Tuesday, December 4, 2012

Database Management System Chapter-9

 

 

Chapter 9:  Object-Relational Databases

 

n Nested Relations

n Complex Types and Object Orientation

n Querying with Complex Types

n Creation of Complex Values and Objects

n Comparison of Object-Oriented and Object-Relational Databases

 

Object-Relational Data Models

 

n Extend the relational data model by including object orientation and constructs to deal with added data types.

n Allow attributes of tuples to have complex types, including non-atomic values such as nested relations.

n Preserve relational foundations, in particular the declarative access to data, while extending modeling power.

n Upward compatibility with existing relational languages.

 

Nested Relations

 

n Motivation:

Ø Permit non-atomic domains (atomic º indivisible)

Ø Example of non-atomic domain:  set of integers,or set of tuples

Ø Allows more intuitive modeling for applications with complex data

n Intuitive definition:

Ø allow relations whenever we allow atomic (scalar) values — relations within relations

Ø Retains mathematical foundation of relational model

Ø Violates first normal form.

 

Example of a Nested Relation

 

n Example:  library information system

 

n Each book has

Ø title,

Ø a set of authors,

Ø Publisher, and

Ø a set of keywords

n Non-1NF relation books

 

 


1NF Version of Nested Relation

n 1NF version of books

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


flat-books

 

 

 

 

4NF Decomposition of Nested Relation

 

n Remove awkwardness of flat-books by assuming that the following multivalued dependencies hold:

Ø title       author

Ø title       keyword

Ø title       pub-name, pub-branch

n Decompose flat-doc into 4NF using the schemas:

Ø (title, author)

Ø (title, keyword)

Ø (title, pub-name, pub-branch)

 

 

 

 

 

 

 

 

 

4NF Decomposition of flatbooks

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Problems with 4NF Schema

 

n 4NF design requires users to include joins in their queries.

n 1NF relational view flat-books defined by join of 4NF relations:

Ø eliminates the need for users to perform joins,

Ø but loses the one-to-one correspondence between tuples and documents.

Ø And has a large amount of redundancy

n Nested relations representation is much more natural here.

Complex Types and SQL:1999

 

n Extensions to SQL to support complex types include:

Ø Collection and large object types

ê Nested relations are an example of collection types

Ø Structured types

ê Nested record structures like composite attributes

Ø Inheritance

Ø Object orientation

ê Including object identifiers and references

n Our description is mainly based on the SQL:1999 standard

Ø Not fully implemented in any database system currently

Ø But some features are present in each of the major commercial database systems

ê Read the manual of your database system to see what it supports

Ø We present some features that are not in SQL:1999

ê These are noted explicitly

 

Collection Types

 

n Set type (not in SQL:1999)

    create table books (
      …..
      keyword-set  setof(varchar(20))
      ……
)

n Sets are an instance of collection types. Other instances include

Ø Arrays (are supported in SQL:1999)

ê E.g.  author-array varchar(20) array[10]

ê Can access elements of array in usual fashion:

  E.g.  author-array[1]

Ø Multisets  (not supported in SQL:1999)

ê I.e., unordered collections, where an element may occur multiple times

Ø Nested relations are sets of tuples

ê SQL:1999 supports arrays of tuples

 

 

 

 

 

Large Object Types

 

n Large object types

Ø clob: Character large objects 

              book-review clob(10KB)

Ø blob:  binary large objects

                  image          blob(10MB)

                  movie          blob (2GB)

n JDBC/ODBC provide special methods to access large objects in small pieces

Ø Similar to accessing operating system files

Ø Application retrieves a locator for the large object and then manipulates the large object from the host language

 

 

Structured and Collection Types

 

n  Structured types can be declared and used in SQL

       create type Publisher as
         
(name            varchar(20),
           branch           varchar(20))
    create type Book as
          (title                varchar(20),
           author-array   varchar(20) array [10],
           pub-date         date,
           publisher        Publisher,
           keyword-set   setof(varchar(20)))

Ø Note: setof declaration of keyword-set is not supported by SQL:1999

Ø Using an array to store authors lets us record the order of the authors

n  Structured types can be used to create tables

            create table books of Book

Ø Similar to the nested relation books, but with array of authors
instead of set

 

n Structured types allow composite attributes of E-R diagrams to be represented directly.

n Unnamed row types can also be used in SQL:1999 to define composite attributes

Ø E.g. we can omit the declaration of type Publisher and instead use the following in declaring the type Book

              publisher   row (name varchar(20),
                                 branch varchar(20))

n Similarly, collection types allow multivalued attributes of E-R diagrams to be represented directly.

 

n  We can create tables without creating an intermediate type

Ø For example, the table books could also be defined as follows:

          create table books

             (title varchar(20),

              author-array varchar(20) array[10],

              pub-date date,

               publisher Publisher

             keyword-list setof(varchar(20)))

n  Methods can be part of the type definition of a structured type:

         create type Employee as (
         name varchar(20),
         salary integer)

    
method giveraise (percent integer)

n  We create the method body separately

       create method giveraise (percent integer) for Employee
  
begin
        set self
.salary = self.salary + (self.salary * percent) / 100;
   end

 

Creation of Values of Complex Types

 

n Values of structured types are created using constructor functions

Ø E.g.   Publisher(‘McGraw-Hill’, ‘New York’)

Ø Note: a value is not an object

n SQL:1999 constructor functions  

Ø E.g.
create function Publisher (n varchar(20), b varchar(20))
returns Publisher
begin
  set
name=n;
  set branch=b;
end

Ø Every structured type has a default constructor with no arguments, others can be defined as required

n Values of row type can be constructed by listing values in parantheses

Ø E.g. given row type   row (name varchar(20),
                                       branch varchar(20))

Ø We can assign (`McGraw-Hill’,`New York’) as a value of above type

 

 

Creation of Values of Complex Types

 

n Array construction

           array [‘Silberschatz’,`Korth’,`Sudarshan’]

n Set value attributes (not supported in SQL:1999)

Ø set( v1, v2, …, vn)

n To create a tuple of the books relation
               (‘Compilers’, array[`Smith’,`Jones’],
                 Publisher(`McGraw-Hill’,`New York’),                                          set(`parsing’,`analysis’))

n To insert the preceding tuple into the relation books

     insert into books
values
   (`Compilers’, array[`Smith’,`Jones’],
      Publisher(‘McGraw Hill’,`New York’ ),                           
      set(`parsing’,`analysis’))

 

 

 

Inheritance

 

n Suppose that we have the following type definition for people:

          create type Person
          
(name varchar(20),
             address varchar(20))

n Using inheritance to define the student and teacher types
      create type Student
       
under Person
       
(degree        varchar(20),
         department  varchar(20))
      create type Teacher
       
under Person
       
(salary          integer,
         department  varchar(20))

n Subtypes can redefine methods by using overriding method in place of method in the method declaration

 

Multiple Inheritance

 

n SQL:1999 does not support multiple inheritance

n If our type system supports multiple inheritance, we can define a type for teaching assistant as follows:
      create type Teaching Assistant
           
under Student, Teacher

n To avoid a conflict between the two occurrences of department we can rename them

                create type Teaching Assistant
          
under
             
Student  with (department as student-dept),
            Teacher  with (department as teacher-dept)

 

Table Inheritance

 

n Table inheritance allows an object to have multiple types by allowing an entity to exist in more than one table at once.

n E.g. people table:    create table people of Person

n We can then define the students and teachers tables as subtables of people

          create table students of Student
          
under people
     
create table teachers of Teacher
                         
under people

n Each tuple in a subtable (e.g. students and teachers) is implicitly present in its supertables (e.g. people)

n  Multiple inheritance is possible with tables, just as it is possible with types.
           create table
teaching-assistants of Teaching Assistant
         
under students, teachers

Ø Multiple inheritance not supported in SQL:1999

 

 

 

Table Inheritance:  Roles

 

n Table inheritance is useful for modeling roles

n permits a value to have multiple types, without having a
most-specific type (unlike type inheritance).

Ø e.g., an object can be in the students and teachers subtables simultaneously, without having to be in a subtable student-teachers that is under both students and teachers

Ø object can gain/lose roles: corresponds to inserting/deleting object from a subtable

n NOTE:  SQL:1999 requires values to have a most specific type

Ø so above discussion is not applicable to SQL:1999

 

Table Inheritance:  Consistency Requirements

 

n Consistency requirements on subtables and supertables.

Ø Each tuple of the supertable (e.g. people) can correspond to at most one tuple in each of the subtables (e.g. students and teachers)

Ø Additional constraint in SQL:1999:

   All tuples corresponding to each other (that is, with the same values for inherited attributes) must be derived from one tuple (inserted into one table).  

ê That is, each entity must have a most specific type

ê We cannot have a tuple in people corresponding to a tuple each in students and teachers

 

Table Inheritance: Storage Alternatives

 

n Storage alternatives

H Store only local attributes and the primary key of the supertable in subtable

ê Inherited attributes derived by means of a join with the supertable

H Each table stores all inherited and locally defined attributes

ê Supertables implicitly contain (inherited attributes of) all tuples in their subtables

ê Access to all attributes of a tuple is faster: no join required

ê If entities must have most specific type, tuple is stored only in one table, where it was created

H Otherwise, there could be redundancy

 

 

Reference Types

 

n Object-oriented languages provide the ability to create and refer to objects.

n In SQL:1999

Ø References are to tuples, and

Ø References must be scoped,

ê I.e., can only point to tuples in one specified table

n We will study how to define references first, and later see how to use references

Reference Declaration in SQL:1999

 

n E.g. define a type Department with a field name and a field head which is a reference to the type Person, with table people as scope

          create type Department(
             name varchar(20),
             head ref(Person) scope people)

n We can then create a table departments as follows

             create table departments of Department

n We can omit the declaration scope people from the type declaration and instead make an addition to the create table statement:
      create table departments of Department
             (head with options scope people)

 

Initializing Reference Typed Values

 

n In Oracle, to create a tuple with a reference value, we can first create the tuple with a null reference and then set the reference separately by using the function ref(p) applied to a tuple variable

n E.g. to create a department with name CS and head being the person named John, we use

    insert into departments

           values (`CS’, null)

    update departments

         set head = (select ref(p)

                           from people as p

                        where name=`John’)

         where name = `CS’

 

 

n SQL:1999 does not support the ref() function, and instead requires a special attribute to be declared to store the object identifier

n The self-referential attribute is declared by adding a ref is clause to the create table statement:        

        create table people of Person
   
ref is oid system generated

Ø Here, oid is an attribute name, not a keyword.

n To get the reference to a tuple, the subquery shown earlier would use

                        select p.oid

     instead of    select ref(p)

 

User Generated Identifiers

 

n SQL:1999 allows object identifiers to be user-generated

Ø The type of the object-identifier must be specified as part of the type definition of the referenced table, and

Ø The table definition must specify that the reference is user generated

Ø E.g.

            create type Person
            (name
varchar(20)
             address varchar(20))
           ref using varchar(20)
        create table people of Person
         
ref is oid user generated

n When creating a tuple, we must provide a unique value for the identifier (assumed to be the first attribute):

       insert into people values
     
(‘01284567’, ‘John’, `23 Coyote Run’)

   

 

n We can then use the identifier value when inserting a tuple into departments

Ø Avoids need for a separate query to retrieve the identifier:

      E.g.  insert into departments
          
values(`CS’, `02184567’)

n It is even possible to use an existing primary key value as the identifier, by including the ref from clause, and declaring the reference to be derived

    create type Person
     
(name varchar(20) primary key,
       address varchar(20))
    ref from(name)
create table people of Person
   
ref is oid derived

n When inserting a tuple for departments, we can then use

    insert into departments
    values
(`CS’,`John’)

 

Path Expressions

 

n Find the names and addresses of the heads of all departments:

          select head –>name, head –>address
      from departments

n An expression such as “head–>name” is called a path expression

n Path expressions help avoid explicit joins

Ø If department head were not a reference, a join of departments with people would be required to get at the address

Ø Makes expressing the query much easier for the user

 

Querying with Structured Types

 

n Find the title and the name of the publisher of each book.

          select title, publisher.name
     
from books

    Note the use of the dot notation to access fields of the composite attribute (structured type) publisher

 

Collection-Value Attributes

 

n Collection-valued attributes can be treated much like relations, using the keyword unnest

Ø The  books relation has array-valued attribute author-array  and set-valued attribute keyword-set

n To find all books that have the word “database” as one of their keywords,             
        select
title
     
from books
     
where ‘database’ in (unnest(keyword-set))

Ø Note: Above syntax is valid in SQL:1999,  but the only collection type supported by SQL:1999 is the array type

n To get a relation containing pairs of the form “title, author-name” for each book and each author of the book

               select B.title, A
       
from books as B, unnest (B.author-array) as A

         

 

n We can access individual elements of an array by using indices

Ø E.g. If we know that a particular book has three authors, we could write:

          select author-array[1], author-array[2], author-array[3]
      from books
     
where title = `Database System Concepts’

 

Unnesting

 

n The transformation of a nested relation into a form with fewer (or no) relation-valued attributes us called unnesting.

n E.g.

      select title, A as author, publisher.name as pub_name,
             publisher.branch as pub_branch, K as keyword

      from books as B, unnest(B.author-array) as A, unnest (B.keyword-list) as K

Nesting

 

n  Nesting is the opposite of unnesting, creating a collection-valued attribute

n  NOTE: SQL:1999 does not support nesting

n  Nesting can be done in a manner similar to aggregation, but using the function set() in place of an aggregation operation, to create a set

n  To nest the flat-books relation on the attribute keyword:

    select title, author, Publisher(pub_name, pub_branch) as publisher,
           set(keyword) as keyword-list
from flat-books
groupby title, author, publisher

n  To nest on both authors and keywords:

      select title, set(author) as author-list,
            Publisher(pub_name, pub_branch) as publisher,
            set(keyword) as keyword-list
from   flat-books
groupby title, publisher

 

 

n Another approach to creating nested relations is to use subqueries in the select clause.

    select title,
      ( select author
         from flat-books as M
         where M.title=O.title) as author-set,
      Publisher(pub-name, pub-branch) as publisher,
      (select keyword
         from flat-books as N
         where N.title = O.title) as keyword-set
from flat-books as O

n Can use orderby clause in nested query to get an ordered collection

Ø Can thus create arrays, unlike earlier approach

 

Functions and Procedures

 

n SQL:1999 supports functions and procedures

Ø Functions/procedures can be written in SQL itself, or in an external programming language

Ø Functions are particularly useful with specialized data types such as images and geometric objects

ê E.g. functions to check if polygons overlap, or to compare images for similarity

Ø Some databases support table-valued functions, which can return a relation as a result

n SQL:1999 also supports a rich set of imperative constructs, including

Ø Loops, if-then-else, assignment

n Many databases have proprietary procedural extensions to SQL that differ from SQL:1999

SQL Functions

 

n Define a function that, given a book title, returns the count of the number of authors (on the 4NF schema with relations books4 and authors).

             create function author-count(name varchar(20))
       returns integer
      begin
           declare
a-count integer;
           select count
(author) into a-count
          
from authors
          
where authors.title=name
          
return a=count;
      
end

n Find the titles of all books that have more than one author.

          select name
     
from books4
     
where author-count(title)> 1

 

SQL Methods

 

n Methods can be viewed as functions associated with structured types

Ø They have an implicit first parameter called self which is set to the structured-type value on which the method is invoked

Ø The method code can refer to attributes of the structured-type value using  the self variable

ê E.g.    self.a

 

n  The author-count function could instead be written as procedure:

    create procedure author-count-proc (in title varchar(20),
                                                             out a-count integer)
     begin
      select count
(author) into a-count
         
from authors
         
where authors.title = title
   
end

n  Procedures can be invoked either from an SQL procedure or from embedded SQL, using the call statement.

Ø E.g. from an SQL procedure

          declare a-count integer;
      call author-count-proc(`Database systems Concepts’, a-count);

n  SQL:1999 allows more than one function/procedure of the same name (called name overloading), as long as the number of
arguments differ, or at least the types of the arguments differ

 

 

 

 

External Language Functions/Procedures

 

n SQL:1999 permits the use of functions and procedures written in other languages such as C or C++

n Declaring external language procedures and functions

    create procedure author-count-proc(in title varchar(20),
                                                            out count integer)
language C
external name’ /usr/avi/bin/author-count-proc’

create function author-count(title varchar(20))
returns integer
language C
external name ‘/usr/avi/bin/author-count’

 

 

n Benefits of external language functions/procedures: 

Ø more efficient for many operations, and more expressive power

n Drawbacks

Ø Code to implement function may need to be loaded into database system and executed in the database system’s address space

ê risk of accidental corruption of database structures

ê security risk, allowing users access to unauthorized data

Ø There are alternatives, which give good security at the cost of potentially worse performance

Ø Direct execution in the database system’s space is used when efficiency is more important than security

 

Security with External Language Routines

 

n To deal with security problems

Ø Use sandbox techniques

ê  that is use a safe language like Java, which cannot be used to access/damage other parts of the database code

Ø Or, run external language functions/procedures in a separate process, with no access to the database process’ memory

ê Parameters and results communicated via inter-process communication

n Both have performance overheads

n Many database systems support both above approaches as well as direct executing in database system address space

 

Procedural Constructs

 

n SQL:1999 supports a rich variety of procedural constructs

n Compound statement

Ø is of the form begin … end,

Ø may contain multiple SQL statements between begin and end.

Ø Local variables can be declared within a compound statements

n While and repeat statements

          declare n integer default 0;

          while n < 10 do

         set n = n+1

          end while

          repeat

           set n = – 1

          until n = 0

          end repeat

 

n For loop

Ø Permits iteration over all results of a query

Ø E.g. find total of all balances at the Perryridge branch

   declare n integer default 0;
  
for as
         select
balance from account
          where branch-name = ‘Perryridge’
    do
         set
n = n + r.balance
   
end for

 

 

n  Conditional statements  (if-then-else)
E.g. To find sum of balances for each of three categories of accounts (with balance <1000, >=1000 and <5000, >= 5000)

          if r.balance < 1000
           then set l = l + r.balance
     
elseif r.balance < 5000
           then set m = m + r.balance
     
else set h = h + r.balance
     
end if

n  SQL:1999 also supports a case statement similar to C case statement

n  Signaling of exception conditions, and declaring handlers for exceptions

          declare out_of_stock condition
      declare exit handler for
out_of_stock
     
begin
     

         ..  signal out-of-stock
      end

Ø The handler here is exit -- causes enclosing begin..end to be exited

Ø Other actions possible on exception

 

 

Comparison of O-O and O-R Databases

 

n Summary of strengths of various database systems:

n Relational systems

Ø simple data types, powerful query languages, high protection.

n Persistent-programming-language-based OODBs

Ø complex data types, integration with programming language, high performance.

n Object-relational systems

Ø complex data types, powerful query languages, high protection.

n Note: Many real systems blur these boundaries

Ø E.g. persistent programming language built as a wrapper on a relational database offers first two benefits, but may have poor performance.

 

Finding all employees of a manager

 

n  Procedure to find all employees who work directly or indirectly for mgr

n  Relation manager(empname, mgrname)specifies who directly works for whom

n  Result is stored in  empl(name)

    create procedure findEmp(in mgr char(10))
begin
      create temporary table
newemp(name char(10));
      create temporary table temp(name char(10));
      insert into newemp    
-- store all direct employees of mgr in newemp
             select empname
            
from manager
            
where mgrname = mgr
     

 

        repeat
       insert into
empl         -- add all new employees found to  empl
      select name
     
from newemp;

     insert into temp         -- find all employees of people already found
   (select manager.empname
    
from newemp, manager
   
where newemp.empname = manager.mgrname;
   )
  except (                   
-- but remove those who were found earlier
     select empname
    
from empl
 
);

         delete from newemp;  -- replace contents of newemp by contents of temp
      insert into newemp
       
select *
        from temp;
     delete from temp;

    until not exists(select* from newemp)   -- stop when no new employees are found
end repeat;
end

 

 

 

 

 

 

 

 

 

End of Chapter

 

 

 

0 comments:

Post a Comment