Databases Object and Declaring Types
outline
History Concepts for Object Databases Object Database Standards, Languages, and Design Object-Relational and Extended-Relational Systems
History of DB Models
1950 File Systems, Punched Cards 1960 Hierarchical
(IMS IBM Mainframes)
1970 Network
(CODASYL, IDMS)
1980 Relational
(ORACLE, DB2, Sybase)
1990 Object-Oriented, Object-Relational
(O2, ORACLE10g)
Main Features
Relation is still the fundamental abstraction, with integration of OO features Structured Types
Non-atomic types
Methods
Special operations can be defined for a type
References
Pointers to tuples
Introduction
Shortcomings of traditional data models: engineering design, manufacturing, scientific experiments, telecommunications, geographic information systems and multimedia. New requirements: complex structure for objects, longer-duration transactions, new data-types for storing images or large textual items, and the need to define non standard application-specific operations.
Introduction
Key feature of OODB: designer can specify both the structure of complex objects and the operations to them. ODMG: a group of object-oriented DBMS venders and users. Two component of an object: state(value) and behavior(operations). Transient objects: exist only during program execution
Introduction
Persistent object: persist beyond program termination and can be retrieved later. Goal of OODB: maintain a direct correspondence between real-world and database objects so that objects do not lose their integrity and identity and can easily be identified and operated upon.
Object Identity &Structure
Object Identity (OID): unique, system-generated, for each independent object, immutable (its common to use long integers) A feature of OODB: objects may have an object structure of arbitrary complexity. The state of a complex object may be constructed from other objects. Object triple representation: (i, c, v) i is a unique OID, c is a type constructor, v is the object state.
Structured Types
Attributes of relation schemas can be
Atomic Relation schemas: Nested Relations
Type Constructors
Basic types: atom, tuple. example: o=(i1, atom, Leeds) Collection types: list, bag, set and array. example: o=(i2, set, {i1, i3, i4}) Represent object as a graph structure: apply the type constructors recursively. Identical states: Identical graph in every respect (OID at every level) Equal states(shallow equality):
Same graph structure Same corresponding atomic values Internal nodes may have different OIDs
Identical vs. Equal State objects
O1 = (i1, tuple, <a1:i4, a2:i6>) O2 = (i2, tuple, <a1:i5,, a2:i6>) O3 = (i3, tuple, <a1:i4, a2:i6>) O4 = (i4, atom, 10) O5 = (i5, atom, 10) O6 = (i6, atom, 20) Is the following objects are identical or equal: O1 and O2, O1 and O3, O4 and O5 ?
Encapsulation of Operation, Methods
Main idea: define the behavior of a type of object based on the operations that can be externally applied to objects of that type. Operation has an interface and implementation (body) Referential integrity can be defined explicitly using operations. Although, recently, the DB system automatically checks integrity using keys and inverse relationships (Oracle 10g). Complete encapsulation is inflexible. Visible attributes: may be directly accessed for reading by external operators, or by high level query language. hidden attributes: completely encapsulated, can be accessed only through predefined operators.
How to make Objects Persistence
Typical mechanism: naming and reachability Naming
give an object a unique persistent name (must be unique within a
particular database). This can be accomplished by defining an operation in the program
Reachability
works by making the object reachable from a persistent object. This can be accomplished by creating a collection type object (set, bag, list) and make it persistent. Then any object that is added to this collection becomes persistent It is possible to define persistent several collection for the same class definition
Class Hierarchies and Inheritance
Attributes and operations are together called functions (attributes resemble functions with zero arguments) Type: a type name and a listing of the names of its visible (public) functions:
PERSON: Name, Address, Birthdate, Age, SSN Student subtype-of Person with major, GPA
Defining a type:
defining all functions Implementing them as attributes or as methods
extents
Extents: collections of objects of the same type Its common in database application that each type or subtype will have an extent associated with it, which holds the collection of all persistent object of that type or subtype. Persistent collection, transient collection
Complex Objects
Two main types: structured and unstructured Structured:
made up of components Defined by applying available type constructors recursively
Unstructured: a data type requires a large amount of storage (image or large textual object) ODBMS can be viewed as having a extensible type system. DBMS internals provide the underlying storage and retrieval capabilities.
Other Concepts
Polymorphism: same name bound to different implementations. Example is the calculate_area () method for different geometry shape classes (circle, square, etc) Multiple inheritance : When a certain subtype T is a subtype of two or more types. Selective inheritance : When a subtype inherits only some of the functions of a supertype
Defining Structured Types SQL99
UDT User Defined Type
A UDT can be the type of a table A UDT can be the type of an attribute belonging to some table A UDT can inherit from another UDT
CREATE TYPE T AS <attribute and method declarations>;
Nested Relations SQL99 Example
CREATE TYPE AddressType AS ( street CHAR(50), city CHAR(20) ); / CREATE TYPE AddressTypeTable AS TABLE OF AddressType; / CREATE TYPE StarType AS ( name CHAR(30), address AddressTypeTable ); / CREATE TABLE MovieStar OF StarType;
Methods SQL99
Special operations defined for a type In SQL, implementation defined with Persistent Stored Modules (PSM) language METHOD m() RETURNS <TYPE>;
Methods SQL99 - Example
CREATE TYPE AddressType AS ( street CHAR(50), city CHAR(20) ) METHOD houseNumber() RETURNS CHAR(10); CREATE METHOD houseNumber() RETURNS CHAR(10) FOR AddressType BEGIN END;
References - I
Allow a tuple t refer to a tuple s rather than ...