UNIT 7 STRUCTURED QUERY LANGUAGE

7.0 Introduction
7.1 Objectives
7.2 Introduction to SQL
7.3 Data Definition Language
7.4 Data Manipulation Language
7.4.1 Data insertion, Updating and Deletion
7.4.2 Data Retrieval
7.5 GROUP BY Clause and Aggregate functions
7.6 Data Control Language
7.7 Summary
7.8 Solutions/Answers

7.0 INTRODUCTION
As discussed in the previous Units, a relational database system consists of relations, which are
normalised to minimise redundancy. The normalisation process involves the concepts of functional
dependency (FD), multi-valued dependency (MVD) and join dependency (JD). The normalisation results
in the decomposition of tables into smaller but non-redundant relations. After designing the normalised
database system, you would like to implement it by using software called relational database management
system (RDBMS). RDBMSs were designed to manage relational data. Some of these RDBMSs are –
Oracle, DB2, SQL Server, MySQL, etc. All these RDBMSs are designed and developed by different
organisations and use different data storage formats. A structured query language (SQL) is one of the
standards, which was created for the transfer of information from a RDBMS. The SQL consists of three
basic languages, viz. Data Definition Language (DDL), which is used for defining the structure of the
relations in the form of SQL tables and indexes; Data Manipulation Language, which is used for input,
modification and deletion of information in SQL tables; and Data Control Language, which is used for
controlling the access rights on the data of SQL tables and indexes. This unit introduces these three
languages and the basic set of commands used in SQL.
You must learn SQL to use database technologies, therefore, you are advised to go through this unit very
carefully. You must practice the concepts learnt in this unit on a commercial RDBMS.

7.1 OBJECTIVES
After going through this unit, you should be able to:
• create SQL objects (tables, indexes, etc.) from a database schema;
• insert data into database tables using SQL commands;
• retrieve data from a database using SQL queries;
• use the Group By and Having clauses of SQL;
• Using aggregate functions of SQL;
• Create access rights on different database objects using SQL.

7.2 INTRODUCTION TO SQL

The Structured Query Language (SQL) was initially designed at IBM by D D Chamberlin and R F Boyce
for the relational model that was proposed in 1970 by E F Codd. SQL was designed as a fourth-generation
programming language to produce multiple record output as a result of a single command. In addition,
SQL commands did not require the specification of indexes or other information for data retrieval. Later,
SQL was standardised by the American National Standard Institute (ANSI) and a number of standards of
SQL have been created, starting from SQL:86, which was proposed in 1986, SQL:92, SQL:99 and
SQL:2003. Several modifications have been proposed in SQL:2003 in the years 2006-2019. One of the
key advantages of using SQL is that it is the query language of most RDBMSs. Thus, it facilitates the
migration of a database from one RDBMS to another RDBMS. However, some differences can be found
in the SQL implementation on several RDBMSs. This unit covers some of the basic features of SQL only.

Why did SQL become popular? One of the major reasons for SQL’s popularity is that it allows you to
write queries without specifying how the query would be executed. For example, in case you want to
JOIN three relations or tables, say A, B, C, then SQL allows you to join these tables without specifying
the sequence of joining. The decision about how to execute the joins, e.g. (A JOIN B) JOIN C or A JOIN
(B JOIN C), what indexes and views are to be used, etc.; are left to the query parser, translator, and query
optimiser of RDBMS. In addition, the syntax of SQL is closer to the English language. Thus, SQL queries
are simpler to write and run. The following are the features of SQL in the context of RDBMS:
• SQL is non-procedural, as in SQL you just need to say what information is required by you and NOT
how that information is to be acquired from the database.
• The syntax of SQL is closer to the English Language; thus, it is very easy to comprehend.
• In general, the output of a SQL query may be a single record or a group of records.
• An interesting feature of SQL is that it can be used at different levels of ANSI’s three-level architecture
of database system.
SQL includes the following three sub-languages:
• Data Definition Language (DDL): The basic focus of DDL is the commands that can be used to convert
database schema into SQL tables, indexes, etc., or change of structure of the tables. These commands
also allow you to define the constraints on the attributes of a table. The DDL commands are explained
in the next section.
• Data manipulation language (DML): The purpose of data manipulation is twofold. First, it has
commands to input, modify and delete the data in the SQL tables. Second, it has the SELECT command,
which helps in the retrieval of data from one or multiple tables.
• Data control language (DCL): The purpose of these commands is to specify user’s privileges to
database users. These commands are very important in client/server databases with different types of
users.

7.3 DATA DEFINITION LANGUAGE

The design of a database consists of the physical, conceptual, and external schema. To implement this
design, you need to define these schemas using SQL. In this unit, we use commands to define the
conceptual schema. The external schema-related commands are discussed in the next unit. As far as the
physical schema is concerned, we will cover only a few commands. The conceptual schema, which
includes the relational schema and attributes, is implemented in RDBMS as a set of tables and columns,
whereas the tuples of the relations are implemented with the help of rows or data records.

We will explain some of the basic DDL commands in this section with the help of an example. Consider
two relations of a UNIVERSITY database system, namely STUDENT and PROGRAMME.

The student relation (STUDENT) has the following data: student ID (STID), which is also the primary
key, the name of the student (STNAME), the programme code in which that student is registered
(PROGCODE) and the mobile phone number of the student (STMOBILE). Please note that this schema
assumes that a student is allowed to register in only a single programme for which s/he is allotted a
unique student ID.

The second relation, PROGRAMME, has the following data: programme code (PROGCODE), which is
also the primary key, the name of the programme (PROGNAME) and the fee for that programme.

STUDENT (STID, STNAME, PROGCODE, STMOBILE) and

PROGRAMME (PROGCODE, PROGNAME, FEE)

The first step would be to define the datatypes of various attributes. We propose to use the following data
types for the attributes.
Relation Name STUDENT
Attribute STID STNAME PROGCODE STMOBILE
Data type Character Character Character Number
Length 4 40 6 12 digits
Constraint PRIMARY KEY NOT NULL FOREIGN KEY -
References
PROGRAMME
table

Relation Name PROGRAMME

Attribute PROGCODE PROGNAME FEE
Data type Character Character DECIMAL
Length 6 30 5
Constraint PRIMARY KEY NOT NULL >1000 and < 50000

Figure 1: Example Relations

Once you have completed the data design, the next step would be to use SQL to create the SQL tables. To
do so, you should learn about different data types in SQL, commands for creating tables in SQL,
commands for creating constraints, etc. Let us discuss each of these.
Data Types in SQL: SQL supports many data types. Figure 2 lists some of the commonly used data
types of SQL. For more data types supported by a DBMS, you may refer to the information manual of
that DBMS. You may select one of these data types for each column.

CHAR (n) It accepts a character string of size n. The character string can include alphabets,
numeric digits, and special characters. The size of each string is fixed, that is, n.
 VARCHAR (n) It accepts a character string up to the size n. The character string can include
alphabets, numeric digits, and special characters.
BOOLEAN It accepts a value False (zero value) or True (non-zero value)
INTEGER (n) It can store signed integers or unsigned integers of length 32 bits. n represents the
display width of the integer.
DECIMAL (n, d) It can store decimal numbers of size n having d digits after the decimal point. The
default value of d is 0.
DATE Represents a valid date. It uses 4 digits for years and 2 digits each for month and
day.
TIMESTAMP It assigns a timestamp, which can be used to determine the recency of data.
Figure 2: Some Basic Data Types
For the relations of Figure 1, you may select CHAR for STID, PROGCODE, STMOBILE (as it is not
required for computation), and VARCHAR for STNAME and PROGNAME, as the names of students
may be of different lengths; so are the names of the programme’s. The data type of FEE is DECIMAL(5).
Creating the Database and the Tables: In order to create the tables as given in Figure 1, you need to
first create a database using the following command:
CREATE DATABASE <name of the database>;
USE <name of the database>; //This command is needed when the DBMS has multiple databases.
The following commands will create the UNIVERSITY database:
CREATE DATABASE UNIVERSITY;
USE UNIVERSITY;
Now, you can create the tables using the create table command. The following is the syntax of this
command:
CREATE TABLE <name of the table> (
ColumnName1 <data type> [constraints],
ColumnName2 <data type> [constraints],
…
);
The following are the descriptions of the create table command:
• The name of a table should begin with an alphabet. It should not contain blank space and special
characters except the underscore character ( _ ).
• You should not use reserved words as a name of a table.
• You should use a unique name for each column in a table.
• The constraints are optional and therefore, shown in [ ] brackets.
• The data type should include the size of the column.
• You may use any of the following constraints on the column:

NOT NULL This column of a table cannot be left blank while data entry or
modification.
UNIQUE This value should be unique across all the values in this column.
PRIMARY KEY The column is or is a part of the primary key.
CHECK It is followed by certain conditions, which should be fulfilled by the
column.
DEFAULT When a default value is specified for a column.
REFERENCES This is used for specifying referential constrains, while creating a table.
Figure 3: Some of the constraints in Create Table Command
Now, you are ready to create the tables. Since the table STUDENT contains a reference to
PROGRAMME table, therefore, you may create the PROGRAMME table first using the following
command:
CREATE TABLE PROGRAMME
(
PROGCODE CHAR(6) PRIMARY KEY,
PROGNAME VARCHAR(30) NOT NULL,
FEE DECIMAL (5),
CHECK (FEE>1000 AND FEE<50000)
);
Now, you are ready to create the STUDENT table. You will use the following command to create the
table.
CREATE TABLE STUDENT
(
STID CHAR(4) PRIMARY KEY,
STNAME VARCHAR(40) NOT NULL,
PROGCODE CHAR (6),
STMOBILE CHAR (12),
FOREIGN KEY PROGCODE REFERENCES PROGRAMME (PROGCODE)
ON DELETE RESTRICT
);

The Referential Action: Please note that in the command of creating the STUDENT table, we have used
a referential action RESTRICT, which will make sure that any deletion of a record in PROGRAMME table
will be restricted in case even one record of STUDENT table contains that PROGCODE. Thus, this action
will ensure that there is no violation of the referential integrity constraint during deletion of a record from
the PROGRAMME table. The other possible referential action is CASCADE. In this case, the deletion of
a record in PROGRAMME table will result in the deletion of all the records of all the students whose
PROGCODE is the same as the record, which is getting deleted in the PROGRAMME table.
Creating an Index: You may notice that the primary key of the STUDENT table is STID, therefore, the
STUDENT table would be organised in the order of STID. However, many database queries may require
the student records in the order of PROGCODE. This would require you to create an index on PROGCODE
in the STUDENT table to enhance the query performance. The following command can be used to create
an index:
CREATE [UNIQUE] INDEX <name of the index>
ON <name of the table> (ColumnName1, ColumnName2, …)
You use the keyword UNIQUE, only if a unique index is to be created. For the given example, you may
create the following index for the STUDENT table.
CREATE INDEX PROGINDEX ON STUDENT (PROGCODE);
Other DDL commands: There are a large number of DDL commands. We will discuss only a few
commands here. You may refer to the DBMS documentation for more commands.
 Commands to alter a Table: For altering a table, you may use an ALTER TABLE command. This
command can be used for performing the following functions:
• You can add a new column to an existing table, or you can modify a column of an existing
table, by using the command:
ALTER TABLE <name of the table> ADD/MODIFY (ColumnName1, <datatype>, …);
• You can add a new constraint in a table using the following command:
ALTER TABLE <name of the table> ADD CONSTRAINT
<name of the constraint> <type of the constraint> (ColumnName);
• You can drop a constraint on a table or enable or disable it using the following command:
ALTER TABLE <name of the table> DROP/ENABLE/DISABLE <name of the constraint>;
Commands to delete a Table or an Index: You can remove a table or an index, which is not required
any more. The following commands can be used for these purposes.
DROP TABLE<name of the table>;
DROP INDEX <name of the Index> ON <name of the table>;
Commands to create a Domain: As defined earlier, SQL has a large set of data types. However, in
many database implementations, you need to create a more meaningful domain that can define the
data types and constraints on the data of a specific attribute or column. The following command
can be used to create domains. It may be noted that the following command may not be defined in
many DBMSs.
CREATE DOMAIN <Name of the Domain> AS <Data type> CHECK (<Constraints>);
You may refer to the DBMS documentation for more details.

Check Your Progress 1

1) List the advantages and disadvantages of using SQL.

…………………………………………………………………………………….……………………
……………………………………………………………….…………………………………………
…………………………………………………………………………………….……………………
……………………………………………………………….…………………………………………
2) Consider a hotel that has three tables: ROOM (RNo, RType, RRent,), BOOKING (CustID, RNo,
BookedFrom, BookedTo) and CUSTOMER (CustID, CustName, CustAddress, CustPhone). Assume
the structure of these tables and create the tables using SQL. You may assume the following constraints
for the HOTEL database:
• The type of hotel rooms can be: Single Room and Double Room, default room type should be
Single room.
• The room rent is between 5000 per day to 25000 per day.
• Rooms are on different floors and numbered from 101 to 150.
• While booking the room the BookedFrom and BookedTo should follow the following
relationship: Today’s Date <= BookedFrom <= BookedTo
• No room should be booked twice for a specific date.
 …………………………………………………………………………………….……………………
……………………………………………………………….…………………………………………
………………………………………….………………………………………………………………
…………………….…………………………………………………………………………………….
……………………………………………………………………………………..

7.4 DATA MANIPULATION LANGUAGE

Once you have created a database and database tables along with the necessary constraints, the next step
is to insert data in the tables. While inserting the data in the tables, you may commit some mistakes or
there may be the need of making certain changes in the data of the tables, therefore, you would be
requiring SQL commands to INSERT, UPDATE and DELETE records in a database table. These are
Data manipulation language (DML) commands. These DML commands allow you to input and edit the
data in the tables. Further, you would like to retrieve selected information from the database. In this
section, first, we discuss the command to insert, update or delete the data followed by commands to
retrieve information from the database. You may please note that the changes made by the DML
statements are made permanent only after these operations are COMMITTED. You will learn about
COMMIT in Unit 9.
7.4.1 Data Insertion, Updating and Deletion
The DML commands are used for inputting and editing data in a database table. To insert data in a table,
you may use the insert command, which is explained next.
Inserting Data: The following command is used to insert a record into a table. The following two formats
of insert commands are used:
If you are inserting a record that had data for all the columns, then you can simply use the command format:
INSERT INTO <name of the table> VALUES (v1, v2, v3, …);
Please note that v1, v2, etc. are the values that are to be inserted into the respective column of the database.
For example, to insert data into the PROGRAMME table, you can use the following INSERT command:
INSERT INTO PROGRAMME
VALUES (“PGDCA”, “Postgraduate Diploma in Computer Applications”, 22000);
Please note the following with respect to the insert command, given above:
• The values inserted into the table would be PROGCODE as PGDCA; PROGNAME as Postgraduate
Diploma in Computer Applications; and FEE as 22000.
• You can use parameters instead of actual values, for example, you can use INSERT INTO
PROGRAMME VALUES (&1, &2, &3); These parameter values can be input at the time of
execution of the query.
• You can use a sub-query (which will be explained in the next unit) instead of the values given in the
command in the (…).
However, if you do not want to insert values in all the columns of the table, then you need to use the
following format:
INSERT INTO <name of the table> (C1, C2, C3, …, Cn) VALUES (v1, v2, v3, …, vn);
Here, C1, C2, … represents the column names and v1, v2, … represents the corresponding value of a
column. Please note that you need to specify n values for the n columns. For example, suppose you do not
know the mobile number of a new student, still his/her information can be entered in the STUDENT table
using the following SQL command:
INSERT INTO STUDENT (STID, STNAME, PROGCODE)
VALUES (“1001”, “RAMESH SHARMA”, “PGDCA”);
Please note the following with respect to the insert command, given above:
• The insertion is possible as STMOBILE has no constraint. It can be left blank.
• The values inserted into the table would be STID as 1001; STNAME as RAMESH SHARMA;
PROGCODE as PGDCA; and STMOBILE as NULL.
• Please also note that the above insert statement will not cause a violation of foreign key constraint,
as we have already inserted the PGDCA programme details in the PROGRAMME table.
Updating Data: For updating data, you may use the update command of SQL. The format of the command
is given below:
UPDATE <name of the table>
SET <C1> = <v1>
WHERE <conditional statement>;

For example, to update the mobile number data of student 1001 to the number, say 8484848484, you can use the
command:
UPDATE STUDENT
SET STMOBILE = “8484848484”
WHERE STID = “1001”;

You can also use a subquery in an update statement (subqueries are covered in unit 8). For example,
consider the fee of all the programmes, which has more than 100 students, is raised by 5%, then the
following update command may be used to update the PROGRAMME table.
UPDATE PROGRAMME
SET FEE = FEE * 1.05
WHERE PROGRAMME.PROGCODE IN ( SELECT STUDENT.PROGCODE
FROM STUDENT
GROUP BY STUDENT.PROGCODE
HAVING COUNT(*) > 100);

The purpose of this subquery will be clear after you go through the next subsection. The subquery will find
those programmes which have more than 100 students.

Deleting Data: You can use the following command to delete one or more records from a table:
DELETE FROM <name of the table>
WHERE <conditional statement>;
The delete command may delete one or more data records at a time. For example, you can try deleting the
data of the PGDCA programme from your database, using the following command.
DELETE FROM PROGRAMME
WHERE PROGCODE= “PGDCA”;
However, as there exists a foreign key constraint with referential action RESTRICT and you have already
inserted a student who has PGDCA as his programme, therefore, DBMS will not allow you to delete the
programme PGDCA from the PROGRAMME table. In case, you want to do so you need to first delete
records of all the students of PGDCA from the STUDENT table and then you can delete the PGDCA
programme from the PROGRAMME table. Please note that you can use subqueries instead of conditional
statement in deletion also.
7.4.2 Data Retrieval
One of the most popular features of any DBMS is the ad-hoc query facility, which requires data retrieval
as per the need and access rights of the user. SELECT statement is one of the most used statements of
DML, as it helps in the retrieval of requisite data. In this section, we discuss various clauses of this
statement.
SELECT Statement: The following is the basic format of the select statement.

SELECT <List of Column names or expressions to be displayed>

FROM <List of Tables that contain the data, whose columns are in select>
WHERE <Conditions for selection of records for display> ;
The following are examples of the use of this statement for the retrieval of data from the two relations given
in Figure 1.
Example 1: List the details of all the programmes of the University.
For answering this query, are using one wildcard character (*), which represents all the columns of a table.
Please note that in case, you have used * in an arithmetic expression in the SELECT clause, it will be treated
as a multiplication sign.
SELECT *
FROM PROGRAMME;
This statement will display the PROGCODE, PROGNAME and FEE of all the records in the
PROGRAMME table.
Example 2: List the programme code and programme names of all the programmes of the university,
whose fee is less than or equal to 5000.
SELECT PROGCODE, PROGNAME
FROM PROGRAMME
WHERE FEE <= 5000 ;
Please note the select clause now contains only those column names, which are to be displayed.
Example 3: List the id, name and mobile number of all the PGDCA students.
SELECT STID, STNAME, STMOBILE
FROM STUDENT
WHERE PROGCODE = “PGDCA” ;
Example 4: This example shows the use of the arithmetic operator and alias. What would be the fee of
the PGDCA programme, if students are given a discount of 15%?
SELECT PROGCODE, PROGNAME, FEE*0.85 AS DISCOUNTEDFEE
FROM PROGRAMME
WHERE PROGCODE = “PGDCA” ;
Please note the computation of the fee and assigning the result of the computation a new name, that is an
alias, DISCOUNTEDFEE. Please note that the syntax of assigning an alias may be different in different
RDBMS.
Even in SQL expressions, the operator precedence is followed. Further, these expressions are executed from
the left towards the right.
Example 5: This example shows the use of the concatenation operator. If you want to display the
names of the programme as: “PGDCA: Postgraduate Diploma in Computer Applications” programme,
you should use the following SQL command.
SELECT PROGCODE + “:” + PROGNAME
FROM PROGRAMME
WHERE PROGCODE = “PGDCA” ;
Please note that in this statement the concatenate operator is ‘+’. However, in different DBMSs it may be
different, for example, MySQL uses CONCAT( ) function, whereas Oracle uses || as a concatenation
operator. In addition, please also note the use of the literal character string “:” in the command. A literal
string may not be part of any column value but can be added to enhance information display.

Example 6: This example demonstrates how duplicate rows can be eliminated using the DISTINCT
operator. Find the programme code of those programmes, which has at least one student.
To find the answer to this query, you may use the STUDENT table and Project it on programme code.
SELECT DISTINCT PROGCODE
FROM STUDENT
In case you do not use DISTINCT then you will get duplicate values of programme code.
Example 7: This example demonstrates the use of the range operator BETWEEN … AND. To find the list
of programmes whose fee is >=5000 but <= 15000, you may use the following command:
SELECT *
FROM PROGRAMME
WHERE FEE BETWEEN 5000 AND 15000;
Please note that both the values, viz. 5000 and 15000 are included in the range.
Example 8: This example demonstrates the use of the set operator IN. Find the students of PGDCA or
BCA programmes. One way of answering that query would be:
SELECT *
FROM STUDENT
WHERE PROGCODE IN (“BCA”, “PGDCA”) ;
Example 9: This example demonstrates the use of LIKE operator for matching a pattern of characters in
the columns which are of CHAR or VARCHAR type. It may be noted that LIKE operator is supported by
several wildcard characters, which may be different in different DBMS. In this example, we demonstrate
the use of % wildcard character that matches zero or many characters in a string. For example, a string like
%COM% will match with strings: COMPUTER, COMMERCE, INCOME, MCOM etc.
To find the list of all the programmes, which have word “Computer”, you may give the command:
SELECT *
FROM PROGRAMME
WHERE PROGNAME LIKE “%Computer%”;
Example 10: This example demonstrates the use of IS NULL operator. Find the list of students, whose
mobile number is not with the University.
SELECT *
FROM STUDENT
WHERE STMOBILE IS NULL ;
Example 11: SQL uses the logical operators, i.e., NOT, AND and OR. The precedence of these operators
is shown in the following table:

Operators in increasing order of

Precedence
Comparison operators (e.g. <, =, >
etc.)
NOT
AND
OR

To print the name of all the students of PGDCA or BCA can also be written as:
SELECT STID, STNAME
FROM STUDENT
WHERE PROGCODE = “BCA” OR PROGCODE = “PGDCA” ;

Ordering of Results using ORDER BY clause

The SELECT statement of SQL, in addition to SELECT, FROM and WHERE clauses, supports three more
clauses, viz. ORDER BY clause, GROUP BY clause and HAVING clause. We will discuss the ORDER
BY clause here and the remaining two clauses are discussed in the next section.
The ORDER BY clause is used, when you want to display the records in the sorted sequence of one or more
columns. This sorted order can be increasing or decreasing. Also, you can use more than one column for
sorting the data. Please note the following points about ORDER BY clause:
• ORDER BY is the last clause of a SELECT statement.
• You can mention ASC for ascending order or DESC for descending order. In case you do not mention
any order, then ordering, by default, is in ascending order.
• You can sort the records on a column, which is part of the table given in FROM statement, even if you
have not displayed that column.
• You can also sort on the alias that has been created by you in the select statement.
Example 12: List the name of the students in the order of programme and within a programme in
alphabetical order.
SELECT PROGCODE, STNAME
FROM STUDENT
ORDER BY PROGCODE, STNAME ;

Check Your Progress 2

1) Write the SQL commands to insert the following data in the STUDENT and PROGRAMME
table. Highlight the errors, if any.
 Relation Name PROGRAMME
PROGCODE PROGNAME FEE
BCA Bachelor of 48000
Computer
Applications
MCA Master of 60000
Computer
Applications

Relation Name STUDENT

STID STNAME PROGCODE STMOBILE
1002 Abc BCA
1003 NULL PGDCA 91999999999
1004 XYZ MCA
……………………………………………………………………………………..……………………
………………………………………………………………..…………………………………………
………………………………………….………………………………………………………………

2) List the various clauses of the SELECT statement giving their purpose.
……………………………………………………………………………………..……………………
………………………………………………………………..………………………………………...
3) Consider the following two relations – S and SP.
S
SupplierNo SupplierName SupplierRanking SupplierCity
S1 ABC 60 Delhi
S2 DEF 30 Mumbai
S3 EFG 50 Bangaluru
S4 FGH 40 Chennai
S5 GHI NULL Delhi
SP
SupplierNo PartNo QuantitySold
S1 PA 1000
S2 PA 500
S2 PB 1200
S3 PB 700
S5 PA 2100
S5 PB 1200

a) Find the name of the suppliers, whose ranking is lower than 40 and the city of the supplier is
Chennai.
b) List the names of the suppliers of Delhi city in the decreasing order of the supplier ranking.
c) List the names of the supplier who have supplied the part PB (use IN operator).
d) List the codes of all the suppliers who have supplied at least one part.
 e) List all the suppliers, who have “EF” in their names.
f) Get part numbers for parts whose quantity sold in a supply is greater than 1000 or are supplied
by S2. (Hint: It is a retrieval using union).
g) List the names of the suppliers, whose ranking is not given.

…………………………………………………………………………………………………………………………
…………………………………………………………………………………………………………..

7.5 GROUP BY CLAUSE AND AGGREGATE FUNCTIONS

In the previous section, you have gone through the concept of data manipulation language (DML). We
have discussed the SELECT statement and its various clauses. In a database system, several queries
require DBMS to produce information about a group. For example, you may be interested in finding the
average marks of the group of PGDCA students vis-à-vis BCA students. The SQL supports a GROUP BY
clause for such cases. In addition, SQL also supports a number of functions that can find aggregate
information for a group of records. These functions are required to find the sum, average, counting of
records etc. These are called aggregate functions. The following table defines some of the important
aggregate functions used in SQL.
count Used to count the number of records
sum Finds the sum of the data of a column
avg Finds the average of the data of a column
max Finds the maximum value from the data of a column
min Find the minimum value from the data of a column
Figure 3: Some Aggregate functions in SQL
Let us explain the use of these functions with the help of a few examples.
Example 13: Find the number of students of PGDCA.
SELECT COUNT(*)
FROM STUDENT
WHERE PROGCODE = “PGDCA”;
Please note that the wildcard *, in this case, indicates all the records to be counted, which fulfil the
WHERE condition.

Example 14: Find the minimum, maximum and average fees of all the programmes.
SELECT MIN(FEE), MAX(FEE), AVG(FEE)
FROM PROGRAMME
GROUP BY clause
GROUP BY clause can be used to group records on certain criteria, e.g., you can group students on the
basis of their Programmes. This clause is added after WHERE clause in the SELECT statement. In a
SELECT statement in which you have used a GROUP BY clause, you can only use the aggregate functions
or the column name on which you have grouped the data in the SELECT clause. The following example
demonstrates this aspect:
Example 15: Find the number of students in every programme of the University.
You can answer this query by grouping the records on PROGCODE and finding the count of the student
ids in each group.
SELECT PROGCODE, COUNT(STID)
FROM STUDENT
GROUP BY PROGCODE;
Please note that in the SELECT clause of the SELECT statement above, we have used only the
PROGCODE and the aggregate function COUNT.
HAVING clause
The HAVING clause can be used to specify a condition for a group. It is different from the WHERE
clause, which is applicable for all the records, whereas the condition specified in the HAVING clause is
to be fulfilled by a group of data. The following example explains the use of the HAVING clause.
Example 15: Count the number of students in each programme, where the mobile number of students is
not given. List only those programmes which have more than 5 such Students.
SELECT PROGCODE, COUNT(STID)
FROM STUDENT
WHERE STMOBILE IS NULL
GROUP BY PROGCODE
HAVING COUNT(STID) < 5;
Please note that in the SELECT clause of the SELECT statement above, we have used only the
PROGCODE and the aggregate function COUNT.

7.6 DATA CONTROL LANGUAGE

The purpose of data control language (DCL) is to create users and assign access rights to them. In general,
these commands are executed by a database administrator. The following are some of the most used DCL
commands.
Creating a new user: You can create a new user using the following command:
CREATE USER < username for database user> IDENTIFIED BY < Password for the user>
For example, you can create a new user with the username “PGDCA_Student” with the password
“PGDCA123”
CREATE USER PGDCA_Student IDENTIFIED BY PGDCA123
Use of GRANT Command: GRANT is used to give different kinds of accesses to a database user. Block3
covers the basic aspects of the GRANT option. In general, SQL supports two kinds of access permissions:
• The permissions that are at the system level.
• The permissions at the level of an object, such as a table, record, column etc.
The system-level permissions are, in general, specific to the DBMS environment, therefore, you may refer
to the system documentation for details on such permissions. In this section, we provide a basic introduction
to object-level permissions with the help of examples.
To give permission to get information from a table, the following SQL command may be used.
GRANT SELECT ON STUDENT, PROGRAMME TO PGDCA_Student;
To give permission for inserting and updating a record in the STUDENT table, you may use the following
SQL command:
GRANT INSERT, UPDATE ON STUDENT TO PGDCA_Student;
In case you want to GRANT the SELECT access rights to more than one user on STUDENT table, then
you may use the following command:
GRANT SELECT ON STUDENT TO PGDCA_Student1, PGDCA_Student2;
Use of REVOKE command: The REVOKE command is used to remove the access permissions that were
granted to a user.
For example, you can also revoke SELECT permission using the following command:
REVOKE SELECT ON STUDENT FROM PGDCA_Student;
The following command will revoke all permissions of a user on STUDENT table.
REVOKE ALL ON STUDENT FROM PGDCA_Student;
Use of DROP command: You can remove a user using the DROP command.
For example, you can drop PGDCA_Student using the following SQL command:
DROP USER PGDCA_Student;

Check Your Progress 3

Consider the supplier relations given in Question 3 of Check Your Progress 2.
1) Write the SQL commands for the following aggregate operations.
a) Find the total supply of part PA.
b) Count the number of suppliers who have made a supply.
c) List the part number and the total quantity supplied for that part.

……………………………………………………………………………………………………………………
……………………………………………………………………

2) Write the SQL commands for the following queries:

a) List the suppliers who have made more than one supply.
b) List the suppliers who have made more than one supply, with each supply being more than
1000.
c) Find the part number and maximum quantity supplied, for the parts for which the total quantity
supplied for that part is more than 1000. You may order the result in ascending order of part
numbers.

……………………………………………………………………………………………………………………
……………………………………………………………………

3) Write the SQL commands for the following:

a) Create a new user ABC having the password XYZ.
b) Grant the user ABC to read table S only.
c) Cancel all the access permissions of ABC on table S.
 ……………………………………………………………………………………………………………………
……………………………………………………………………

7.7 SUMMARY

SQL is one of the most important languages for any RDBMS. It allows a user to create tables, insert and
update data in the tables and retrieve information from the tables. In addition, data of a table can be made
available to only authorised users. This unit first introduces you to basic aspects of SQL and thereafter
discusses the DDL commands. It is important that while creating tables you also include the constraints
that are to be fulfilled by the tables. The constraints in SQL may be implemented using CHECK clause,
PRIMARY KEY clause, FOREIGN KEY clause etc. You must also implement the referential action in
case of referential integrity. You can also alter the tables if needed. This unit discusses the DDL
commands for all the above operations. After creating the table using the DDL commands, next you use
the DML commands to insert data into the tables. In case of any changes in the data values in the table,
you may use the UPDATE command of DML. The commands for data insertion and updating have been
discussed in the unit. One of the most important DML commands is SELECT which allows the retrieval
of data from the table based on various criteria. This unit discusses various clauses of SELECT statement,
viz., SELECT, FROM, WHERE, GROUP BY, HAVING and ORDER BY clauses. Finally, the unit
discusses the CREATE USER, GRANT, REVOKE and DROP commands of DCL. You may please note
that this unit does not cover all the SQL commands. You must practice these commands and learn more
SQL commands from the user documentation of DBMS that you may use.

7.8 SOLUTIONS/ANSWERS
Check Your Progress 1

1) The following are the advantages of using SQL:

• It is a standard query language across RDBMSs; therefore, you can port your data from one DBMS
to another. Also, you just need to master SQL to use any RDBMS.
• SQL is more English-like, so is easy to learn and use.
• SQL exists for both interactive environments and as embedded SQL. Thus, you can use SQL for
ad-hoc queries as well as in programs, like web-based programs.
Some of the disadvantages of SQL are:
• A query can be implemented in many ways, this may confuse a user.
• The language, in its present form, is very large.
• Some of the functions of the SQL are not portable across DBMS.
• The result of an SQL command may allow duplicates, which are not relational in nature.

2) In this implementation, a lot of constraints have been defined, so instead of directly putting them in
tables, first, we define the domains and use these domains in the CREATE table command. This will
be a neater and more maintainable implementation.
CREATE DOMAIN TYPEOFROOM AS CHAR (1) CHECK (VALUE IN (S, D));
// This makes the domain based on the first constraint. S represents a single room and D
represents a double room.
CREATE DOMAIN ROOMRENT AS DECIMAL(5, 0)
CHECK( VALUE BETWEEN 5000 AND 25000);
// Implements the second constraint
CREATE DOMAIN ROOMNUMBER AS INTEGER (3)
CHECK (VALUE > = 101 AND VALUE <=151);
//Implements the third constraint. Now you are ready to create the Tables.
CREATE TABLE ROOM (
RNo ROOMNUMBER PRIMARY KEY,
RType TYPEOFROOM NOT NULL DEFAULT S,
RRent ROOMRENT NOT NULL,
);
// Default room type stated in the first constraint is implemented in this statement.
// Next, we create the CUSTOMER table, BOOKING table will have references to both ROOM and
CUSTOMER tables.
CREATE TABLE CUSTOMER (
CustID CHAR(5) PRIMARY KEY,
CustName VARCHAR(40) NOT NULL,
CustAddress VARCHAR(60) NOT NULL
CustPhone CHAR (12) NOT NULL
);
CREATE DOMAIN DATEOFBOOKING AS DATETIME
CHECK (VALUE > = GETDATE() );
// Please note that GETDATE() function will output the current date.
// For a different RDBMS, this function may be different.
CREATE TABLE BOOKING (
CustID CHAR(5) NOT NULL,
RNo ROOMNUMBER NOT NULL,
BookedFrom DATEOFBOOKING NOT NULL,
BookedTo DATEOFBOOKING NOT NULL,
PRIMARY KEY (CustID, RNo, BookedFrom),
FOREIGN KEY (RNo) REFERENCES ROOM
ON DELETE RESTRICT ON UPDATE CASCADE,
FOREIGN KEY (CustID) REFERENCES CUSTOMER
ON DELETE RESTRICT ON UPDATE CASCADE,
CONSTRAINT BOOKINGDATEVERIFICATION
CHECK ( NOT EXISTS ( SELECT *
FROM BOOKING X BOOKING Y
WHERE X.RNo = Y.RNo AND
X.BookedFrom > Y.BookedFrom AND
X.BookedFrom < Y.BookedTo)
);
 //Please note this constraint may not work on certain DBMS, where NOT EXISTS clause is not
supported.

Check Your Progress 2

1) You may use the following sequence of instructions:

INSERT INTO PROGRAMME
VALUES (“BCA”, “Bachelor of Computer Applications”, 48000);
The above insertion will be successful.
INSERT INTO PROGRAMME
VALUES (“MCA”, “Master of Computer Applications”, 60000);
The above insertion will fail, as the FEE is more than the allowable fee value.
INSERT INTO STUDENT (STID, STNAME, PROGCODE)
VALUES (“1002”, “Abc”, “PGDCA”);
The above insertion will be successful, as PGDCA was inserted in Section 7.4.1
INSERT INTO STUDENT (STID, PROGCODE, STMOBILE)
VALUES (“1003”, “BCA”, “91999999999”);
The above insertion will fail, as the Student name column value must be given, as it has a
constraint NOT NULL.
INSERT INTO STUDENT (STID, STNAME, PROGCODE)
VALUES (“1003”, “XYZ”, “MCA”);
The above insertion will fail, as the MCA programme insertion had failed earlier, so it
will result in a violation of the referential constraint.
2) The following are the basic clauses of SELECT statements.
SELECT is used to specify the columns or expressions that are to be displayed in the result.
FROM is used to list the tables that are to be used for data output.
WHERE is used to specify conditions for the selection of data for the display.
GROUP BY clause is used to specify the columns, whose values should be used to group the records
of the table.
HAVING is used to specify conditions, which should be fulfilled by each group for selection for
display.
ORDER BY is used to specify the ordering of the records for the output.

3) a) SELECT SupplierName
FROM S
WHERE SupplierRanking < 40 AND SupplierCity = “Chennai”;
The output of this will be:

SupplierName
FGH

b) SELECT SupplierName
FROM S
WHERE SupplierCity = “Delhi”
 ORDER BY SupplierRanking DESC;
The output of this will be:
SupplierName
ABC
GHI

c) SELECT SupplierName
FROM S
WHERE SupplierNo IN (SELECT DISTINCT SupplierNo
FROM SP
WHERE PartNo= “PB”) ;

The output of this will be:

SupplierName
DEF
EFG
GHI

d) SELECT DISTINCT SupplierID

FROM SP;
The output of this will be:

SupplierNo
S1
S2
S3
S5

e) SELECT *
FROM S
WHERE SupplierName LIKE %EF% ;
The output of this will be:
SupplierNo SupplierName SupplierRanking SupplierCity
S2 DEF 30 Mumbai
S3 EFG 50 Bangaluru

f) SELECT DISTINCT PartNo

FROM SP X
WHERE X.QuantitySold > 1000
UNION
(SELECT DISTINCT PartNo
FROM SP Y
WHERE Y.SupplierNO = “S2” ) ;
The output of this will be:
PartNo
 PA
PB

g) SELECT SupplierName
FROM S
WHERE SupplierRanking IS NULL ;
The output of this will be:
SupplierName
GHI

Check Your Progress 3

1)
a) SELECT sum(QuantitySold)
FROM SP
WHERE PartNo = “PA” ;
b) SELECT count(DISTINCT SupplierNo)
FROM SP;
c) SELECT PartNo, sum(QuantitySold)
FROM SP
GROUP BY PartNo

2)
a) SELECT SupplierNo
FROM SP
GROUP BY SupplierNo
HAVING Count(PartNo) > 1 ; // As each supply is of one part only.
b) SELECT SupplierNo
FROM SP
WHERE QuantitySold > 1000
GROUP BY SupplierNo
HAVING Count(PartNo) > 1 ; // As each supply is of one part only.
c) SELECT PartNo, Max(QuantitySold)
FROM SP
GROUP BY PartNo
HAVING Sum(QuantitySold) > 1000
ORDER BY PartNo;

3)
a) CREATE USER ABC IDENTIFIED BY XYZ;
b) GRANT SELECT ON S TO ABC;
c) REVOKE ALL ON S FROM ABC;