12-sql73

7/28/2019 12-SQL73

1/73

1

SQLSQL

(Structured Query Language)(Structured Query Language)

7/28/2019 12-SQL73

2/73

2

Disclaimer & Acknowledgements

Even though Sang Shin is a full-timeemployee of Sun Microsystems, the contentshere are created as his own personal

endeavour and thus does not reflect anyofficial stance of Sun Microsystems.

Sun Microsystems is not responsible for anyinaccuracies in the contents.

7/28/2019 12-SQL73

3/73

3

Revision History 01/24/1998: version 1, created with speaker notes (Edward Haynes) 02/02/2003: version 2, cosmetic change (Sang) 10/10/2004: version 3, cosmetic change (Sang)

7/28/2019 12-SQL73

4/73

4

AgendaWhat is SQL?

SQL Terminology

Creating and Deleting Tables

Select Statement for queryingJoin Statement

Aliases

Sub-query

Comparison, Character match, Numericalexpression

Insert, Update, Delete

SQL Data-types

7/28/2019 12-SQL73

5/73

5

What is SQL?What is SQL?

7/28/2019 12-SQL73

6/73

6

What is SQL?

SQL provides a way to retrieve andmanipulate data in a relational database

SQL is used for all relational databaseoperations, including administration,schema creation and data retrieval

Data definition

Data manipulation

7/28/2019 12-SQL73

7/737

What SQL IS NotSQL is a not a programming language

It does not provide any flow-of-controlprogramming constructs, function definitions,

do-loops, or if-then-else statements Unlike modern programming languages that

enable you to define a data type for a specificpurpose, SQL forces you to choose from a setof predefined data types when you create ormodify a column.

SQL is not a procedural language You use SQL to tell the database what data to

retrieve or modify without telling it how to do it

7/28/2019 12-SQL73

8/738

Similarity with Programming

LanguageThey both usually give you more than oneway to accomplish the same goal

Various SQL statements may achieve the sameresults (but differ in processing efficiency andclarity of code)

7/28/2019 12-SQL73

9/739

Case Sensitivity

SQL is not case sensitive

You can mix uppercase and lowercase

when referencing SQL keywords(SELECT, INSERT, ...), table names andcolumn names

However, case does matter when

referring to the contents of a column; ifyou ask for all the store's where the storename is lower case, but in the databaseall the names are stored in upper case,

you wont retrieve any rows at all

7/28/2019 12-SQL73

10/7310

SQL TerminologySQL Terminology

7/28/2019 12-SQL73

11/7311

SQL TermsTable

A set of rows

Analogous to a file

ColumnA column is analogous to a field of a record

Each column in a given row has a single value

7/28/2019 12-SQL73

12/7312

SQL Terms (Continued)Row

Analogous to a record of a file

All rows of a table have the same set ofcolumns

Primary Key

One of more columns whose contents areunique within a table and thus can be used toidentify a row of that table

7/28/2019 12-SQL73

13/73

13

Creating & DeletingCreating & DeletingTables and otherTables and otherOperations via DDLOperations via DDL

7/28/2019 12-SQL73

14/7314

Data Definition Language (DDL)DDL is used for defining the databasestructure

Create Table, Alter Table, Drop table

Create Index, Drop Index

Create User, Alter User, Drop User

Create Role, Drop RoleCreate Schema, Drop Schema

Create View, Drop View

Create Trigger, Drop Trigger

7/28/2019 12-SQL73

15/7315

CREATE TABLEcreate table store

(store_id NUMERIC,

name VARCHAR(25),city VARCHAR(25),

state VARCHAR(2),

zip VARCHAR(25));

7/28/2019 12-SQL73

16/73

16

ALTER TABLEALTER TABLE store ADD UNIQUE (C1);

ALTER TABLE store ADD ORDER_NUM INT;

ALTER TABLE store ADD CONSTRAINTconstraint_0 FOREIGN KEY (C1)REFERENCES T1 (C1);

ALTER TABLE store DROP ORDER_NUMCASCADE;

ALTER TABLE store RENAME TO T1;

7/28/2019 12-SQL73

17/73

17

DROP TABLEDROP TABLE store;

7/28/2019 12-SQL73

18/73

18

CREATE INDEX & DROP INDEXCREATE INDEX customer_idx

ON "PBPUBLIC"."CUSTOMER_TBL"

( "CUSTOMER_NUM" ASC ,"NAME" ASC )

DROP INDEX ORDER_TBL.ORDER;

7/28/2019 12-SQL73

19/73

19

CREATE/ALTER/DROP USER

CREATE USER PoInT PASSWORD BaSE;

CREATE USER "PoInT" PASSWORD "BaSE";

ALTER USER Scott PASSWORD lion;

ALTER USER Scott DEFAULT ROLE CEO;

DROP USER ENGINEERING_MANAGERCASCADE;

7/28/2019 12-SQL73

20/73

20

CREATE/DROP VIEW

CREATE VIEW customer_order

AS selectorder_num,order_tbl.customer_num,customer

_tbl.nameFROM order_tbl,customer_tbl

WHERE product_num = 10;

DROP VIEW customer_order;

7/28/2019 12-SQL73

21/73

21

CREATE/DROP TRIGGERCREATE TRIGGER trigger2

BEFORE UPDATE ON product_tbl

REFERENCING NEW AS NEWROW

FOR EACH ROW

WHEN (NEWROW.qty_on_hand < 0)

SET NEWROW.qty_on_hand = 0;

DROP TRIGGER trigger2

7/28/2019 12-SQL73

22/73

22

Select StatementSelect Statement

7/28/2019 12-SQL73

23/73

23

SELECT StatementSELECT [ DISTINCT | ALL ]

column_expression1, column_expression2, ....

[ FROM from_clause ][ WHERE where_expression ]

[ GROUP BY expression1, expression2, .... ]

[ HAVING having_expression ]

[ ORDER BY {column | expression} [asc| desc] ]

distinct keyword eliminates duplicates

7/28/2019 12-SQL73

24/73

24

SELECT StatementSelect * from customer_tbl;

7/28/2019 12-SQL73

25/73

25

FROM ClauseComma delimited list of tables to retrievedata from

With or without aliases

7/28/2019 12-SQL73

26/73

26

WHERE ClauseDetermines exactly which rows areretrieved

Qualifications in the where clause

Comparison operators (=,>,=,

7/28/2019 12-SQL73

27/73

27

WHERE Clause

(Continued)Not negates any Boolean expressionsand keywords such as like, null, between

and in

7/28/2019 12-SQL73

28/73

28

Example: WHERE Clauseselect Customer_num, discount_code, zip, name

from customer_tbl

where discount_code in ('N','M','L')

and zip like '33%'

7/28/2019 12-SQL73

29/73

29

GROUP BY ClauseThe group by function organises data intogroups based on the contents of a column(s)

Usually used with an aggregate function in the selectlist

The aggregate is calculated for each group

All null values in the group by column are treated asone group

Grouping can be done by a column_name orby any expression that does not contain anaggregate function

7/28/2019 12-SQL73

30/73

30

GROUP BY Clause Cont.The group by function usually containsall columns and expressions in the

select list that are not aggregates

In conjunction with the where clause,rows are eliminated before they go into

groupsApplies a condition on a table before thegroups are formed

7/28/2019 12-SQL73

31/73

31

Example: GROUP BY

Clauseselect customer_num,

count(customer_num),

sum(quantity)from order_tbl

group by customer_num

7/28/2019 12-SQL73

32/73

32

HAVING ClauseSet conditions on groups

Restricts groups

Applies a condition to the groups afterthey have been formed

having is usually used in conjunction

with an aggregate function

7/28/2019 12-SQL73

33/73

33

Example: HAVING Clauseselect product_num,

quantity,

shipping_costfrom order_tbl

group by product_num,

quantity,

shipping_cost

having shipping_cost >250

7/28/2019 12-SQL73

34/73

34

ORDER BY ClauseThe order by clause sorts the queryresults (in ascending order by default)

Items named in an order by clause neednot appear in the select list

When using order by, nulls are listed first

7/28/2019 12-SQL73

35/73

35

ORDER BY Clause

Exampleselect product_num,

quantity,

shipping_costfrom order_tbl


quantity,

shipping_cost

having shipping_cost >250

order by quantity,

shipping_cost;

7/28/2019 12-SQL73

36/73

36

Join StatementJoin Statement

7/28/2019 12-SQL73

37/73

37

Join StatementRetrieves data from two or more tables

Combines tables by matching valuesfrom rows in each table

Joins are done in the where clause

Columns in join don't have to be in select

7/28/2019 12-SQL73

38/73

38

Join Statement (Continued)If you do not specify how to join rowsfrom different tables, the database

server assumes you want to join everyrow in each table

Cartesian Product

Very costly (cpu time)

May take a while to return large result set

Column names that appear in more thenone table should be prefixed by table

name

7/28/2019 12-SQL73

39/73

39

Example: Join Statementselect c.name,

o.quantity,

o.shipping_costfrom order_tbl o,

customer_tbl c

where o.customer_num =c.customer_num;

7/28/2019 12-SQL73

40/73

40

Outer JoinsWith a join, if one row of a table isunmatched, row is omitted from result

table.The outer join operations retain rowsthat do not satisfy the join condition List branches and properties that are in same

city along with any unmatched branches.

SELECT b.*, p.*

FROM branch1 b LEFT JOIN

property_for_rent1 p ON b.bcity = p.pcity;

7/28/2019 12-SQL73

41/73

41

bno bcity pno pcity

---------------------------------------------

B3 Glasgow PG4 Glasgow

B4 Bristol NULL NULL

B2 London PL94 London

There is a LEFT JOIN and a RIGHT JOINThese can be very vendor specific (check withyour vendor for outer join specifications)

Oracle uses (+) for outer joins, MS SQL Server

uses LEFT OUTER JOIN

Outer Join cont.

7/28/2019 12-SQL73

42/73

42

AliasAlias

7/28/2019 12-SQL73

43/73

43

AliasUse of temporary names for tableswithin a query

Can be used anywhere in query

Reduces amount of typing

ex. select pub_name, title_id

from titles t, publishers p

where t.pub_id = p.pub_id

7/28/2019 12-SQL73

44/73

44

Example: Aliasselect type_code,

city,

regionfrom office_tbl;

7/28/2019 12-SQL73

45/73

45

Example: Alias cont.select t.description,

o.city,

o.region

from office_tbl o,

office_type_code_tbl t

where o.type_code =

t.type_code;

7/28/2019 12-SQL73

46/73

46

Sub-QuerySub-Query

7/28/2019 12-SQL73

47/73

47

Sub-QueriesA select statement, used as anexpression in part of anotherselect,update, insert ordelete

The sub-query is resolved first and theresults are substituted into the outerquery's where clause

Used because it can be quicker to understandthan a join

Perform otherwise impossible tasks (i.e. usingan aggregate)

7/28/2019 12-SQL73

48/73

48

Example: Sub-Queryselect product_num,

description,

purchase_cost

from product_tbl


description,

purchase_cost

having purchase_cost >

(select avg

(shipping_cost)

from order_tbl);

7/28/2019 12-SQL73

49/73

49

EXISTS and NOT EXISTSEXISTS and NOT EXISTS are for useonly with sub-queries.

They produce a simple true/false resultEXISTS is true if and only if there existsat least one row in result table returnedby sub-query

It is false if sub-query returns an emptyresult table

NOT EXISTS is the opposite of EXISTS

7/28/2019 12-SQL73

50/73

50

EXISTS cont.select o.city,

s.first_name || ' '||s.last_name

from sales_rep_tbl s,

office_tbl o

where s.office_num =o.office_num

and exists

(select 1

from manufacture_tbl m

where m.city = o.city);

7/28/2019 12-SQL73

51/73

51

Union, IntersectionUnion, Intersection

7/28/2019 12-SQL73

52/73

52

Union, Intersection and

DifferenceCan use normal set operations of union,intersection, and difference to combine

results of two or more queries into asingle result table.

Union of two tables, A and B, is table

containing all rows in either A or B orboth.

Intersection is table containing all rowscommon to both A and B.

7/28/2019 12-SQL73

53/73

53

Union cont.

Difference is table containing all rows inA but not in B.

Two tables must be union compatible.(SELECT area

FROM branch

WHERE area IS NOT NULL) UNION(SELECT area

FROM property_for_rent

WHERE area IS NOT NULL);

7/28/2019 12-SQL73

54/73

54

Comparison,Comparison,Character match,Character match,Numerical ExpressionNumerical Expression

7/28/2019 12-SQL73

55/73

55

Comparison Operators

= equal to

> greater than

< less than!= not equal

not equal

!> not greater than

!= greater than / equalto

7/28/2019 12-SQL73

56/73

56

Character Matches

Use to select rows containing field thatmatch specified portions of character

stringUse with character data only

Can use wildcards

% any string of zero or more characters

Enclose wildcards and character stringsin quotes

7/28/2019 12-SQL73

57/73

57

Numerical Expressions

Numerical expressions can be used inany numeric column or in any clause

that allows an expression+ addition

- subtration

* multiplication

/ division% modulos

7/28/2019 12-SQL73

58/73

58

AggregateAggregateFunctionsFunctions

7/28/2019 12-SQL73

59/73

59

Aggregate Functions

count(*) number of selected rows

count(col_name) number of nul-null values in column

max(col_name) highest value in column

min(col_name) lowest value in column

sum(col_name) total of values in column

avg(col_name) average of values in column

Aggregates ignore null values (exceptcount(*))

sum and avg only work with numericvalues

7/28/2019 12-SQL73

60/73

60

Aggregate Functions Cont.

If a group by clause is not used, onlyone row is returned

Aggregates may not be used in a whereclause

Aggregates can be applied to all rows in

a table or to a subset of a tableDistinct keyword eliminates duplicatevalues before applying the aggregate

function

7/28/2019 12-SQL73

61/73

61

NULL ValuesA null value is an unknown value

null is a special value which means noinformation available

is null should be used to match columnscontains null values

= null can also be used, but is notrecommended

One null value is never equal to anothernullvalue

null's sort and group together

Some columns are defined to permit null

values

7/28/2019 12-SQL73

62/73

62

Insert, Update,Insert, Update,DeleteDelete

7/28/2019 12-SQL73

63/73

63

INSERT

INSERT INTO table_name [ ( col_name1, col_name2, .... ) ]

VALUES ( expression1_1, expression1_2, .... ),

( expression2_1, expression2_2, .... ), ....

INSERT INTO table_name [ ( col_name1, col_name2, .... ) ]

SELECT ...

INSERT INTO table_name

SET col_name1 = expression1, col_name2 = expression2, ....

7/28/2019 12-SQL73

64/73

64

INSERT cont.

The column list is optional.

If omitted, SQL assumes a list of allcolumns in their original CREATETABLE order.

Any columns omitted must have been

declared as NULL when table wascreated, unless DEFAULT was specifiedwhen creating column.

7/28/2019 12-SQL73

65/73

65

INSERT cont.

Data value list must match column list asfollows:

Number of items in each list must be the same.

Must be direct correspondence in position of itemsin two lists.

Data type of each item in data_value_list must be

compatible with data type of corresponding column.

7/28/2019 12-SQL73

66/73

66

INSERT Examples

Not specifying column names:

INSERT INTO Dept VALUES (1, 'Engineering', ProjectSet());

Specifying column names:

INSERT INTO staff (sno, fname, lname, position, salary, bno)

VALUES ('SG44', 'Anne', 'Jones', 'Assistant', 8100, 'B3');

Insert using a select statement

INSERT INTO stores (store_name, total_sales)

SELECT ...

7/28/2019 12-SQL73

67/73

67

UPDATE

UPDATE table_name

SET col_name1 = expression1, col_name2 = expression2, ....

[ WHERE expression ]

[ LIMIT limit_amount ]

SET clause specifies names of one or more columns that are tobe updated.

Where clause is optional.

If omitted, named columns are updated for all rows intable.

7/28/2019 12-SQL73

68/73

68

Update Example

UPDATE emp

SET sal = sal * 1.1WHERE empno NOT IN (SELECT empno FROM bonus);

UPDATE staffSET salary = salary*1.05

WHERE position = 'Manager';

7/28/2019 12-SQL73

69/73

69

DELETE

DELETE FROM table_name

[ WHERE expression ]

[ LIMIT limit_amount ]

where is optional; if omitted, all rows are deleted from table. Thisdoes not delete table. If the where is specified, only those rowsthat satisfy condition are deleted.

7/28/2019 12-SQL73

70/73

70

DELETE examples

delete from store where store_num = 221;

Delete all information from the store table:

delete from store;

7/28/2019 12-SQL73

71/73

71

SQL Data TypesSQL Data Types

7/28/2019 12-SQL73

72/73

72

SQL Data Types

Data Type Declarationcharacter CHAR, VARCHAR

bit BIT, BIT VARYINGnumeric NUMERIC, DECIMAL, INTEGER

approx. numeric FLOAT, REAL, DOUBLE

datetime DATE, TIME, TIMESTAMP

7/28/2019 12-SQL73

73/73

Passion!Passion!

12-sql73

Documents