Object Oriented

1
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

2
Outline for today 11/24/03

• Objectives
• Introduction
• Need for OO/OR-DBMS
• OO Fundamentals
• How (OO)DBMS incorporates OO ideas
• Database Modeling/Querying
• Conclusions

OODBMS ORDBMS
Conceptual ODL/UML EER/PEER
3
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

4
Learning Objectives

• Basic concepts of OO and OR models
• Extensions at conceptual modeling
• Mapping Conceptual Model into Logical Model
• Exposure to additional features in SQL1999
  standard.
• Ability to model and implement a broader class of
  applications (spatial, multimedia, engineering,
  biological, scientific, )

5
Introduction

• Why OODBMS?
• Let us start with modeling this simple example in
  a RDBMS
• Assume that we are given the following diagram

6
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 inside the rectangle
  7.

7
Introduction Motivating Example

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

POINT
RECTANGLE
EDGE
p1 3 4
p2 3 10
p3 10 10
p4 10 4
r5 e1
r5 e2
r5 e3
r5 e4
e1 p1
e1 p2
e2 p2
e2 p3
e3 p3
e3 p4
e4 p4
e4 p1
8
Introduction Motivating Example

• Logical Flow

9
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 ltgt 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

10
Introduction Motivating Example

• Discussion Question
• Print the square names

11
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)

12
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
• What is lacking
• Support for complex data types
• Support for predicates (e.g., area, intersect,
  length)

13
OODBMS-Fundamentals

• Has its origin in OO programming languages
• Object
• State current value
• Behavior - what operations are permitted
• Difference between PL Object and DB Object
• 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

14
OODBMS-Fundamentals

• OID unique system generated object identifier
  for each object
• Compare this with RDBMS primary key
• Object structure
• Arbitrarily complex in order to contain all
  necessary information about the object
• Compare this with RDBMS
• (rectangle object) information about complex
  object is often scattered
• The state of complex object may be constructed
  from other objects using type constructors

15
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)

16
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, ltDNAMEi3, DCHAIRi1gt)

17
OODBMS-Fundamentals

• Type hierarchies and Inheritance
• OODB permits definition of new types based on
  other predefined types, leading to a type
  hierarchy
• Example
• 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

18
OODBMS-Fundamentals

• Exercise Consider the geometry objects defined
  in our first example and defined class
  hierarchies

19
OODBMS-Fundamentals

• example -
• GEOMETRY_OBJECT Shape, Area, ReferencePoint
• RECTANGLE subtype-of GEOMETRY_OBJECT
  (Shaperectangle) edge1, edge2, edge3, edge4

20
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

21
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

22
Database Modeling

• Example
• We use two examples (school-DB) and park-DB (from
  SDB-A Tour).
• Identify typical objects (and hierarchies) in
  school-DB
• Person, Student, Faculty, Department
• Identify relationships
• 11, 1M, MN, partial participation,
• 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

23
Database Modeling - EER
24
Database Modeling - EER

• Find a suitable entity and define a overlapping
  type of relationship

25
Database Modeling - UML

• Person
• SSN
• dob
• firstName
• lastName

26
Database Modeling - PEER
OGC-Geometry
27
Database Modeling - PEER

• State Park (ER Model)

28
Database Modeling - PEER

• Pictogram miniature version of geographic
  object inserted inside of a box
• Entity pictograms
• Relationship pictograms
• Advantages
• Ease of use
• Complete grammar
• Translation rules
• Pictogram inserted ER diagram to SQL3 level
  constructs

29
Database Modeling - PEER

• Pictograms
• ltPictogramgt ltShapegt
• // any possible shape
• ! // user-defined shapes
• ltShapegt ltBasic Shapegt
• ltMulti-Shapegt
• ltDerived Shapegt
• ltAlternate Shapegt

30
Database Modeling - PEER

• ltBasic Shapegt

ltCardinalitygt
0,1 1 1,n 0,n n
31
Database Modeling - PEER
ltBasic Shapegt

• ltDerived Shapegt

ltBasic Shapegt
ltBasic Shapegt
ltDerived Shapegt
ltBasic Shapegt
32
Database Modeling - PEER

• Any Possible Shape

User Defined Shape
Raster Shape
Raster
TIN
Thesian
33
Database Modeling - PEER

• State Park (PEER)

34
Summary and Conclusions

• Learning Objectives
• Understand OO Concepts
• Conceptual Modeling
• RDBMS limitations
• No support for complex data types and predicates
• OO
• Rich set of features encapsulation,
  inheritance, ..
• Helps manage complexity
• Conceptual Modeling Simple Extensions
• EER, UML, PEER

35
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

36
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)