Transcript PPT

C20.0046: Database
Management Systems
Lecture #13
Matthew P. Johnson
Stern School of Business, NYU
Spring, 2004
M.P. Johnson, DBMS, Stern/NYU, Sp2004
1
Agenda

Last time: G&A, modifications, defining
schemata, indices

This time:, Constraints, Triggers
Views
2. Constraints
3. Triggers
Homework 2 is due now
Future:
 Lots of programming for SQL
1.


M.P. Johnson, DBMS, Stern/NYU, Sp2004
2
Quick review

Examples from sqlzoo.net
M.P. Johnson, DBMS, Stern/NYU, Sp2004
3
New topic: Views (6.7)




Stored relations physically exist and persist
Views are relations that don’t
 in text, “table” = stored relation = “base table”
Basically names/references given to queries
 maybe a relevant subset of a table
Employee(ssn, name, department, project, salary)
CREATE VIEW Developers AS
SELECT name, project
FROM Employee
WHERE department = ‘Development’

Payroll has access to Employee, others only to Developers
M.P. Johnson, DBMS, Stern/NYU, Sp2004
4
A Different View



Person(name, city)
Purchase(buyer, seller, product, store)
Product(name, maker, category)
CREATE VIEW NYC-view AS
SELECT buyer, seller, product, store
FROM Person, Purchase
WHERE Person.city = ‘New York’ AND
Person.name = Purchase.buyer

We have a new virtual table:
NYC-view(buyer, seller, product, store)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
5
A Different View
CREATE VIEW NYC-view AS
SELECT buyer, seller, product, store
FROM Person, Purchase
WHERE Person.city = ‘New York’ AND
Person.name = Purchase.buyer

Now we can query the view:
SELECT name, store
FROM
NYC-view, Product
WHERE NYC-view.product = Product.name AND
Product.category = ‘shoes’
M.P. Johnson, DBMS, Stern/NYU, Sp2004
6
What happens when we query a view ?
SELECT name, NYC-view.store
FROM
NYC-view, Product
WHERE NYC-view.product = Product.name AND
Product.category = ‘shoes’
SELECT name, Purchase.store
FROM Person, Purchase, Product
WHERE Person.city = ‘NYC’ AND
Person.name = Purchase.buyer AND
Purchase.product = Product.name AND
Product.category = “shoes”
M.P. Johnson, DBMS, Stern/NYU, Sp2004
7
Can rename view fields
CREATE VIEW NYC-view(nycbuyer, nycseller,
prod, store) AS
SELECT buyer, seller, product, store
FROM Person, Purchase
WHERE Person.city = ‘NYC’ AND
Person.name = Purchase.buyer

Views info: USER_VIEWS
M.P. Johnson, DBMS, Stern/NYU, Sp2004
8
More complex views




NYC-View is just a subset of a table
Views can be arbitrarily complex
Imagine the database of student and course
records
Students get one view:


Professors get another view


For each semester, My-courses
For each course, My-students
Bursar gets another view, etc.
M.P. Johnson, DBMS, Stern/NYU, Sp2004
9
Types of Views

Views discussed here:




Used in databases
Computed only on-demand – slow at runtime
Always up to date
Sometimes talk about “materialized” views




Used in data warehouses
Pre-computed offline – fast at runtime
May have stale data
More later…
M.P. Johnson, DBMS, Stern/NYU, Sp2004
10
Updating Views


How can I insert a tuple into a table that doesn’t exist?
Employee(ssn, name, department, project, salary)
CREATE VIEW Developers AS
SELECT name, project
FROM Employee
WHERE department = ‘Development’
If we make the
following insertion:
INSERT INTO Developers
VALUES(‘Conrad’, ‘Optimizer’)
It INSERT INTO Employee(ssn, name, department, project, salary)
becomes: VALUES(NULL, ‘Conrad’, NULL, ‘Optimizer’, NULL)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
11
Non-Updatable Views


Person(name, city)
Purchase(buyer, seller, product, store)
CREATE VIEW City-Store AS
SELECT Person.city, Purchase.store
FROM Person, Purchase
WHERE Person.name = Purchase.buyer




How can we add the following tuple to the view?
(‘NYC’, ‘Nine West’)
We don’t know the name of the purchaser
cannot set to NULL (why?)
To delete: DROP VIEW City-Store
M.P. Johnson, DBMS, Stern/NYU, Sp2004
12
New topic: Constraints & Triggers (7)

Certain properties we’d like our database to hold

Enforce application semantics


prevent inconsistencies



e.g., sid is a key
e.g., sname has to be a string, age must be < 200
Modification of tables may break these properties
Constrains enforced by either


Preventing actions causing illegal states
Performing compensating actions

Triggers are one option
M.P. Johnson, DBMS, Stern/NYU, Sp2004
13
Constraints in SQL







Domain constraints
Keys
Foreign keys/referential integrity
Attribute-level constraints
Tuple-level constraints
Global constraints: assertions
simplest
Most
complex
The more complex, the harder/more expensive to
check and enforce
M.P. Johnson, DBMS, Stern/NYU, Sp2004
14
Denoting primary keys

In the CREATE TABLE statement, use:

PRIMARY KEY or UNIQUE
CREATE TABLE MovieStar (
Example:

name VARCHAR(30) PRIMARY KEY,
address VARCHAR(255),
gender CHAR(1)
)

If NAME and ADDRESS are together the primary key, it must
be written as:
CREATE TABLE MovieStar (
name VARCHAR(30),
address VARCHAR(255),
gender CHAR(1),
PRIMARY KEY (name, address) )
M.P. Johnson, DBMS, Stern/NYU, Sp2004
15
Effects of primary keys

No two tuples may agree on all key atts




No nulls allowed in any of the key atts
Only one primary key is allowed per table
NB: SQL key is slightly different from Relational
Model key



Have key  no duplicates at all
In R.M., key fields values  all other field values
In SQL, key field values really are unique
NB: no concept of superkey in SQL

Of course, chosen keys should be minimal (why?)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
16
Multiple keys
CREATE TABLE Product (
productID CHAR(10),
name CHAR(30),
category VARCHAR(20),
price INT,
UPC CHAR(20) UNIQUE,
PRIMARY KEY (productID),
UNIQUE (name, category))
There is at most one PRIMARY KEY;
there can be many UNIQUE
M.P. Johnson, DBMS, Stern/NYU, Sp2004
17
PRIMARY KEY v. UNIQUE

Can use the UNIQUE keyword (instead of PRIMARY KEY) in
(almost) same way


Unique keys are…not unique
NULLs allowed, and multiple NULLs allowed (why?)

Example:

If NAME and ADDRESS are together unique, we can write:
CREATE TABLE MovieStar (
name VARCHAR(30) UNIQUE,
address VARCHAR(255),
gender CHAR(1) )
CREATE TABLE MovieStar (
name VARCHAR(30),
address VARCHAR(255),
gender CHAR(1),
UNIQUE (name, address) )
M.P. Johnson, DBMS, Stern/NYU, Sp2004
18
Indices & keys




Close affinity between indices and keys
Want to keep values unique (key)  need
easy way to check for existing values (index)
In many implementations (like Oracle):
creating keys (primary or unique)
automatically creates indices
Explicit unique index creation:
CREATE UNIQUE INDEX MyIndex on MyTable(a,b,c)


In effect, automatically creates key
Index info: USER_INDEXES
M.P. Johnson, DBMS, Stern/NYU, Sp2004
19
Foreign Key Constraint
ActedIn(ActorName, MovieName)
Movies(MovieName, year)
If MovieName in ActedIn is a foreign key for
Movies, then:



1.
2.
The referenced atts must be a key in the ref-ed table
All non-null foreign key values must appear in the
referenced table
ActorName
MovieName MovieYear
(1)
Movies
ActedIn
Title
Year
(2)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
20
Foreign key example

Recall, ActedIn has FK MovieName...




Movies(MovieName, year)
(Fatal Attraction, 1987)
ActedIn(ActorName, MovieName)
(Michael Douglas, Fatal Attraction)
ActorName
MovieName MovieYear
M.Douglas
Fatal Attract 1987
ActedIn
(2)
(1)
Movies
Title
Year
Fatal Attraction
1987
M.P. Johnson, DBMS, Stern/NYU, Sp2004
21
Declaring Foreign Key Constraints

REFERENCES keyword:
CREATE TABLE ActedIn (
ActorName VARCHAR(30),
MovieName VARCHAR(30)
REFERENCES Movies(MovieName) )

Or summarize at end:
CREATE TABLE ActedIn (
ActorName VARCHAR(30),
MovieName VARCHAR(30),
CONSTRAINT const_name
FOREIGN KEY (MovieName)
REFERENCES Movies(MovieName) )
M.P. Johnson, DBMS, Stern/NYU, Sp2004
22
Constraint example

Q: Is this the same as REFERENCES?
CREATE TABLE StarrtedIn (
ActorName VARCHAR(30),
MovieName VARCHAR(30)
CHECK (MovieName IN
(SELECT MovieName FROM Movies) ) )
ActorName
MovieName MovieYear
M.Douglas
Fatal Attract 1987
ActedIn
(2)
(1)
Movies
Title
Year
Fatal Attraction
1987
M.P. Johnson, DBMS, Stern/NYU, Sp2004
23
Dangling tuples result from
Can result from:



In StarredIn: insert/update
In Movies: delete/update
Responses:

1.
2.
3.
Reject
Cascade
Set-null
M.P. Johnson, DBMS, Stern/NYU, Sp2004
24
What happens during updates?




SQL has three policies for maintaining
referential integrity:
Reject violating modifications (default)
Cascade: after a delete/update do a
delete/update
Set-null set foreign-key field to NULL
M.P. Johnson, DBMS, Stern/NYU, Sp2004
25
Cascading and set-null


Cascading:
Applies to last two causes:





Deleting exec
Bad exec update
I.e., loss/change of exec referred to
Strategy: del/change exec referred to 
del/change the reference to exec
Loss/change of exec referred to  set
reference to null
M.P. Johnson, DBMS, Stern/NYU, Sp2004
26
Cascading and set-null

Cascade/set-null choices independent by att:
CREATE TABLE Studio (
name CHAR(30) PRIMARY KEY,
address VARCHAR(255),
presC# int REFERENCES MovieExec(cert#)
ON DELETE SET NULL
ON UPDATE CASCADE
)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
27
Circular dependencies

Sometimes one table refers to another and
the other refers to the first


Soln involves



Neither can be inserted without the other
Transactions (xacts)
Deferring constraint check until end of xact
Constraints can be NOT DEFERABLE
(default) or DEFERABLE


DEFERABLE INITIALLY DEFERRED
DEFERABLE INITIALLY IMMEDIATE
M.P. Johnson, DBMS, Stern/NYU, Sp2004
28
Deferring constraints
CREATE TABLE Studio (
name CHAR(30) PRIMARY KEY,
address VARCHAR(255),
presC# INT UNIQUE
REFERENCES MovieExec(cert#)
DEFERRABLE INITIALLY DEFERRED
)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
29
Constraints
Constraints can be applied to

1.
2.

On attributes
On tuples
Set update and delete actions inde’ly:
CREATE TABLE StarredIn (
ActorName VARCHAR(30),
MovieName VARCHAR(30)
REFERENCES Movies(MovieName)
ON DELETE SET CASCADE )

ON UPDATE not supported by Oracle
M.P. Johnson, DBMS, Stern/NYU, Sp2004
30
Constraining Attribute Values

Constrain invalid values
name CHAR(30) NOT NULL



gender CHAR(1) CHECK (gender IN (‘F’, ‘M’))
Any test legal in a WHERE clause


Subqueries (but not in Oracle!), etc.
The constraint is checked whenever the local
attributes it references are changed or added
M.P. Johnson, DBMS, Stern/NYU, Sp2004
31
Is this the same as a
foreign key?
CREATE TABLE Purchase (
prodName CHAR(30)
CHECK (prodName IN
SELECT Product.name
FROM Product),
date DATETIME NOT NULL)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
32
Naming domains to constrain values

Can define new domains to use as the
attribute type:
CREATE DOMAIN GenderDomain
CHAR(1)
CHECK (VALUE IN (‘F’, ‘M’))

Then update our attribute definition...



gender GenderDomain
Note use of VALUE to refer to the att value
Named domains are not supported by Oracle
M.P. Johnson, DBMS, Stern/NYU, Sp2004
33
More complex constraints

Among several attributes
in one table
Specify at the end of
CREATE TABLE
CHECK (gender = ‘F’ OR name
NOT LIKE ‘Ms.%’)


Checked whenever a tuple of
the relation is added or
updated.
CREATE TABLE Sailors
( sid INTEGER,
sname CHAR(10),
rating INTEGER,
age REAL,
PRIMARY KEY (sid),
CHECK ( rating >= 1
AND rating <= 10)
)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
34
Adding/deleting named constraints

Add:
ALTER TABLE Movie
ADD CONSTRAINT Const1 PRIMARY KEY(name)

Delete:
ALTER TABLE Movie
DROP CONSTRAINT Const1

Constraints info: USER_CONSTRAINTS
M.P. Johnson, DBMS, Stern/NYU, Sp2004
35
Assertions (7.4)

Constraint on entire database (not table!)






Checked upon any change to DB
Not supported by Oracle
CREATE ASSERTION <name> CHECK (<condition>)
Analogous to our assertions in R.A.:
Studio(Name, Address, PresC#);
MovieExec(Name, Address, Cert#, NetWorth);
CREATE ASSERTION RichPres CHECK
(NOT EXISTS
(SELECT *
FROM Studio, MovieExec
WHERE presC# = cert#
AND netWorth < 10,000,000))
M.P. Johnson, DBMS, Stern/NYU, Sp2004
36
Another assertion example

Movie(Title, Year, Length, inColor, StudioName,
ProducerC#)
CREATE ASSERTION SumLength
CHECK
(10000 >= ALL
(SELECT SUM(length)
FROM Movie
GROUP BY studioName))


Q: Can we write this as a tuple constraint?
Q: Could we if >=  < ?
M.P. Johnson, DBMS, Stern/NYU, Sp2004
37
Another example

Q: What does this check?
CREATE ASSERTION myAssert CHECK
NOT EXISTS(
SELECT Product.name
FROM Product, Purchase
WHERE Product.name = Purchase.prodName
GROUP BY Product.name
HAVING count(*) > 200)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
38
Different Constraint Types
Type
Where
Declared
When
activated
Attribute
with attribute
on insertion or not if subquery
update
Tuple
relation
schema
insertion or
update to
relation
not if subquery
Assertion
database
schema
on change to
any relation
mentioned
Yes

Guaranteed to
hold?
Q: Why not always use assertions?
M.P. Johnson, DBMS, Stern/NYU, Sp2004
39
Triggers

Constraints state what must remain true



Triggers are instructions to perform at explicitly
specified times
Three aspects:




DBMS decides when to check
An event (e.g., update to an attribute)
A condition (e.g., a query to check)
An action (the trigger’s effect) (deletion, update, insertion)
When the event occurs, DBMS checks the
constraint, and if it is satisfied, performs the action
M.P. Johnson, DBMS, Stern/NYU, Sp2004
40
Triggers – important points





Can replace old row (result of event) with
new row
Action may be performed before or after
event
Can refer to old row and new row
WHEN clauses tests whether to continue
Action may be performed either



For each row involved in event
Once per event
Oracle does triggers as PL/SQL programs
M.P. Johnson, DBMS, Stern/NYU, Sp2004
41
Elements of Triggers





Timing of action execution: before, after or instead
of triggering event
The action can refer to both the old and new state of
the database
Update events may specify a particular column or
set of columns
A condition is specified with an optional WHEN
clause
The action can be performed either for


once for every tuple or
once for all the tuples that are changed by the database
operation
M.P. Johnson, DBMS, Stern/NYU, Sp2004
42
Simple trigger example

R(id, data, last-modified)




Goal: whenever data is modified, update lastmodified date
Could modify all scripts/programs that touch this
table
CREATE TRIGGER UpdateDateTrigger


data is a large string
Last-modified is a newly added date field
Bad idea
Better: user a trigger
BEFORE UPDATE OF data ON R
REFERENCING
NEW ROW AS NewTuple
FOR EACH STATEMENT
BEGIN
NewTuple.last-modified = sysdate;
END;
M.P. Johnson, DBMS, Stern/NYU, Sp2004
43
Triggers: Row-level example


MovieExec(name, address, cert#, netWorth)
“If someone decreases a movie executive’s net worth, I
want the database to reset itself to the previous net
worth.”
CREATE TRIGGER NetWorthTrigger
AFTER UPDATE OF netWorth ON MovieExec
REFERENCING
NEW ROW AS NewTuple
OLD ROW AS OldTuple
FOR EACH ROW
WHEN (OldTuple.netWorth>NewTuple.netWorth)
UPDATE MovieExec
SET netWorth = oldTuple.netWorth
WHERE cert# = newTuple.cert#)
M.P. Johnson, DBMS, Stern/NYU, Sp2004
44
Triggers: Table-level example


MovieExec(name, address, cert#, netWorth)
“If someone updates the net worth of one movie exec so that the
average net worth of all movie execs becomes less than $50,000, I
want the database to reset itself.”
CREATE TRIGGER AvgNetWorthTrigger
AFTER UPDATE OF netWorth ON MovieExec
REFERENCING
OLD TABLE AS OldStuff,
NEW TABLE AS NewStuff
FOR EACH STATEMENT
WHEN (50000 > (SELECT AVG(netWorth) FROM
MovieExec))
BEGIN
DELETE FROM MovieExec
WHERE (Name, address, cert#, netWorth) IN
NewStuff;
INSERT INTO MovieExec
(SELECT * FROM OldStuff);
END;
M.P. Johnson, DBMS, Stern/NYU, Sp2004
45
End

Examples from sqlzoo.net

For next time: read 8.1-2
M.P. Johnson, DBMS, Stern/NYU, Sp2004
46