SQL for Beginners

By Anne-Marie Wright

In this article I will be trying to show how sql statements work.

The database model

-------Name Table------------- -------Address Table----------
NameId type Long AddressId Type Long
Surname type String Line1 type string
Firstname type String line2 type string
Middlename type string City type string
Male type boolean ZipCde type string
AddressId type Long

The two tables are linked by the AddressId in a one to many relationship.

This means that there can be many Names linked to one address.

The data in the tables

Name Table

NameID Surname FirstName MiddleName Male AddressId

1 Smith Andrew John true 1
2 Smithe Fred John true 2
3 Wright Anne false 3
4 Jones Emily Anne false 1
5 Wright David Peter true 3

Address Table

AddressId Line1 Line2 City ZipCode

1 A Street London
2 A Road A Town Oxon
3 A House Village Oxon OX1 3ED

So lets get down to the SQL.

SELECT
SYNTAX: SELECT [{tableName}.]{fieldname}[,[{tablename}.] {fieldname}] FROM
{tablename}

This allows us to return all, or a subset, of the data in the tables.

SQL: to return all the fields and records in the name table
SELECT * FROM Name;
Result: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
2 Smithe Fred John true 2
3 Wright Anne false 3
4 Jones Emily Anne false 1
5 Wright David Peter true 3
Or we could only return certain fields

SQL: to return the id, surname, firstname fields of all the records of the name table

SELECT NameId, Surname, FirstName FROM Name;

Result: NameID Surname FirstName
1 Smith Andrew
2 Smithe Fred
3 Wright Anne
4 Jones Emily
5 Wright David

WHERE
To enable us to have a subset of the data we can add a Where clause to the end of the statement.

SQL: to return all the fields, but only the records that contain Smith in the Surname field

SELECT * FROM Name WHERE Surname='Smith';

Result: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1

If we want to do a search.

SQL: to return all the fields, but only the records that start with 'An' in the Firstname table

SELECT * FROM Name WHERE Firstname Like 'An%';

Result: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
3 Wright Anne false 3

There is no limited to the number of fields we can add to the WHERE clause

SQL: to return all fields from the records that have the Surname 'Wright AND that are Male

SELECT * FROM Name WHERE Surname='Wright' AND Male=True;

Result: NameID Surname FirstName MiddleName Male AddressId
5 Wright David Peter true 3

SQL: to return all the records that have the Surname 'Wright' OR that are Male

SELECT * FROM Name WHERE Surname='Wright' OR Male=True;

Result: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
2 Smithe Fred John true 2
3 Wright Anne false 3
5 Wright David Peter true 3
SQL:

SELECT * FROM Name WHERE (Surname Like 'Smith%' AND

MiddleName='John') Or Male=False;
Result: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
2 Smithe Fred John true 2
3 Wright Anne false 3
4 Jones Emily Anne false 1

ORDER BY
It is possible to have the records returned in a certain order

o Ascending order is A to Z, 0 to 9
o Descending order is Z to A, 9 to 0

If nothing is specified then it is sorted into ascending order

SQL: to return the Surname and Firstname from all the records sorted in ascending order by the
Firstname

SELECT Firstname, Surname FROM Name ORDER BY

FirstName;
or
SELECT Firstname, Surname FROM Name ORDER BY FirstName ASC;
Result: FirstName Surname
Andrew Smith
Anne Wright
David Wright
Jones Emily
Fred Smithe

SQL: to return the Surname and Firstname from all the records sorted in descending order by the
Firstname

SELECT Firstname, Surname FROM Name ORDER BY FirstName DESC;

Result: FirstName Surname
Fred Smithe
Jones Emily
David Wright
Anne Wright
Andrew Smith

GROUP BY
It is also possible to Group identical information together, but you have to put the fields that you want
returned. It is not possible to put a * to say the whole table as we have been doing in the previous
examples.

SQL: to return all the Male records and grouping the Middlename fields together, then the Surname and
finally the FirstName

SELECT Surname, Firstname, MiddleName FROM Name WHERE Male=True

GROUP BY Middlename, Surname, Firstname;
Result: Surname FirstName MiddleName
Smith Andrew John
Smithe Fred John
Wright David Peter

What if we want to select the address for the names. For that we need to use a JOIN (The way tables are
joined together in a SQL statement depends on the database so I will give you 2 types Access and Oracle)

Access SQL:

SELECT Address.*, Name.* FROM Address INNER JOIN Name ON

Address.AddressId = Name.AddressId;

Oracle SQL:

SELECT * FROM Address, Name WHERE

Address.AddressID=Name.AddressID
Result: AddressId Line1 Line2 City ZipCode NameId Surname FirstName MiddleName Male AddressId
A
1 London 1 Smith Andrew John true 1
Street
A
1 London 4 Jones Emily Anne false 1
Street
A A
2 Oxon 2 Smithe Fred John true 2
Road Town
A OX1
3 Village Oxon 3 Wright Anne false 3
House 3ED
A OX1
3 Village Oxon 5 Wright David Peter true 3
House 3ED

We can put a Where clause statement on the end

Access SQL:

SELECT Address.*, Name.* FROM Address INNER JOIN Name ON

Address.AddressId = Name.AddressId WHERE Name.Surname='Wright';

Oracle SQL:

SELECT * FROM Address, Name WHERE AddressID=Name.AddressID AND

Name.Surname='Wright';
Result: AddressId Line1 Line2 City ZipCode NameId Surname FirstName MiddleName Male AddressId
A
3 Village Oxon OX1 3ED 3 Wright Anne false 3
House
A
3 Village Oxon OX1 3ED 5 Wright David Peter true 3
House

Modifying records
It's all very well being able to select the records but now we are looking at how to modify them.

The select statements are not going to return any errors if the sql is correct, they might return nothing,
but they will work. The queries that modify records can return errors. You must make sure that all the
fields that must have something in them are populated and that the fields have the correct type of data
(no letters in number fields etc). Otherwise it will not be able to save the record and will return an error.
Another hick-up might be if you had a relationship between two or more tables, you may find that you
cannot add data to one table before having a corresponding record in another table (i.e We have to have
an address in the address table before we can create a record in the Name table to link to it). This could
cause problems with deleting a record as well. There might be records in another table that are joined to
the record you are trying to delete. This again will cause an error and stop the process.

UPDATE
SYNTAX: UPDATE {tablename} SET [{tablename}.]{fieldname}=newvalue WHERE {criteria}

So if we want to change the record.

SQL:

UPDATE Name SET Surname="Dickens" WHERE NameID=3;

Before update: NameID Surname FirstName MiddleName Male AddressId
3 Wright Anne false 3
After update: NameID Surname FirstName MiddleName Male AddressId
3 Dickens Anne false 3

SQL:

UPDATE Name SET Surname="Wright", FirstName="Ann" Where

NameID=3;
Before update: NameID Surname FirstName MiddleName Male AddressId
3 Dickens Anne false 3
After update: NameID Surname FirstName MiddleName Male AddressId
3 Wright Ann</EM< td> false 3

INSERT
OK so now we need to add new records to the table. For this we use the INSERT command

SYNTAX: INSERT INTO {Tablename}({fieldname}[,{fieldname}]) VALUES ({value}[,{value}])

So to add a recordSQL:

INSERT INTO Name(NameId, Surname, FirstName, Male) VALUES

(6, "Davis", "Ivan", true)
Result: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
2 Smithe Fred John true 2
3 Wright Anne false 3
4 Jones Emily Anne false 1
5 Wright David Peter true 3
6 Davis Ivan true

To add records to a table from another table assume we had another table called OtherNames:

OtherNames

Surname Name Age

Green Vicky 12
Black Steve 32
Howells Zara 25

SQL:

INSERT INTO Name(Surname, FirstName) FROM SELECT Surname, Name

FROM OtherNames;
Result: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
2 Smithe Fred John true 2
3 Wright Anne false 3
4 Jones Emily Anne false 1
5 Wright David Peter true 3
Green Vicky
Black Steve
Howells Zara

Notice the select statement it is wirtten in just the same way as if it was a SQL query on its own. So we
could have had a subset of OtherNames added to the Name table.

i.e SQL:

INSERT INTO Name(Surname, FirstName) FROM SELECT Surname, Name

From OtherNames WHERE Age < 30;
Result: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
2 Smithe Fred John true 2
3 Wright Anne false 3
4 Jones Emily Anne false 1
5 Wright David Peter true 3
Green Vicky
Howells Zara

DELETE
Deleting a record This is acheived by using the DELETE command

SYNTAX: DELETE FROM {TableName} WHERE {criteria}

So if we wanted a table with just the women in we could use the following:

SQL:

DELETE FROM Names WHERE Male=true;

Before Delete: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
2 Smithe Fred John true 2
3 Wright Anne false 3
4 Jones Emily Anne false 1
5 Wright David Peter true 3
After Delete: NameID Surname FirstName MiddleName Male AddressId
3 Wright Anne false 3
4 Jones Emily Anne false 1

Or if we wanted to delete just one record

SQL:

DELETE FROM Names WHERE NameId=3

Before Delete: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
2 Smithe Fred John true 2
3 Wright Anne false 3
4 Jones Emily Anne false 1
5 Wright David Peter true 3
After Delete: NameID Surname FirstName MiddleName Male AddressId
1 Smith Andrew John true 1
2 Smithe Fred John true 2
4 Jones Emily Anne false 1
5 Wright David Peter true 3

This I hope has given you a simple idea of how SQL works. There is a lot more too it, but knowing this
should allow you to create small database applications.