SELECT, FROM & the Sample Database · SQL & Databases

SQL is a language for asking questions of data. We start with the one statement you'll write most: SELECT.

SQL (Structured Query Language) is how you talk to a relational database — a set of tables, each a grid of rows (records) and columns (fields). You describe what data you want and the database works out how to get it. The workhorse is the SELECT statement.

The sample database

Every example in this book uses the same small company database — departments, employees and projects. Learn it once and you can follow every query that follows.

Our sample schema and data

-- departments
dept_id | dept_name   | location
   10   | Engineering | Bengaluru
   20   | Sales       | Mumbai
   30   | HR          | Pune

-- employees
emp_id | name    | dept_id | salary | manager_id | hire_date
  1    | Asha    |   10    | 90000  |   NULL     | 2019-03-01
  2    | Ravi    |   10    | 60000  |    1       | 2021-06-15
  3    | Meena   |   20    | 75000  |    1       | 2020-01-20
  4    | John    |   20    | 50000  |    3       | 2022-09-10
  5    | Priya   |   30    | 55000  |    1       | 2021-11-05
  6    | Sam     |   10    | 60000  |    1       | 2023-02-01

-- projects
proj_id | proj_name | dept_id | budget
  101   | Apollo    |   10    | 500000
  102   | Orion     |   20    | 300000
  103   | Helix     |   10    | 150000

The basic SELECT

Pick columns from a table

SELECT name, salary          -- columns you want
FROM   employees;            -- table to read from

SELECT *                     -- all columns
FROM   departments;

DISTINCT, ORDER BY and LIMIT

SELECT DISTINCT dept_id      -- unique values only
FROM   employees;

SELECT name, salary
FROM   employees
ORDER  BY salary DESC        -- highest first
LIMIT  3;                    -- only the top 3 rows

⚡ The edge

SELECT chooses columns, FROM chooses the table. Use SELECT * only while exploring — in real queries, name the columns so the result is stable and readable.
ORDER BY defaults to ascending; add DESC for descending. Pair it with LIMIT (or TOP in SQL Server) to get 'the top N' rows.

Worked example

List the three highest-paid employees, showing only their name and salary.

Choose the columns: name and salary.
Sort by salary, highest first: ORDER BY salary DESC.
Keep only the first three rows: LIMIT 3.

Answer: SELECT name, salary FROM employees ORDER BY salary DESC LIMIT 3;

Worked example

How many distinct departments do employees belong to?

You want unique department values, so use DISTINCT on dept_id.
To count them, combine COUNT with DISTINCT.
The query is COUNT(DISTINCT dept_id).

Answer: SELECT COUNT(DISTINCT dept_id) FROM employees; -- returns 3

⚠ Watch out

SELECT * is convenient but fragile — column order/additions can break code that depends on it.
ORDER BY is ascending by default; don't assume descending.
Top-N syntax differs by dialect: LIMIT n (MySQL/Postgres) vs TOP n (SQL Server).