DS 5110 – Lecture 4

SQL Part II

Roi Yehoshua
Agenda
 Subqueries
 Common table expressions (CTEs)
 Recursive queries
 Query execution plans
 Views
 Stored procedures
 Functions
 Transactions
 Triggers
 Indexes

2 Roi Yehoshua, 2024

Subqueries
 A subquery is a SELECT query nested inside another query
 Subqueries allow you to perform SQL operations in multiple steps
 The subquery is executed before its parent query and its results are passed to the outer query
 A subquery may be nested inside
 The SELECT clause
 The FROM clause
 The WHERE clause
 The HAVING clause
 Subqueries must be enclosed inside brackets ()
 Subqueries can return an individual value or a list of records

3 Roi Yehoshua, 2024

Subqueries
 Example: Find the name of the instructor with the highest salary
 This query cannot be written without sorting the whole table
 The way to write it with a subquery:

4 Roi Yehoshua, 2024

Correlated Subqueries
 A correlated subquery references one or more columns from the outer query
 A correlated subquery is executed once for each candidate row in the outer query
 Example: Find all instructors whose salary is higher than the average salary in their
own department

5 Roi Yehoshua, 2024

Multiple-Row Subqueries
 If the subquery returns more than one row, you cannot use it with the regular
comparison operators like <, >, =
 There are 4 additional operators that allow to use the results of these subqueries:
 IN / NOT IN
 ALL
 ANY
 EXISTS / NOT EXISTS

6 Roi Yehoshua, 2024

The IN Operator
 Checks whether a column contains a value that is within the results of the subquery
 Example: Find all the courses that were offered both in Fall 2017 and Spring 2018

7 Roi Yehoshua, 2024

The IN Operator
 You can compare more than one column with the results of the subquery
 Example: Find the total number of (distinct) students who have been taught by Katz

8 Roi Yehoshua, 2024

The NOT IN Operator
 Checks if the column doesn’t contain any value within the results of the subquery
 Example: Find all the courses that were offered in Fall 2017 but not in Spring 2018

9 Roi Yehoshua, 2024

Class Exercise
 Print the id and names of students that have taken a course in Computer Science,
but have never taken a course in Physics

10 Roi Yehoshua, 2024

Solution

11 Roi Yehoshua, 2024

The ALL Operator
 The ALL operator returns true if all of the subquery results meet the condition
 Used in conjunction with one of the comparison operators (=, <>, >, >=, <, or <=)
 Example: Find the department with the highest number of courses

12 Roi Yehoshua, 2024

The ALL Operator
 Note that the following attempt to write the same query won’t work

 You cannot layer aggregate functions on top of each other in the same SELECT

13 Roi Yehoshua, 2024

The ANY Operator
 The ANY operator returns true if any of the subquery results meets the condition
 Used in conjunction with one of the comparison operators (=, <>, >, >=, <, or <=)
 Example: Find all the instructors whose salary is higher than any of the instructors in
the Biology department

14 Roi Yehoshua, 2024

The EXISTS Operator
 The EXISTS operator tests the existence of any result in the subquery
 It returns true if the subquery returns one or more rows
 For example, another way to write the query “Find all the courses that were offered
both in Fall 2017 and Spring 2018”

15 Roi Yehoshua, 2024

The NOT EXISTS Operator
 You can also use NOT EXISTS to check if the subquery didn’t return any rows
 Example: Find instructors who haven’t taught any course

16 Roi Yehoshua, 2024

The NOT EXISTS Operator
 Example: Find students who have taken all the courses in Music

17 Roi Yehoshua, 2024

The NOT EXISTS Operator
 Another way to write the same query:

18 Roi Yehoshua, 2024

Class Exercise
 Print the instructors that taught all the courses in their department

19 Roi Yehoshua, 2024

Subqueries in the FROM Clause
 SQL allows you to write subqueries in the FROM clause
 In MySQL, the subquery must be given a name using the AS keyword
 The subquery results can be joined with other tables or other subqueries
 Example: Find the highest number of courses offered in any given department

20 Roi Yehoshua, 2024

Subqueries in the SELECT Clause
 You can write subqueries in the select list of another SELECT clause
 These subqueries need to return a single value for each row of the outer SELECT
 Example: List all departments along with the number of instructors in each one

21 Roi Yehoshua, 2024

Updates with Subqueries
 You can also use subqueries in UPDATE and DELETE statements
 Example: recompute and update the tot_cred value for each student
UPDATE student AS s
SET tot_cred = (
SELECT SUM(c.credits)
FROM takes AS t
JOIN course AS c
WHERE t.course_id = c.course_id
AND s.id = t.id
AND t.grade IS NOT NULL
AND t.grade <> 'F'
);

 Sets tot_cred to NULL for students who have not taken any course
 Instead of SUM(credits) can use COALESCE(SUM(credits), 0)
 This would return SUM(credits) if it is not NULL, and 0 otherwise

22 Roi Yehoshua, 2024

Common Table Expressions (CTEs)
 CTE is a named temporary result set that only exists for the duration of the query
 CTEs are useful for breaking down complex queries into simpler parts
 Basic syntax of CTE:
WITH cte_name (column_name1, column_name2, ...)
AS (
-- Your SQL query here
)
-- Your main SQL statement using the CTE here

23 Roi Yehoshua, 2024

Common Table Expressions (CTEs)
 Example: find all the instructors that earn more than the average salary

24 Roi Yehoshua, 2024

Common Table Expressions (CTEs)
 Show the distribution of number of courses taken by students
 i.e., how many students took only one course, how many students took two courses, etc.

25 Roi Yehoshua, 2024

Recursive Queries
 A recursive query is a query that refers back to itself
 Used to deal with hierarchical or tree-structured data
 A recursive query is written using a CTE that consists of two subqueries:
 The base (non-recursive) query provides the starting point of the recursion
 The recursive query which has a reference to the CTE itself

26 Roi Yehoshua, 2024

Recursive Queries
 Example: Find for each course all its prerequisites (both direct and indirect)

27 Roi Yehoshua, 2024

Summary: SQL Execution Order

28 Roi Yehoshua, 2024

Query Execution Plans
 A query execution plan is an ordered set of steps used to access data in the database
 Defines which algorithm to use for each operation, and how to coordinate the operations
 When a query is submitted to the database, the query optimizer evaluates the
different plans for executing the query and returns what it considers the best option
 Cost difference between execution plans for a query can be enormous
 Estimation of plan cost is based on:
 Statistical information about tables
 e.g., number of rows, number of distinct values in a column
 Cost of different algorithms (e.g., join vs. subquery)
 Also computed using statistics
 Sometimes we need to manually examine and tune the optimizer plans

29 Roi Yehoshua, 2024

Query Execution Plans
 To view the execution plan, click on Execution Plan next to the query results

30 Roi Yehoshua, 2024

Class Exercise
 Write the following query: Find IDs and names of all the instructors whose salary is
greater than the salary of the instructor Katz
 Using join
 Using subselect
 Examine the execution plans of these queries
 Which query is more efficient?

31 Roi Yehoshua, 2024

Solution
 Using join

32 Roi Yehoshua, 2024

Solution
 Using subselect

33 Roi Yehoshua, 2024

Views
 A view is a virtual table based on the results set of an SQL statement
 Reasons to use views:
 Data security – allows you to hide certain data from view of certain users
 e.g., some users only need to know the instructor IDs and names, but not their salaries
 Hiding complexity of the underlying tables
 A view is created with the CREATE VIEW statement:
CREATE VIEW view_name AS
SELECT column1, column2, ...
FROM table_name
WHERE condition;

 A view always shows up-to-date data

 The database engine recreates the view whenever it is used in a query

34 Roi Yehoshua, 2024

Views
 Example: A view that shows all Physics instructors without their salary
CREATE VIEW physics_faculty AS
SELECT id, name
FROM instructor
WHERE dept_name = 'Physics';

 We can query the view above as follows:

 View names may appear in a query anywhere a table name may appear
35 Roi Yehoshua, 2024
Views
 You can see all the available views in the database under the View folder:

 To update a view use the CREATE OR REPLACE VIEW command:

CREATE OR REPLACE VIEW view_name AS
SELECT column1, column2, ...
FROM table_name
WHERE condition;

 To delete a view use the DROP VIEW command:

DROP VIEW view_name;

36 Roi Yehoshua, 2024

Updatable Views
 Some views are updatable
 i.e., you can use them in statements like INSERT or UPDATE to update the underlying tables
 For a view to be updatable, there must be a one-to-one relationship between the
rows in the view and the rows in the underlying table
 More specifically, the view isn’t updatable if it contains any of the following:
 Aggregate functions like MIN() or AVG()
 DISTINCT
 GROUP BY or HAVING
 Certain joins are not allowed
 Subquery in SELECT
 Subquery in the WHERE clause that refers to a table in the FROM clause
 UNION or UNION ALL

37 Roi Yehoshua, 2024

Stored Procedures
 A stored procedure (SP) is a group of SQL statements that is stored in the database
 You can also pass parameters to a stored procedure
 The benefits of using stored procedures:
 Reusable code that can be used by different applications
 Allows a single point of change in case the business rules change
 Faster execution
 Reduce the network traffic
 More secure than ad-hoc queries

38 Roi Yehoshua, 2024

Creating a Stored Procedure
 The syntax for creating a new stored procedure in MySQL:
DELIMITER $$ Redefine the delimiter temporarily
so that ; can be used to separate
CREATE PROCEDURE procedure_name (
statements inside the procedure
IN | OUT | INOUT parameter1 datatype,
IN | OUT | INOUT parameter2 datatype,
...
) Use ; in the body of the
BEGIN stored procedure
-- SQL statements
END $$ Use the delimiter $$ to end the
DELIMITER ; stored procedure
Change the default delimiter back to ;
 Parameter modes:
 in (default) – the parameter’s value is passed into the SP and cannot be changed inside the SP
 out – the parameter’s value is passed back to the calling program, must be a variable
 inout – the parameter’s value is passed to the SP and a new value can be assigned to it

39 Roi Yehoshua, 2024

Creating a Stored Procedure
 Example: a stored procedure that returns all courses in a given department

40 Roi Yehoshua, 2024

Creating a Stored Procedure
 You can view your procedure under the Stored Procedures folder of the schema

41 Roi Yehoshua, 2024

Calling a Stored Procedure
 To execute the store procedure, you use the CALL keyword:
CALL procedure_name(parameter1, parameter2, ...);

 For example, let’s call the get_courses_by_dept with Comp. Sci. as the argument:

42 Roi Yehoshua, 2024

Calling a Stored Procedure
 You can also execute the SP by clicking on the Execute button next to its name

43 Roi Yehoshua, 2024

OUT Parameter
 A stored procedure that returns the number of students in a given department

44 Roi Yehoshua, 2024

OUT Parameter
 To call a SP with an OUT parameter you need to pass a session variable to it
 This variable will receive the return value from the procedure
 Session variables are variables that start with @
 Don’t require any declaration
 Can be used inside any SQL query or statement
 Exist until the end of the current session
 Assignments to the variable are performed using a SET statement
 To display the value of a variable use the SELECT statement

45 Roi Yehoshua, 2024

InOut Parameter Example
 The following example demonstrates how to use an inout parameter in a SP:

46 Roi Yehoshua, 2024

InOut Parameter Example
 The following statements illustrate how to call the update_counter SP:

47 Roi Yehoshua, 2024

Class Exercise
 Create a stored procedure that gets a student ID and returns the IDs and titles of the
courses he/she has taken

48 Roi Yehoshua, 2024

Solution

49 Roi Yehoshua, 2024

Programming Language Constructs
 Stored procedures support programming language constructs like ifs and loops:
IF boolean expression THEN
statement_list
[ELSEIF boolean expression THEN
statement_list]
...
[ELSE statement_list]
END IF;

WHILE boolean expression DO

statement_list
END WHILE;

REPEAT
statement_list
UNTIL boolean expression
END REPEAT;

50 Roi Yehoshua, 2024

Programming Language Constructs
 The following procedure registers a student to a course section only if there is
enough room left in the classroom allocated for this section

51 Roi Yehoshua, 2024

Programming Language Constructs

52 Roi Yehoshua, 2024

Programming Language Constructs
 Calling the procedure:

53 Roi Yehoshua, 2024

Functions
 A function is a special type of a stored program that returns a single value
 You can use a stored function in SQL statements wherever an expression is used
 The syntax for creating a new stored function in MySQL:
DELIMITER $$

CREATE FUNCTION function_name ( • For a function, parameters are always

parameter1 datatype, regarded as in parameters
parameter2 datatype, • A function can only return values of type
... {STRING|INTEGER|REAL|DECIMAL}
)
RETURNS datatype
• A deterministic function always returns
[NOT] DETERMINISTIC
BEGIN the same result for the same parameters
-- SQL statements • A non-deterministic function returns
END $$ different results for the same parameters
• MySQL uses the not deterministic option
DELIMITER ; by default

54 Roi Yehoshua, 2024

Function Example
 The following function returns the number of students in a given department

55 Roi Yehoshua, 2024

Calling a Function
 You can call a function inside a SELECT statement:

 Or you can set the result of the function into a session variable:

56 Roi Yehoshua, 2024

Transactions
 A transaction is a sequence of SQL statements that represents a single unit of work
 For example, consider a Bank amount transfer, which involves two operations:
 Withdrawal of money from account A
 Deposit money to Account B
 If the system crashes after subtracting the amount from A but before adding it to B,
the bank balances will be inconsistent

57 Roi Yehoshua, 2024

ACID
 A set of properties of database transactions intended to guarantee data validity
 Atomicity – All the operations of a transaction either succeed completely or fail completely
 Consistency – Execution of a transaction preserves the consistency of the database
 Isolation – Multiple transactions may execute independently without interference
 Durability – The changes made by a successful transaction persist in the database even in the
case of a system failure

58 Roi Yehoshua, 2024

Transactions
 To define a new transaction, use the START TRANSACTION statement
 All the subsequent SQL statements will belong to the transaction
 The transaction must end with one of the following statements:
 COMMIT: The updates performed by the transaction become permanent in the database
 ROLLBACK: All updates performed by the transaction are undone

59 Roi Yehoshua, 2024

Transactions
 In MySQL, transactions need to be written inside a stored procedure
 So you can automatically rollback the transaction whenever an error occurs
 The following statement needs to be written in the beginning of the procedure:
DECLARE EXIT HANDLER FOR SQLEXCEPTION ROLLBACK;

 The following procedure updates a student’s grade in a course

 This procedure needs to update both the takes table and the student’s tot_cred
 These two updates should be performed in a single transaction

60 Roi Yehoshua, 2024

Transaction Example

61 Roi Yehoshua, 2024

Transaction Example

62 Roi Yehoshua, 2024

Transaction Example
 Calling the procedure with valid parameters results in a successful commit:

63 Roi Yehoshua, 2024

Transaction Example
 Calling the procedure with invalid parameters results in a rollback:

64 Roi Yehoshua, 2024

Triggers
 A trigger is a statement that is executed automatically by the database
 as a side effect of performing a modification on the database
 The syntax for creating a trigger in MySQL:
CREATE TRIGGER trigger_name
{BEFORE | AFTER} {INSERT | UPDATE | DELETE} ON
table_name
FOR EACH ROW
BEGIN
trigger_body;
END

 The triggering event can be INSERT, UPDATE or DELETE

 The trigger action time can be either before or after the event
 The statements following FOR EACH ROW execute once for each row affected by the event
 Values of attributes before and after the update can be referenced using OLD and NEW
65 Roi Yehoshua, 2024
Trigger Example
 A trigger that keeps the tot_cred attribute of a student up-to-date when a new grade
is assigned to them in the takes relation
CREATE TRIGGER credits_earned AFTER UPDATE ON takes
FOR EACH ROW
BEGIN
IF NEW.grade <> 'F' AND NEW.grade IS NOT NULL
AND (OLD.grade = 'F' OR OLD.grade IS NULL) THEN
UPDATE student
SET tot_cred = tot_cred + (
SELECT credits FROM course
WHERE course.course_id = NEW.course_id
)
WHERE student.id = NEW.id;
END IF;
END

66 Roi Yehoshua, 2024

Triggers in MySQL
 You can also define a trigger from the MySQL Workbench menu
 Click on the tools icon next to the table
 Click on the Triggers tab and choose the desired trigger type
 Enter the trigger’s code in the code editor

67 Roi Yehoshua, 2024

Triggers in MySQL
 Click Apply

68 Roi Yehoshua, 2024

Triggers in MySQL
 You can view all the triggers in the database using the command show triggers

69 Roi Yehoshua, 2024

When Not to Use Triggers
 In the past, triggers were used for tasks such as
 Maintaining summary data (e.g., total salary of each department)
 Replicating databases by recording changes to special delta relations
 There are better ways of doing these now:
 Databases today provide built-in materialized view facilities to maintain summary data
 Databases provide built-in support for replication
 There are situations where you want to disable execution of triggers, e.g.,
 When loading data from a backup copy
 When replicating updates at a remote site

70 Roi Yehoshua, 2024

Indexes
 An index is a data structure that can speed up queries/searches
 e.g., searching for instructors in a specific department

 Drawbacks of indexes
 Causes updates to the table to become slower (since the index also needs to be updated)
 Storage space

71 Roi Yehoshua, 2024

Creating Indexes
 We create an index using the CREATE INDEX command
CREATE [UNIQUE] INDEX index_name
ON table_name (column1, column2, ...);

 A UNIQUE index creates a constraint such that all values in the index must be distinct
 For example, let’s define an index on the name column in the instructor table:
CREATE INDEX idx_name
ON instructor (name);
 You can see the new index under the Indexes folder of the table:

72 Roi Yehoshua, 2024

Clustered Indexes
 A clustered index defines the physical order in which table records are stored
 There can be only one clustered index per table
 By default a clustered index is created on a primary key column
 Accessing a row through the clustered index is fast
 Since the index search leads directly to the page that contains the row data

73 Roi Yehoshua, 2024