL Joins'
+    WHEN 2 THEN 'Database Performance Tips'
+    WHEN 3 THEN 'Building REST APIs'
+    WHEN 4 THEN 'Security Best Practices'
+  END || ' - Part ' || ((row_number() OVER () - 1) / 5 + 1),
+  'This is an in-depth article about ' ||
+  CASE (row_number() OVER ()) % 5
+    WHEN 0 THEN 'PostgreSQL fundamentals and getting your first database running.'
+    WHEN 1 THEN 'different types of SQL joins and when to use them.'
+    WHEN 2 THEN 'optimizing queries and indexing strategies.'
+    WHEN 3 THEN 'designing and implementing RESTful APIs.'
+    WHEN 4 THEN 'securing your applications and databases.'
+  END,
+  (row_number() OVER ()) % 3 != 0,  -- 2/3 are published
+  (random() * 1000)::integer,
+  NOW() - ((row_number() OVER ()) || ' days')::interval
+FROM users u
+CROSS JOIN generate_series(1, 3)
+ON CONFLICT DO NOTHING;
+
+-- Insert post_tags relationships
+INSERT INTO post_tags (post_id, tag_id)
+SELECT p.id, t.id
+FROM posts p
+CROSS JOIN tags t
+WHERE random() < 0.4
+ON CONFLICT DO NOTHING;
+
+-- Insert comments
+INSERT INTO comments (post_id, user_id, content, created_at)
+SELECT
+  p.id,
+  u.id,
+  CASE (random() * 4)::integer
+    WHEN 0 THEN 'Great article! Very helpful.'
+    WHEN 1 THEN 'Thanks for sharing this information.'
+    WHEN 2 THEN 'I learned something new today.'
+    WHEN 3 THEN 'Could you elaborate on this topic?'
+    ELSE 'Interesting perspective!'
+  END,
+  p.created_at + ((random() * 7)::integer || ' days')::interval
+FROM posts p
+CROSS JOIN users u
+WHERE p.published = true
+  AND u.id != p.user_id
+  AND random() < 0.5
+LIMIT 20
+ON CONFLICT DO NOTHING;
+`,
+}
+
+/**
+ * E-Commerce Dataset
+ * Products, orders, and customers for business analytics
+ */
+export const ECOMMERCE_DATASET: SampleDataset = {
+  id: 'ecommerce',
+  name: 'E-Commerce Store',
+  description: 'Products, orders, and customers - great for analytics and reporting queries',
+  tags: ['ecommerce', 'products', 'orders', 'analytics', 'business'],
+  tables: ['customers', 'products', 'categories', 'orders', 'order_items'],
+  rowCounts: {
+    customers: 10,
+    products: 20,
+    categories: 5,
+    orders: 25,
+    order_items: 50,
+  },
+  complexity: 'moderate',
+  schemaSql: `
+-- Categories table
+CREATE TABLE IF NOT EXISTS categories (
+  id SERIAL PRIMARY KEY,
+  name TEXT UNIQUE NOT NULL,
+  description TEXT,
+  parent_id INTEGER REFERENCES categories(id)
+);
+
+-- Customers table
+CREATE TABLE IF NOT EXISTS customers (
+  id SERIAL PRIMARY KEY,
+  first_name TEXT NOT NULL,
+  last_name TEXT NOT NULL,
+  email TEXT UNIQUE NOT NULL,
+  phone TEXT,
+  address TEXT,
+  city TEXT,
+  country TEXT DEFAULT 'USA',
+  created_at TIMESTAMP DEFAULT NOW()
+);
+
+-- Products table
+CREATE TABLE IF NOT EXISTS products (
+  id SERIAL PRIMARY KEY,
+  name TEXT NOT NULL,
+  description TEXT,
+  category_id INTEGER REFERENCES categories(id),
+  price DECIMAL(10, 2) NOT NULL,
+  cost DECIMAL(10, 2),
+  stock_quantity INTEGER DEFAULT 0,
+  sku TEXT UNIQUE,
+  active BOOLEAN DEFAULT true,
+  created_at TIMESTAMP DEFAULT NOW()
+);
+
+-- Orders table
+CREATE TABLE IF NOT EXISTS orders (
+  id SERIAL PRIMARY KEY,
+  customer_id INTEGER NOT NULL REFERENCES customers(id),
+  order_date TIMESTAMP DEFAULT NOW(),
+  status TEXT DEFAULT 'pending' CHECK (status IN ('pending', 'processing', 'shipped', 'delivered', 'cancelled')),
+  shipping_address TEXT,
+  total_amount DECIMAL(10, 2) DEFAULT 0,
+  notes TEXT
+);
+
+-- Order items table
+CREATE TABLE IF NOT EXISTS order_items (
+  id SERIAL PRIMARY KEY,
+  order_id INTEGER NOT NULL REFERENCES orders(id) ON DELETE CASCADE,
+  product_id INTEGER NOT NULL REFERENCES products(id),
+  quantity INTEGER NOT NULL CHECK (quantity > 0),
+  unit_price DECIMAL(10, 2) NOT NULL,
+  discount_percent DECIMAL(5, 2) DEFAULT 0
+);
+
+-- Indexes
+CREATE INDEX IF NOT EXISTS idx_products_category ON products(category_id);
+CREATE INDEX IF NOT EXISTS idx_products_active ON products(active) WHERE active = true;
+CREATE INDEX IF NOT EXISTS idx_orders_customer ON orders(customer_id);
+CREATE INDEX IF NOT EXISTS idx_orders_status ON orders(status);
+CREATE INDEX IF NOT EXISTS idx_orders_date ON orders(order_date);
+CREATE INDEX IF NOT EXISTS idx_order_items_order ON order_items(order_id);
+CREATE INDEX IF NOT EXISTS idx_order_items_product ON order_items(product_id);
+`,
+  dataSql: `
+-- Insert categories
+INSERT INTO categories (name, description) VALUES
+  ('Electronics', 'Electronic devices and accessories'),
+  ('Clothing', 'Apparel and fashion items'),
+  ('Books', 'Books and reading materials'),
+  ('Home & Garden', 'Home decor and garden supplies'),
+  ('Sports', 'Sports equipment and fitness gear')
+ON CONFLICT (name) DO NOTHING;
+
+-- Insert customers
+INSERT INTO customers (first_name, last_name, email, phone, city, country) VALUES
+  ('John', 'Doe', 'john.doe@email.com', '555-0101', 'New York', 'USA'),
+  ('Jane', 'Smith', 'jane.smith@email.com', '555-0102', 'Los Angeles', 'USA'),
+  ('Michael', 'Johnson', 'michael.j@email.com', '555-0103', 'Chicago', 'USA'),
+  ('Emily', 'Williams', 'emily.w@email.com', '555-0104', 'Houston', 'USA'),
+  ('David', 'Brown', 'david.b@email.com', '555-0105', 'Phoenix', 'USA'),
+  ('Sarah', 'Davis', 'sarah.d@email.com', '555-0106', 'Seattle', 'USA'),
+  ('Chris', 'Miller', 'chris.m@email.com', '555-0107', 'Boston', 'USA'),
+  ('Lisa', 'Wilson', 'lisa.w@email.com', '555-0108', 'Miami', 'USA'),
+  ('James', 'Taylor', 'james.t@email.com', '555-0109', 'Denver', 'USA'),
+  ('Amy', 'Anderson', 'amy.a@email.com', '555-0110', 'Portland', 'USA')
+ON CONFLICT (email) DO NOTHING;
+
+-- Insert products
+INSERT INTO products (name, description, category_id, price, cost, stock_quantity, sku) VALUES
+  ('Laptop Pro 15', 'High-performance laptop', 1, 1299.99, 900.00, 50, 'ELEC-001'),
+  ('Wireless Mouse', 'Ergonomic wireless mouse', 1, 29.99, 15.00, 200, 'ELEC-002'),
+  ('USB-C Hub', '7-in-1 USB-C adapter', 1, 49.99, 25.00, 150, 'ELEC-003'),
+  ('Mechanical Keyboard', 'RGB mechanical keyboard', 1, 89.99, 45.00, 100, 'ELEC-004'),
+  ('Cotton T-Shirt', 'Premium cotton t-shirt', 2, 24.99, 8.00, 500, 'CLOTH-001'),
+  ('Denim Jeans', 'Classic fit denim jeans', 2, 59.99, 25.00, 300, 'CLOTH-002'),
+  ('Hoodie', 'Comfortable pullover hoodie', 2, 44.99, 18.00, 250, 'CLOTH-003'),
+  ('Running Shoes', 'Lightweight running shoes', 2, 79.99, 35.00, 200, 'CLOTH-004'),
+  ('SQL Handbook', 'Complete SQL reference', 3, 39.99, 15.00, 100, 'BOOK-001'),
+  ('Clean Code', 'Software craftsmanship guide', 3, 44.99, 20.00, 75, 'BOOK-002'),
+  ('Design Patterns', 'Gang of Four patterns book', 3, 54.99, 25.00, 60, 'BOOK-003'),
+  ('Table Lamp', 'Modern LED table lamp', 4, 34.99, 15.00, 150, 'HOME-001'),
+  ('Plant Pot Set', 'Ceramic plant pots (3 pack)', 4, 29.99, 12.00, 200, 'HOME-002'),
+  ('Throw Pillow', 'Decorative throw pillow', 4, 19.99, 7.00, 300, 'HOME-003'),
+  ('Yoga Mat', 'Non-slip exercise mat', 5, 24.99, 10.00, 250, 'SPORT-001'),
+  ('Dumbbells Set', 'Adjustable dumbbell pair', 5, 149.99, 75.00, 80, 'SPORT-002'),
+  ('Jump Rope', 'Speed jump rope', 5, 14.99, 5.00, 400, 'SPORT-003'),
+  ('Water Bottle', 'Insulated sports bottle', 5, 19.99, 8.00, 350, 'SPORT-004'),
+  ('Webcam HD', '1080p HD webcam', 1, 69.99, 35.00, 120, 'ELEC-005'),
+  ('Monitor Stand', 'Adjustable monitor riser', 1, 39.99, 18.00, 180, 'ELEC-006')
+ON CONFLICT (sku) DO NOTHING;
+
+-- Insert orders with varying dates and statuses
+INSERT INTO orders (customer_id, order_date, status, shipping_address, total_amount)
+SELECT
+  (random() * 9 + 1)::integer,
+  NOW() - ((random() * 60)::integer || ' days')::interval,
+  CASE (random() * 4)::integer
+    WHEN 0 THEN 'pending'
+    WHEN 1 THEN 'processing'
+    WHEN 2 THEN 'shipped'
+    WHEN 3 THEN 'delivered'
+    ELSE 'delivered'
+  END,
+  '123 Main St, City, State 12345',
+  0
+FROM generate_series(1, 25)
+ON CONFLICT DO NOTHING;
+
+-- Insert order items
+INSERT INTO order_items (order_id, product_id, quantity, unit_price, discount_percent)
+SELECT
+  o.id,
+  p.id,
+  (random() * 3 + 1)::integer,
+  p.price,
+  CASE WHEN random() < 0.3 THEN (random() * 15)::integer ELSE 0 END
+FROM orders o
+CROSS JOIN products p
+WHERE random() < 0.15
+ON CONFLICT DO NOTHING;
+
+-- Update order totals
+UPDATE orders SET total_amount = (
+  SELECT COALESCE(SUM(quantity * unit_price * (1 - discount_percent/100)), 0)
+  FROM order_items
+  WHERE order_items.order_id = orders.id
+);
+`,
+}
+
+/**
+ * HR/Employee Dataset
+ * Employees, departments, and salaries for organizational queries
+ */
+export const HR_DATASET: SampleDataset = {
+  id: 'hr',
+  name: 'HR & Employees',
+  description: 'Employees, departments, and salaries - ideal for hierarchical and analytical queries',
+  tags: ['hr', 'employees', 'departments', 'salaries', 'organization'],
+  tables: ['departments', 'employees', 'salaries', 'job_history'],
+  rowCounts: {
+    departments: 6,
+    employees: 20,
+    salaries: 20,
+    job_history: 15,
+  },
+  complexity: 'moderate',
+  schemaSql: `
+-- Departments table
+CREATE TABLE IF NOT EXISTS departments (
+  id SERIAL PRIMARY KEY,
+  name TEXT UNIQUE NOT NULL,
+  location TEXT,
+  budget DECIMAL(12, 2),
+  manager_id INTEGER -- Will be FK to employees after employees created
+);
+
+-- Employees table with self-referencing manager
+CREATE TABLE IF NOT EXISTS employees (
+  id SERIAL PRIMARY KEY,
+  first_name TEXT NOT NULL,
+  last_name TEXT NOT NULL,
+  email TEXT UNIQUE NOT NULL,
+  phone TEXT,
+  hire_date DATE NOT NULL,
+  job_title TEXT NOT NULL,
+  department_id INTEGER REFERENCES departments(id),
+  manager_id INTEGER REFERENCES employees(id),
+  is_active BOOLEAN DEFAULT true
+);
+
+-- Add manager FK to departments
+ALTER TABLE departments
+ADD CONSTRAINT fk_department_manager
+FOREIGN KEY (manager_id) REFERENCES employees(id)
+ON DELETE SET NULL;
+
+-- Salaries table (history of salary changes)
+CREATE TABLE IF NOT EXISTS salaries (
+  id SERIAL PRIMARY KEY,
+  employee_id INTEGER NOT NULL REFERENCES employees(id) ON DELETE CASCADE,
+  amount DECIMAL(10, 2) NOT NULL,
+  effective_date DATE NOT NULL,
+  end_date DATE,
+  CONSTRAINT valid_date_range CHECK (end_date IS NULL OR end_date > effective_date)
+);
+
+-- Job history table
+CREATE TABLE IF NOT EXISTS job_history (
+  id SERIAL PRIMARY KEY,
+  employee_id INTEGER NOT NULL REFERENCES employees(id) ON DELETE CASCADE,
+  job_title TEXT NOT NULL,
+  department_id INTEGER REFERENCES departments(id),
+  start_date DATE NOT NULL,
+  end_date DATE,
+  reason TEXT
+);
+
+-- Indexes
+CREATE INDEX IF NOT EXISTS idx_employees_department ON employees(department_id);
+CREATE INDEX IF NOT EXISTS idx_employees_manager ON employees(manager_id);
+CREATE INDEX IF NOT EXISTS idx_employees_hire_date ON employees(hire_date);
+CREATE INDEX IF NOT EXISTS idx_salaries_employee ON salaries(employee_id);
+CREATE INDEX IF NOT EXISTS idx_salaries_effective ON salaries(effective_date);
+CREATE INDEX IF NOT EXISTS idx_job_history_employee ON job_history(employee_id);
+`,
+  dataSql: `
+-- Insert departments
+INSERT INTO departments (name, location, budget) VALUES
+  ('Engineering', 'San Francisco', 2000000),
+  ('Marketing', 'New York', 800000),
+  ('Sales', 'Chicago', 1200000),
+  ('Human Resources', 'San Francisco', 500000),
+  ('Finance', 'New York', 600000),
+  ('Operations', 'Chicago', 900000)
+ON CONFLICT (name) DO NOTHING;
+
+-- Insert employees (executives first, then managers, then staff)
+INSERT INTO employees (first_name, last_name, email, phone, hire_date, job_title, department_id, manager_id) VALUES
+  -- CEO (no manager)
+  ('Robert', 'Chen', 'robert.chen@company.com', '555-1001', '2018-01-15', 'CEO', NULL, NULL),
+  -- Department heads (report to CEO)
+  ('Sarah', 'Martinez', 'sarah.m@company.com', '555-1002', '2018-06-01', 'VP Engineering', 1, 1),
+  ('Michael', 'Thompson', 'michael.t@company.com', '555-1003', '2018-07-15', 'VP Marketing', 2, 1),
+  ('Jennifer', 'Lee', 'jennifer.l@company.com', '555-1004', '2019-01-10', 'VP Sales', 3, 1),
+  ('William', 'Garcia', 'william.g@company.com', '555-1005', '2019-03-20', 'HR Director', 4, 1),
+  ('Elizabeth', 'Brown', 'elizabeth.b@company.com', '555-1006', '2019-05-01', 'CFO', 5, 1),
+  -- Managers
+  ('David', 'Wilson', 'david.w@company.com', '555-2001', '2019-08-15', 'Engineering Manager', 1, 2),
+  ('Jessica', 'Taylor', 'jessica.t@company.com', '555-2002', '2019-09-01', 'Marketing Manager', 2, 3),
+  ('Christopher', 'Moore', 'chris.m@company.com', '555-2003', '2020-01-15', 'Sales Manager', 3, 4),
+  -- Staff
+  ('Amanda', 'Anderson', 'amanda.a@company.com', '555-3001', '2020-03-01', 'Senior Engineer', 1, 7),
+  ('Daniel', 'Jackson', 'daniel.j@company.com', '555-3002', '2020-06-15', 'Software Engineer', 1, 7),
+  ('Michelle', 'White', 'michelle.w@company.com', '555-3003', '2020-09-01', 'Junior Engineer', 1, 7),
+  ('Ryan', 'Harris', 'ryan.h@company.com', '555-3004', '2021-01-10', 'Marketing Specialist', 2, 8),
+  ('Lauren', 'Martin', 'lauren.m@company.com', '555-3005', '2021-03-15', 'Content Writer', 2, 8),
+  ('Kevin', 'Thompson', 'kevin.t@company.com', '555-3006', '2021-06-01', 'Sales Representative', 3, 9),
+  ('Nicole', 'Robinson', 'nicole.r@company.com', '555-3007', '2021-08-15', 'Sales Representative', 3, 9),
+  ('Brandon', 'Clark', 'brandon.c@company.com', '555-3008', '2022-01-10', 'HR Specialist', 4, 5),
+  ('Stephanie', 'Lewis', 'stephanie.l@company.com', '555-3009', '2022-04-01', 'Accountant', 5, 6),
+  ('Andrew', 'Walker', 'andrew.w@company.com', '555-3010', '2022-07-15', 'Operations Analyst', 6, 1),
+  ('Rachel', 'Hall', 'rachel.h@company.com', '555-3011', '2023-01-05', 'Software Engineer', 1, 7)
+ON CONFLICT (email) DO NOTHING;
+
+-- Update department managers
+UPDATE departments SET manager_id = 2 WHERE name = 'Engineering';
+UPDATE departments SET manager_id = 3 WHERE name = 'Marketing';
+UPDATE departments SET manager_id = 4 WHERE name = 'Sales';
+UPDATE departments SET manager_id = 5 WHERE name = 'Human Resources';
+UPDATE departments SET manager_id = 6 WHERE name = 'Finance';
+UPDATE departments SET manager_id = 1 WHERE name = 'Operations';
+
+-- Insert current salaries
+INSERT INTO salaries (employee_id, amount, effective_date)
+SELECT
+  id,
+  CASE
+    WHEN job_title LIKE '%CEO%' THEN 350000
+    WHEN job_title LIKE '%CFO%' THEN 280000
+    WHEN job_title LIKE '%VP%' THEN 220000
+    WHEN job_title LIKE '%Director%' THEN 180000
+    WHEN job_title LIKE '%Manager%' THEN 130000
+    WHEN job_title LIKE '%Senior%' THEN 120000
+    WHEN job_title LIKE '%Specialist%' OR job_title LIKE '%Analyst%' THEN 80000
+    WHEN job_title LIKE '%Junior%' THEN 65000
+    ELSE 75000
+  END + (random() * 20000)::integer,
+  hire_date
+FROM employees
+ON CONFLICT DO NOTHING;
+
+-- Insert some job history
+INSERT INTO job_history (employee_id, job_title, department_id, start_date, end_date, reason)
+SELECT
+  e.id,
+  'Previous Role',
+  e.department_id,
+  e.hire_date - INTERVAL '1 year',
+  e.hire_date - INTERVAL '1 day',
+  'Promoted'
+FROM employees e
+WHERE random() < 0.4 AND e.hire_date < '2022-01-01'
+ON CONFLICT DO NOTHING;
+`,
+}
+
+/**
+ * IoT/Sensors Dataset
+ * Time-series data from sensors for time-based analytics
+ */
+export const IOT_DATASET: SampleDataset = {
+  id: 'iot',
+  name: 'IoT Sensors',
+  description: 'Time-series sensor data - perfect for time-based queries and analytics',
+  tags: ['iot', 'sensors', 'time-series', 'analytics', 'monitoring'],
+  tables: ['locations', 'sensors', 'readings'],
+  rowCounts: {
+    locations: 4,
+    sensors: 8,
+    readings: 200,
+  },
+  complexity: 'complex',
+  schemaSql: `
+-- Locations table
+CREATE TABLE IF NOT EXISTS locations (
+  id SERIAL PRIMARY KEY,
+  name TEXT UNIQUE NOT NULL,
+  address TEXT,
+  latitude DECIMAL(9, 6),
+  longitude DECIMAL(9, 6),
+  timezone TEXT DEFAULT 'UTC'
+);
+
+-- Sensors table
+CREATE TABLE IF NOT EXISTS sensors (
+  id SERIAL PRIMARY KEY,
+  location_id INTEGER NOT NULL REFERENCES locations(id),
+  name TEXT NOT NULL,
+  type TEXT NOT NULL CHECK (type IN ('temperature', 'humidity', 'pressure', 'co2', 'motion')),
+  unit TEXT NOT NULL,
+  min_threshold DECIMAL(10, 2),
+  max_threshold DECIMAL(10, 2),
+  is_active BOOLEAN DEFAULT true,
+  installed_at TIMESTAMP DEFAULT NOW(),
+  UNIQUE (location_id, name)
+);
+
+-- Readings table (time-series data)
+CREATE TABLE IF NOT EXISTS readings (
+  id SERIAL PRIMARY KEY,
+  sensor_id INTEGER NOT NULL REFERENCES sensors(id),
+  value DECIMAL(10, 4) NOT NULL,
+  recorded_at TIMESTAMP NOT NULL DEFAULT NOW(),
+  is_anomaly BOOLEAN DEFAULT false,
+  metadata JSONB
+);
+
+-- Indexes for time-series queries
+CREATE INDEX IF NOT EXISTS idx_readings_sensor ON readings(sensor_id);
+CREATE INDEX IF NOT EXISTS idx_readings_time ON readings(recorded_at);
+CREATE INDEX IF NOT EXISTS idx_readings_sensor_time ON readings(sensor_id, recorded_at);
+CREATE INDEX IF NOT EXISTS idx_readings_anomaly ON readings(is_anomaly) WHERE is_anomaly = true;
+`,
+  dataSql: `
+-- Insert locations
+INSERT INTO locations (name, address, latitude, longitude, timezone) VALUES
+  ('Main Office', '100 Tech Park, San Francisco, CA', 37.7749, -122.4194, 'America/Los_Angeles'),
+  ('Data Center', '200 Server Lane, Austin, TX', 30.2672, -97.7431, 'America/Chicago'),
+  ('Warehouse A', '300 Storage Blvd, Phoenix, AZ', 33.4484, -112.0740, 'America/Phoenix'),
+  ('Factory Floor', '400 Manufacturing Dr, Detroit, MI', 42.3314, -83.0458, 'America/Detroit')
+ON CONFLICT (name) DO NOTHING;
+
+-- Insert sensors
+INSERT INTO sensors (location_id, name, type, unit, min_threshold, max_threshold) VALUES
+  (1, 'Office Temp 1', 'temperature', 'celsius', 18, 26),
+  (1, 'Office Humidity', 'humidity', 'percent', 30, 60),
+  (2, 'Server Room Temp', 'temperature', 'celsius', 15, 24),
+  (2, 'Server Room Humidity', 'humidity', 'percent', 40, 60),
+  (3, 'Warehouse Temp', 'temperature', 'celsius', 10, 35),
+  (3, 'Warehouse CO2', 'co2', 'ppm', 400, 1000),
+  (4, 'Factory Pressure', 'pressure', 'bar', 0.95, 1.05),
+  (4, 'Factory Motion', 'motion', 'count', 0, 1000)
+ON CONFLICT (location_id, name) DO NOTHING;
+
+-- Insert readings for the past 7 days (multiple per sensor)
+INSERT INTO readings (sensor_id, value, recorded_at, is_anomaly, metadata)
+SELECT
+  s.id,
+  CASE s.type
+    WHEN 'temperature' THEN 20 + (random() * 6 - 3) + (2 * sin(extract(hour from ts) * pi() / 12))
+    WHEN 'humidity' THEN 45 + (random() * 20 - 10)
+    WHEN 'pressure' THEN 1.0 + (random() * 0.08 - 0.04)
+    WHEN 'co2' THEN 600 + (random() * 300 - 100)
+    WHEN 'motion' THEN CASE WHEN extract(hour from ts) BETWEEN 8 AND 18 THEN (random() * 500)::integer ELSE (random() * 50)::integer END
+  END,
+  ts,
+  random() < 0.02, -- 2% chance of anomaly
+  jsonb_build_object('quality', CASE WHEN random() < 0.95 THEN 'good' ELSE 'degraded' END)
+FROM sensors s
+CROSS JOIN generate_series(
+  NOW() - INTERVAL '7 days',
+  NOW(),
+  INTERVAL '1 hour'
+) AS ts
+ON CONFLICT DO NOTHING;
+
+-- Mark true anomalies (values outside thresholds)
+UPDATE readings r
+SET is_anomaly = true
+FROM sensors s
+WHERE r.sensor_id = s.id
+  AND (r.value < s.min_threshold OR r.value > s.max_threshold);
+`,
+}
+
+/**
+ * All available sample datasets
+ */
+export const SAMPLE_DATASETS: SampleDataset[] = [
+  BLOG_DATASET,
+  ECOMMERCE_DATASET,
+  HR_DATASET,
+  IOT_DATASET,
+]
+
+/**
+ * Get a dataset by ID
+ */
+export function getDatasetById(id: string): SampleDataset | undefined {
+  return SAMPLE_DATASETS.find((d) => d.id === id)
+}
+
+/**
+ * Get datasets by tag
+ */
+export function getDatasetsByTag(tag: string): SampleDataset[] {
+  return SAMPLE_DATASETS.filter((d) => d.tags.includes(tag.toLowerCase()))
+}
+
+/**
+ * Get all dataset tags
+ */
+export function getAllDatasetTags(): string[] {
+  const tags = new Set<string>()
+  for (const dataset of SAMPLE_DATASETS) {
+    for (const tag of dataset.tags) {
+      tags.add(tag)
+    }
+  }
+  return Array.from(tags).sort()
+}
+
+/**
+ * Get the SQL to load a dataset (schema + data)
+ */
+export function getDatasetSql(dataset: SampleDataset): string {
+  return `-- Dataset: ${dataset.name}
+-- ${dataset.description}
+
+-- Schema
+${dataset.schemaSql}
+
+-- Data
+${dataset.dataSql}
+`
+}
+
+/**
+ * Get combined SQL for the default blog dataset
+ * (Used by loadSampleData in playground.ts)
+ */
+export function getDefaultDatasetSql(): string {
+  return getDatasetSql(BLOG_DATASET)
+}