Object Oriented & Object Relational Databases

Ranga Raju Vatsavai

Teaching Mentor (Prof. Shekhar) CSci 5708 : Architecture and Implementation of Database Management Systems, Fall 2003 Week 15 (11/24, 11/26/03) Lecture Notes

Outline for today 11/24/03

Objectives Introduction

Need for OO/OR-DBMS How (OO)DBMS incorporates OO ideas OODBMS ORDBMS EER/PEER

OO Fundamentals

Database Modeling/Querying
Conceptual ODL/UML

Conclusions

Outline for next class 11/26/03

Objectives ORDBMS Fundamentals

How ORDBMS incorporates OO ideas

Mapping Conceptual Model into Logical Model OQL vs. SQL-99 Comparison of OODBMS and ORDBMS Conclusions

Learning Objectives

Basic concepts of OO and OR models Extensions at conceptual modeling Mapping Conceptual Model into Logical Model Exposure to additional features in SQL:1999 standard. Ability to model and implement a broader class of applications (spatial, multimedia, engineering, biological, scientific, )

Introduction

Why OODBMS? Let us start with modeling this simple example in a RDBMS

Assume that we are given the following diagram 1 6 5 4 0, 0 8 2 7 3

Introduction Motivating Example

Our objective is to store this graph in a RDBMS and support queries on these simple geometries

What kind of relations we need? Print names of rectangles which are squares Find all objects which are inside1the rectangle 7. 6 5 4 0, 0 8 2 7 3

Introduction Motivating Example

Relations

POINT(pname, x, y), EDGE(ename, aPoint), RECTANGLE(rname, anEdge).

POINT EDGE

RECTANGL E r5 e1

e1 e1 e2 e2 e3 e3 e4 e4

p1 p2 p2 p3 p3 p4 p4 p1

p1 3 p2 3

4 10

r5 r5 r5

e2 e3 e4

p3 10 10 p4 10 4

1 6 5 4 0, 0 8 2 7 3

Introduction Motivating Example

Logical Flow
rname Edge_length (eName, eLength)

Rectangle r Edge_Pnt_List (eName, stPnt, endPnt) Point p1 Point p2 Edge e1 E1.ename = e2.ename E1.aPoint <> e2.aPoint Edge e2

Introduction Motivating Example

CREATE VIEW pnt_list (ename, stPnt, endPnt) AS SELECT e.ename, e1.aPoint, e2.aPoint FROM edge e1, edge e2 WHERE e1.ename = e2.ename AND e1.aPoint <> e2.aPoint; CREATE VIEW edge_length (ename, elength) AS SELECT e.ename, sqrt(sq(p1.x p2.x) + sq(p1.y p2.y) FROM pnt_list e, point p1, point p2 WHERE e.stPnt = p1.pname AND e.endPnt = p2.pname;

Introduction Motivating Example

Discussion Question Print the square names

Introduction Motivating Example

Solution SELECT r.rname FROM rectangle r, edge_length e WHERE r.anEdge = e.ename GROUP BY r.rname HAVING max(e.elength) ~ MIN (e.elength);

Introduction

Though we can model and query these simple geometric objects, its

Not clean Complexity increases with complex objects (like irregular polygons) Even with these simple objects, how do you answer queries like find nearest objects to point 2 or find all intersecting objects Support for complex data types Support for predicates (e.g., area, intersect, length)

What is lacking

OODBMS-Fundamentals

Has its origin in OO programming languages Object

State current value Behavior - what operations are permitted PL Objects exist only during program execution (transient objects). DB Objects persist beyond program termination; stored permanently on a secondary storage DB allows sharing of objects among multiple programs and applications

Difference between PL Object and DB Object

OODBMS-Fundamentals

OID unique system generated object identifier for each object

Compare this with RDBMS primary key Arbitrarily complex in order to contain all necessary information about the object Compare this with RDBMS

Object structure

(rectangle object) information about complex object is often scattered

The state of complex object may be constructed from other objects using type constructors

OODBMS-Fundamentals

Type constructors

Basic atom, tuple, and set Others list, bag, and array (collection/bulk types) tuple also called as a structured type

Formally an object can be thought of as a triple (I,C,V)

I OID C type constructor V object state (current value)

OODBMS-Fundamentals

Examples

o1 = (i1, atom, Pen-Chung Yew) o2 = (i2, atom, Minneapolis) o3 = (i3, atom, Computer Science) o4 = (i4, set, {i1,i2,i3}) o5 = (i4, tuple, <DNAME:i3, DCHAIR:i1>)

OODBMS-Fundamentals

Type hierarchies and Inheritance

OODB permits definition of new types based on other predefined types, leading to a type hierarchy TYPE_NAME: function, function, , function PERSON: firstName, lastName, dob, SSN STUDENT: firstName, lastName, dob, SSN, status, GPA FACULTY: firstName, lastName, dob, SSN, rank, salary STUDENT subtype_of PERSON: major, GPA FACULTY subtype_of PERSON: rank, salary

Example

OODBMS-Fundamentals

Exercise: Consider the geometry objects defined in our first example and defined class hierarchies
1 6 5 4 0, 0 8 2 7 3

OODBMS-Fundamentals

example

GEOMETRY_OBJECT: Shape, Area, ReferencePoint RECTANGLE subtype-of GEOMETRY_OBJECT (Shape=rectangle): edge1, edge2, edge3, edge4

OODBMS-Fundamentals

Standards:

ODMG Made up of several parts

Object Model Object Definition Language (ODL) Object Query Language (OQL) Binding to OO programming languages (C++, Smalltalk, Java)

OOBMS

O2 ObjectStore Jasmine

Database Modeling

Enhance Entity Relationship (EER Chap 4, Elmasri/Navathe) Pictogram Enhanced Entity Relationship (PEER Chapter 2, SD A Tour). Unified Modeling Language (UML) Object Definition Language

Database Modeling

Example

We use two examples (school-DB) and park-DB (from SDB-A Tour). Identify typical objects (and hierarchies) in schoolDB

Person, Student, Faculty, Department 1:1, 1:M, M:N, partial participation,

Identify relationships

Let us start with EER

Includes all the modeling concepts of ER Additionally includes oo concepts of subclass, superclass, specialization and generalization, attribute and relationship inheritance

Database Modeling - EER

SSN do b status Student
N

Person d

firstNam e lastNam e Faculty

N

rank

major
1

worksin
1

code

Department

name

Database Modeling - EER

Find a suitable entity and define a overlapping type of relationship

Database Modeling - UML

Person SSN dob firstName lastName

{disjoint, mandatory}

Student status

Faculty 0..1 Departmen chai rank t r chairOf majorsIn 1 Code worksIn * dept major Name 1

Database Modeling - PEER

1+

Exactly one Many (0 or More) Optional (0 or One) One or More Numerically Specified Aggregation Inheritance Derived Class

OGC-Geometry

Database Modeling - PEER

State Park (ER Model)

Volume supplies N Name M Length River Name M 1 within Belongs_to 1 1 1 Geometry Crosses Geometry Elevation Forest 1 N captures M Fire_Image Image-id Name NoOfLanes Geometry Road M access N Part_of M

Geometry Facility Name

Geometry Forest_Stand Stand-id Specie

M Fire_Station Name

manages

Geometry

Image

Database Modeling - PEER

Pictogram : miniature version of geographic object inserted inside of a box

Entity pictograms Relationship pictograms Ease of use Complete grammar Translation rules Pictogram inserted ER diagram to SQL3 level constructs

Advantages

Database Modeling - PEER

Pictograms
<Pictogram> <Shape> * // any possible shape ! // user-defined shapes <Shape> <Basic Shape> <Multi-Shape> <Derived Shape> <Alternate Shape>

Database Modeling - PEER

<Basic Shape> <Cardinality >

0,1 1 1,n 0,n n

Point

Line

Polygon

Pictograms for Basic Shapes

n

0, n

Pictograms for Multi-Shapes

Database Modeling - PEER

<Derived Shape>

<Basic Shape>

Pictograms for Derived <Basic Shape> <Basic Shape> Shapes

<Derived Shape> <Basic Shape>

Pictograms for Alternate Shapes

Database Modeling - PEER

Any Possible Shape User Defined Shape Raster Shape

*
Raster TIN Thesian

! Raste TIN r Thesia n

Pictograms for Raster Shapes

Part_of (Network)

Part_of (Partition)

Pictograms for Relationships

Database Modeling - PEER

State Park (PEER)

Volume supplies N Name M Length River Name Elevation Facility Belongs_to Forest 1 1 N captures M Fire_Image Image-id Crosses Name NoOfLanes Road M access N Part_of M Forest_Stand Stand-id Specie

Name

manages

Fire_Station Name

Summary and Conclusions

Learning Objectives

Understand OO Concepts Conceptual Modeling No support for complex data types and predicates Rich set of features encapsulation, inheritance, .. Helps manage complexity

RDBMS limitations

OO

Conceptual Modeling Simple Extensions EER, UML, PEER

Next Week

ORDMBS (SQL-3) in depth Mapping of Conceptual Model onto ORDBMS (SQL-3 DDL constructs) Examples using Postgresql/PostGIS Comparison of OODBMS and ORDBMS Conclusions

Additional Readings

http://www.cs.umn.edu/~vatsavai/oo-ordbms.ppt Main Database Management Systems by Raghu Ramakrishnan (Chapter 24: Object Database Systems). Additional References (relevant chapters):

Fundamentals of Database Systems Elmasri & Navathe A First Course In Database Systems Ullman & Widom.

http://www-users.cs.umn.edu/~shekhar/5705/

Spatial Database A Tour (Chapter 2 and 3)