ObjectOriented Technology Dr' Koichiro Isshiki Computer Information Department Cal Poly, Pomona kris - PowerPoint PPT Presentation

Title: ObjectOriented Technology Dr' Koichiro Isshiki Computer Information Department Cal Poly, Pomona kris


1
Object-Oriented TechnologyDr. Koichiro
IsshikiComputer Information DepartmentCal
Poly, Pomonakrisshiki_at_csupomona.edu

2
Outline

• Introduction of OO Concepts
• Introduction of UML
• Using Nouns
• Example 1 Electronic Filing Program
• Exercise Computer-Based Gas Station
• RDD Approach
• Example 2 Electronic Checkbook
• Example 3 Elevator Simulation System
• Exercise Coffee Machine Controller

3
Introduction of OO Concepts

4
Methodology

• A process for the organized production of
  software using a collection of predefined
  techniques and notational conventions.
• Early Years

Problem Domain
Solution Domain (Some Ad-hoc Model)
5
Structured Methodology
Problem Domain
Solution Domain data structure data
structure
Function 1
Function 2
Function 3
6
OO Methodology

Problem Domain
Solution Domain Message Message
Obj 1
Obj 2
Obj 3
7
Procedural vs. OO

• Procedural Approach - a program is a collection
  of procedures and functions
• read_from_tape ()
• read_from_file ()
• write_to_tape ()
• write_to_file ()
• Object-Oriented Approach - a program is a
  collection of cooperating objects
• For object tape, define operations read write
• For object file, define operations read write

8
Encapsulation - combine data and code

• e.g. visit to procedural doctor
• - go to library and get medical books
• - go to doctors office
• - Doc, here are the books and here are my
  symptoms
• - tell me my diagnosis
• e.g. visit to OO doctor
• - go to doctors office
• - Doc, here are my symptoms
• - tell me my diagnosis

9
Encapsulation (2)

• e.g. procedural bread making
• - get baking pan, mixing bowl, mixer
• - get flour, yeast, oil
• - get cookbook
• - follow directions pour, mix, knead, raise,
  knead, raise, bake, time, remove
• e.g. OO bread making
• - get flour, yeast, oil
• - pour into automatic bread maker
• - turn on

10
Traditional System Development

• Problems
• Software Projects
• Maintenance Cost Escalate
• Project Deadlines Slide
• Cost Estimate Difficult
• Processing
• Excessive Transformations
• Decreased Productivity

11
OO System Development

• Iterative rather than sequential
• User-centered analysis process
• Concepts of objects appear in analysis and is
  carried through design and implementation
• Early easy prototypes

12
Object-Oriented Analysis

• A problem domain
• A stage in the development cycle in which a
  real-world problem is examined to understand its
  requirements without planning the implementation
• The process of studying user needs to arrive at a
  definition of system or software requirements
• OOA contains the following activities
• finding the objects
• organizing the objects
• describing how the objects interact
• defining the operations of the objects

13
Object-Oriented Design

• A solution model
• The creation of an abstraction of a software
  system that is consistent with the requirements
  specifications and provides a reasonable
  description for implementation.
• A series of decisions on what is most
  cost-effective implementation
• OOD contains the following activities
• building system architecture
• providing full definition of classes user
  interface
• designing algorithms to implement class operations

14
OOD Guidelines

• Coupling - the strength of interdependencies
  between discrete components in a system.
• Its something to be minimized e.g. use
  encapsulation.
• For an OO system, we are concerned with the
  coupling between classes and objects that are not
  part of a gen-spec or whole-part hierarchy e.g.,
  we look for ways to minimize the number of
  messages between objects, as well as the
  complexity and content of the messages.

15
OOD Guidelines (2)

• Cohesion - the strength of association of
  elements within a system component (i.e. the
  degree to which the responsibilities of a single
  component form a meaningful unit)
• An entity, such as a class, an operation, or a
  module, is coherent if it is organized on a
  consistent plan and all its parts fit together
  toward a common goal.
• Its something to be maximized e.g., a single
  method should not contain both policy and
  implementation.

16
OOD Guidelines (3)

• Reuse
• Code reuse
• cut, copy, and paste
• subroutines and includes
• binary links
• Design reuse
• code reuse typically occurs at the bottom levels
  of a system design hierarchy while design reuse
  results in whole branches of the tree being
  reused
• Specification reuse
• eliminate completely the effort involved in
  designing, coding, and testing an implementation
  of that specification

17
Introduction of UML

18
UML

• Unified Modeling Language
• It unifies the methods of Booch, Rumbaugh, and
  Jacobson
• Model -- an abstraction of something for the
  purpose of understanding it before building it
• architectural models to show customers
• airplane models for wind-tunnel tests

19
UML (2)

• It is a modeling language, not a method
• A method consists of both a modeling language and
  a process
• The modeling language is the notation that
  methods use to express designs
• The process is their advice on what steps to take
  in doing a design

20
UML (3)

• Different parts of UML
• Views - show different aspects of the system
• Diagrams - graphs that describe the contents in a
  view
• Model elements - represent common oo concepts
  such as classes, objects, messages, and
  relationships
• General mechanisms - provide extra comments,
  information, or semantics about a model element

21
Views

• Use-case view A view showing the functionality
  of the system as perceived by external actors.
• Logical view A view showing how the
  functionality is designed inside the system, in
  terms of the systems static structure and
  dynamic behavior.

22
Diagrams

• For static parts of a system
• class diagram
• object diagram
• component diagram
• deployment diagram
• For dynamic parts of a system
• use case diagram
• sequence diagram
• collaboration diagram
• state diagram
• activity diagram

23
Static Dynamic

• All systems have a static structure and dynamic
  behavior.
• Static view represents structural, data aspects
  of a system. It does not describe the
  time-dependent behavior of the system.
• Dynamic modeling is to demonstrate how the
  objects interact dynamically at different times
  during the execution of the system, i.e. how the
  objects collaborate through communication and how
  the objects within the system change state during
  the systems lifetime.

24
Use Case Modeling

• Originates from Ivar Jacobsons methodology
  (OOSE)
• Purposes for use cases
• To provide a complete specification, from the
  users point of view, of how the system will
  function
• To provide a basis for performing system tests
  that verify the system
• Defines actors and use cases
• Users are classified as actors in the system. For
  each actor, specific use cases are played against
  the system. Collectively the use cases detail
  everything that the users can do with the system.

25
Use Case Modeling (2)

• Actor - someone or something that interacts with
  the system (send or receive messages to and from
  the system) an actor can be a human being or
  another system
• Finding actors - by answering a number of
  questions
• Who will use the main functionality of the
  system?
• Who will need support from the system to do
  their daily tasks?
• Who will need to maintain, administrate, and
  keep the system working?
• Which hardware devices does the system need to
  handle?
• With which other systems does the system need to
  interact?
• Who or what has an interest in the results that
  the system produces?

26
Use Case Modeling (3)

• Use Case - a set of sequences of actions a system
  performs that yield an observable result of value
  to a particular actor
• is always initiated by an actor (The actor must
  directly or indirectly order the system to
  perform the use case)
• provides value to an actor
• is complete (not a function call a use case is
  not complete until the end value is produced)
• Finding use cases - for each of the actors
  identified, ask
• Which functions does the actor require from the
  system?
• What does the actor need to do?
• What input/output does the system need? Where
  does this I/O come from or go to?

27
Use Case Modeling (4)

• Example - recycling machine
• Two actors the Customer and the Operator. Brian
  is the operator of the machine, so he normally
  acts as the Operator actor. However, sometimes he
  deposits his own bottles and cans, and then he
  acts as a Customer.
• Use cases
• Customers deposit cans and/or bottles.
• Customers print receipts.
• Customers collect money.
• Operators collect cans and/or bottles.
• Operators replace receipt paper.
• .

28
Use Case Diagram
Recycling Machine
Deposit cans/bottles
customer
Collect cans/bottles
Print receipt
Collect money
operator
29
Finding the Objects

• Using nouns (see example 1)
• Responsibility driven approach (see example 2,3)
• Using the things to be modeled (e.g. identify
  individual or group things, such as persons,
  organizations, locations, logs, reports, forms,
  etc.)
• Using decomposition
• Using generalization
• Reusing an application framework

30
Using Decomposition

• Patient -gt Surgical patient, Referral patient

Patient
referralPatient
nameString ssnNumber
surgicalPatient
rphys setPhysician() physicianIs()
surgery setSurgery() surgeryIs()
init() nameIs() SSNis()
31
Using Generalization

• Surgical patient, Referral patient -gt Patient

referralPatient
nameString ssnNumber
surgicalPatient
rphys setPhysician() physicianIs()
surgery setSurgery() surgeryIs()
init() nameIs() SSNis()
32
Reusing An Application Framework

• The steps of this technique follow
• identify one or more relevant application
  frameworks in the same application domain
• reuse both objects and classes from previously
  developed frameworks (note that some of the
  classes may need to be modified to be reused in
  your specific application)

33
Using Nouns

34
Example 1 Electronic Filing Program

35
The Electronic Filing System

• Problem Statement
• An electronic filing program (EFP) can be used to
  store and retrieve text documents. Any document
  created by a word processor may be stored in the
  electronic filing system. Documents may be filed
  along with keywords, authors, and/or a document
  description or abstract describing the document.
  Documents filed in the system may also be removed
  or deleted.

36
The EFP (2)

• Problem Statement (contd)
• Documents stored using the EFP are indexed to
  enable rapid retrieval. Users may retrieve or
  locate documents based on their content,
  description, author(s), or user-defined keywords.
  Therefore, the document description, authors,
  keywords, and/or the actual text document itself
  may be searched.
• A user may specify search criteria, which results
  in a number of documents being found that meet
  the specified search criteria. The user may then
  continue to specify additional search criteria,
  successively narrowing down the search until the
  required documents are found. Documents found
  that meet the users search criteria may then be
  viewed or printed.

37
The EFP (3)

• Problem Statement (contd)
• The user is provided with the capability of
  specifying any junk words, which if found in
  the content of the document will not be searched
  or indexed - for example, and, or, not, the, if.
  The user can also specify which alphanumeric
  characters in the text document will be indexed
  and searchable (the filing character set),
  thereby limiting the search and index to only
  portions of a document(s).

38
Identifying Object Classes

• Grammatical Inspection - nouns/noun phrases
  represent candidate objects/classes
• Potential Object Classes
• abstract, alphanumeric character, author,
• content, description, document, document
  description,
• EFP, electronic filing system, filing character
  set, index,
• junk word, keyword, number of documents,
• portion of the documents, search, search
  criteria,
• system, text document, user, user-defined
  keyword,
• word processor
• By general knowledge
• line, word, page

39
Identifying Object Classes (2)

• Problem Statement
• An electronic filing program (EFP) can be used to
  store and retrieve text documents. Any document
  created by a word processor may be stored in the
  electronic filing system. Documents may be filed
  along with keywords, authors, and/or a document
  description or abstract describing the document.
  Documents filed in the system may also be removed
  or deleted.

40
Identifying Object Classes (3)

• Problem Statement (contd)
• Documents stored using the EFP are indexed to
  enable rapid retrieval. Users may retrieve or
  locate documents based on their content,
  description, author(s), or user-defined keywords.
  Therefore, the document description, authors,
  keywords, and/or the actual text document itself
  may be searched.
• A user may specify search criteria, which results
  in a number of documents being found that meet
  the specified search criteria. The user may then
  continue to specify additional search criteria,
  successively narrowing down the search until the
  required documents are found. Documents found
  that meet the users search criteria may then be
  viewed or printed.

41
Identifying Object Classes (4)

• Problem Statement (contd)
• The user is provided with the capability of
  specifying any junk words, which if found in
  the content of the document will not be searched
  or indexed - for example, and, or, not, the, if.
  The user can also specify which alphanumeric
  characters in the text document will be indexed
  and searchable (the filing character set),
  thereby limiting the search and index to only
  portions of a document(s).

42
Discarding Unnecessary Incorrect Classes

• Redundant Classes - keep the most descriptive
  object/class name for object classes that express
  the same information
• Irrelevant Classes - classes that have little or
  nothing to do with the problem domain should be
  eliminated
• Vague Classes - classes should be specific
  (should not have ill-defined boundaries or be too
  broad in scope)

43
Discarding Unnecessary Incorrect Classes (2)

• Attributes - attributes are properties of object
  classes and not object classes themselves
• Operations - a name describes an operation that
  is applied to objects should be discarded as a
  potential object class
• Implementation Constructs - anything that has to
  do with how the problem is solved should not be a
  part of the analysis model

44
Redundant Classes

• Document and text document are redundant. Text
  document is retained because it is more
  descriptive. Content refers to the document
  itself and is redundant with text document.
• Filing character set and alphanumeric characters
  are redundant. Filing character set is retained
  because its more descriptive and makes more
  sense in the application domain.

45
Redundant Classes (2)

• Description, document description, and abstract
  are redundant. Abstract is retained because it is
  a document description and is a familiar term in
  business and academia.
• User-defined keyword and keyword are redundant.
  All keywords, documents, and authors must be
  specified by the user in the EFP. Therefore,
  user-defined is not relevant and keyword will be
  retained.

46
Irrelevant Classes

• Word processor is irrelevant since the creation
  of a document is outside the scope of the EFP
  software.
• Portion of the document and number of documents
  have nothing to do with the problem and will also
  be eliminated.
• User also has little to do with the problem. All
  systems have users, and later on we will need to
  create a user interface.

47
Vague Classes

• System is too broad in scope. The system is
  really the EFP, therefore system will be
  discarded as a potential object class.
• EFP and electronic filing system are rather broad
  in scope. It is the application we are attempting
  to build, and not an object class in itself.

48
Attributes

• A search operation requires search criteria in
  order to perform the search, but beyond that
  search criteria does not affect the problem. So
  search criteria should be treated as an attribute
  of a search operation. Search criteria is
  discarded as a potential object class.

49
Operations

• Search involves a sequence of actions to locate
  one or more documents satisfying some search
  criteria. Search is discarded as a potential
  object class.

50
Implementation Constructs

• Filing character set is an implementation
  construct. The exact representation of the
  collection of filing characters is a design
  issue. Filing character set is renamed filing
  character.

51
Candidate Object Classes

• abstract keyword
• author line
• filing character page
• index text document
• junk word word

52
Data Dictionary

• The data dictionary provides a place to collect
  information about an object class that is
  considered important and to clarify the meaning
  of the object class.
• Abstract - Brief statement of the essential
  thoughts of the document. The abstract may
  consist of several sentences describing what the
  specific document is all about.
• Author - Name of the person who wrote or
  originated the document. More than one author may
  be associated with a document.

53
Data Dictionary (2)

• Filing character - An alphanumeric character that
  is recognized by the EFP. Some applications in
  which electronic filing is used may require that
  only specific characters or numbers be indexed
  and retrievable. Only electronic filing
  characters are indexed and retrievable.
• Index - Each document is indexed in some manner
  to enable rapid retrieval of data. The indexing
  technique is not specifed at this point in time.

54
Data Dictionary (3)

• Junk word - A word that is common in the
  vocabulary of the language and is not of any
  value for retrieval purposes. Junk words are
  therefore not indexed. Examples of junk words are
  the following and, the, or, if, is, are.
• Keyword - A word that helps the user to uniquely
  identify a document. Multiple keywords may be
  associated with a document.

55
Data Dictionary (4)

• Line - Text documents are made of up to lines of
  words. A document must have at least one line.
  Each line in a text document may contain zero or
  more words.
• Page - A text document that is not an empty
  document is made up of one or more pages. Each
  page consists of one or more lines.

56
Data Dictionary (5)

• Text document - Document possibly created by an
  editor, word processor, desktop publishing tool,
  etc. A text document is divided into pages. A
  document must have one ore more pages. A page may
  consist of multiple lines. Each line may consist
  of multiple words. A word consists of one or more
  letters.
• Word - An alphabetic character or some meaningful
  combination of alphabetic characters that
  represents a unit of language.

57
Adding AssociationsBetween Classes

• Grammatical Inspection - verb/verb phrases
  represent potential associations between object
  classes
• Explicit verb phrases
• Implicit verb phrases
• Knowledge of problem domain
• (Note that Associations that do not come
  directly from the problem statement should be
  verified with the user and the problem domain
  expert.)

58
Explicit verb phrases

• Documents filed in system Documents filed with
  abstract
• Documents filed with author Documents filed with
  keyword
• Documents removed/deleted from system
• Documents is retrievable EFP indexes documents
• EFP retrieves text documents EFP stores text
  documents
• User prints documents
• User specifies indexable/searchable characters
• User specifies junk words User specifies search
  criteria
• User views documents Word processor creates
  document

59
Implicit verb phrases

• Document referenced by index
• Documents created external to EFP
• Documents must be text only
• Documents retrievable by abstract
• Documents retrievable by author
• Documents retrievable by content
• Documents retrievable by keyword
• Search criteria has abstract
• Search criteria has author
• Search criteria has keyword
• Search criteria has word

60
Knowledge of problem domain

• Abstract describes document
• Documents contain pages
• Keyword identifies document
• Lines contain words
• Pages contain lines
• Text document identified by author
• Words contain alphanumeric characters

61
Discarding Unnecessary Incorrect Associations

• Associations between eliminated classes - any
  association stated in terms of those eliminated
  classes must also be eliminated or restated in
  terms of some other class
• Irrelevant or Implementation associations - any
  associations dealing with implementation
  constructs or outside the problem domain are
  eliminated
• Actions - potential associations describing
  transient events are discarded only those that
  describe a permanent relationship between object
  classes should be retained

62
Associations Between Eliminated Classes

• The EFP was eliminated as a potential object
  class, so we can eliminate EFP indexes documents,
  EFP retrieves text documents, EFP stores text
  documents, Documents created external to EFP.
• User was eliminated, so we can eliminate User
  prints documents, User specifies
  indexable/searchable characters, User specifies
  junk words, User specifies search criteria, User
  views documents.
• Any associations specifying document are changed
  to text document.

63
Irrelevant or Implementation associations

• Documents must be text only is eliminated because
  it is outside the domain of the problem.
  Documents are created by programs other than the
  EFP, such as editors, word processors. Documents
  that are not recognized as text will be rejected
  by the EFP.

64
Actions

• Documents filed with abstract, Documents filed
  with author, Documents filed with
  keyword,Documents retrievable by abstract,
  Documents retrievable by author, Documents
  retrievable by keyword describe interactions
  between the user and the EFP and not permanent
  relationships. They are all eliminated.

65
Candidate Associations

• Abstract describes document
• Documents contain pages
• Document referenced by index
• Keyword identifies document
• Lines contain words
• Pages contain lines
• Text document identified by author
• Words contain alphanumeric characters
• Note that a number of these associations actually
  represent aggregation.

66
A New Object Class

• Alphanumeric character was discovered while
  analyzing potential associations between object
  classes. It was discarded as a potential object
  class in favor of filing character. It now seems
  more relevant since word contains alphanumeric
  characters which may be nonfiling.
• The discovery of new object classes means the
  discovery of new associations - Word contains
  alphanumeric character.

67
Other Missing Associations

• Other associations are further identified between
  object classes
• abstract contains lines
• line contains junk words
• index references text document

68
Initial EFP Class Diagram

references
identifies

1..
Author
Index

identifies


Text document
Keyword
describes
1..
0..1
Page
Abstract
1..
1..
Line
Junk word




Word




Alphanumeric char
filing char
69
Adding Attributes

• Attributes are data values that are held by the
  objects in a class.
• Look for nouns followed by possessive phrases
  (e.g., the length of the bridge, the width of the
  road, the voltage of the computer, the number of
  characters in the document) in the problem
  statement to identify attributes.
• Attributes can also be found by drawing on your
  knowledge of the application domain and the real
  world.

70
Potential Attributes

• Based on knowledge of application domain
• Attribute Object
• Author name Author
• Character Filing character
• Character Alphanumeric character
• Date created Text document
• Document description Abstract
• Document file name Text document
• Document size Text document
• Index file name Index

71
Potential Attributes (2)

• (contd)
• Attribute Object
• Index size Index
• Keyword name Keyword
• Line number Line
• Number of characters Word
• Page number Page
• String Word
• Word Junk word

72
Discarding Unnecessary Incorrect Attributes

• Objects - entities that have features of their
  own are objects, not attributes. Attributes are
  pure data values. (no objects disguising
  themselves as attributes in our attributes list)
• Qualifiers - a qualifier is a special attribute
  that reduces the effective multiplicity of an
  association and increase the semantic accuracy.
  Any attributes that appear to be qualifiers
  should be restated as such.

73
Discarding Unnecessary Incorrect Attributes (2)

• Link Attributes - if an attribute depends on the
  presence of the link, then it is a link attribute
  and not an object attribute. Use association
  class for link attributes. (no link attributes
  masquerading as object attributes in our list)
• Internal Values - attributes that describe the
  internal state of an object and are invisible
  outside the object should not be part of the
  object model during analysis. Postpone this level
  of detail until design.

74
Qualifiers

• Qualification improves semantic accuracy. For
  instance, it is more informative to be told that
  a directory and file name combine to identify a
  file, rather than be told that a directory has
  many files.

Directory
file
file name


Directory
file
75
Qualifiers (2)

• Author name, keyword name, and index file name
  qualify the associations - author identifies text
  document, keyword identifies text document, and
  index references text document, respectively.
  Author name, keyword name, and index file name
  are changed to qualifiers.

76
Link Attributes

• A link attribute is a property of the links in an
  association. For instance, access permission is
  an attribute of Accessible by since access
  permission is a joint property of File and User,
  and cannot be attached to either File or User
  alone without losing information.

Accessible by
User


File
access permission
/etc/termcap (read) John Doe
/etc/tempcap (read-write) Mary Brown
/usr/doe/.login (read-write) John Doe
77
Link Attributes (2)

• Another link attributes example

Works-for
Person
Company


boss
name ssn address
0..1
name address
salary job title

worker
performance rating
78
Internal Values

• Size of the index object, size of the text
  document object, and number of characters in the
  word object are attributes that describe the
  internal states of the objects. They are
  eliminated from our analysis model but are added
  to the data dictionary so that they may be
  incorporated later into our object design model.

79
Candidate Attributes

• Candidate Attribute Object
• Character ASCII character
• Date created Text document
• Document description Abstract
• Document file name Text document
• Line number Line
• Page number Page
• String Word
• Word Junk word

80
Use Inheritance to Share Common Structure

• Note that filing characters will most likely be
  alphanumeric characters. Therefore, a word should
  consist of filing and nonfiling characters. Both
  filing and nonfiling characters are ASCII
  characters.
• However, showing filing character and nonfiling
  character as subclasses of ASCII character would
  be trivial inheritance. The filing character and
  nonfiling character subclasses would not have
  distinctive attributes, associations, or
  behavior. We choose the representation to show
  ASCII character as having attribute, character,
  which is of the type filing or nonfiling.

81
Initial EFP Class Diagram with Attributes
identifies

references

Text document
Author
author name
Index
identifies
index file name
document file name creation date
Keyword
keyword name

1..
describes
Page
0..1
Abstract
page number
description
1..
Line
Junk word
1..


line number
word


Ascii character

1..
Word
string
characterfiling or nonfiling
82
Summary of Deliverables

• The deliverables for developing the object model
  are
• the problem statement
• the initial data dictionary
• the object model

83
Identifying Actors Use Cases

• The major actor in the context of the EFP is the
  user. No physical devices (other than the parts
  of the computer system itself) or other computer
  applications are involved. Our scenarios will
  therefore be based on interactions between the
  user and the EFP.
• A set of use cases for the EFP
• filing a document
• searching for, or finding, a document
• viewing a document
• deleting a document
• printing a document
• changing the filing character set
• changing the junk words

84
Use Case Diagram

Filing a doc
Finding a doc
Viewing a doc
Deleting a doc
Printing a doc
user
Changing filing char set
Changing junk word
85
Filing a Document

• Scenario 1. Normal scenario for filing a
  document.
• The user wishes to file a document.
• The user specifies a directory and a document
  file.
• The user enters an abstract of the document,
  keywords, and/or authors.
• The user requests the document to be filed.
• The document is filed.
• A possible format for the filing a document use
  case is shown in the following two figures.

86
Filing a Document (2)

File a Document
File Name
path
OK
Files
Directories
jim.wri susan.wri article.doc computer.rpt
windows borland (..)
(a) (b) (c)
Cancel
Reference Information
87
Filing a Document (3)

Document Reference Information
Document Abstract
Keywords
Authors
Cancel
OK
88
Filing a Document (4)

• Scenario 2. Filing scenario with exceptions.
• The user wishes to file a document.
• The user specifies a directory.
• The user requests the document to be filed.
• The EFP displays an error message and asks the
  user to specify both a directory and file name
  for the document.
• The user cancels the operation.

89
Searching for a Document

• Scenario 3. Normal scenario requiring a single
  search for a document.
• The user wishes to search for a document(s)
  meeting some specific criteria.
• The user enters an abstract query for an abstract
  search, a word or phrase for a content search,
  one or more keywords for a keyword search, and/or
  one or more authors for an author search.
• The user requests the search to start.
• The EFP returns the names of the documents
  satisfying the search criteria.
• The user selects a document.
• The user views a document satisfying the search
  criteria.
• The user ends the request for a search operation.

90
Searching for a Document (2)

• Scenario 4. Search scenario with exceptions.
• The user wishes to search for a document(s)
  meeting some specific criteria.
• The user requests the search to start.
• A message box is displayed showing an error
  message and asking the user to enter abstract,
  content, keyword, and/or author information as a
  query.
• The user ends the request for a search operation.

91
Searching for a Document (3)

• Scenario 5. Normal scenario requiring multiple
  searches for a document.
• The user wishes to search for a document(s)
  meeting some specific criteria.
• The user enters one or more keywords for a
  keyword search.
• The user requests the search to start.
• The EFP returns the names of the documents
  satisfying the search criteria.
• The user selects a document.
• The user views a document satisfying the search
  criteria.
• The user enters additional keywords to narrow
  down the search.
• The user requests the search to start.
• The EFP returns the names of the documents
  satisfying the search criteria.
• The user selects a document.
• The user views a document satisfying the search
  criteria.
• The user ends the request for a search operation.

92
Searching for a Document (4)

Search for A Document(s)
Search Results
Content Query
Document Name
View
(and)
Abstract Query
jim.wri computer.doc
(and)
Keyword Query
Print
(and)
Author Query
Done
Search
Done
93
Viewing a Document

• Only documents that have been retrieved from a
  search may be viewed by the EFP. Scenario 3 shows
  an example of a normal scenario for viewing a
  document found from a search. Scenario 6 shows an
  example with exception.
• Scenario 6. Viewing scenario with exceptions.
• The user wishes to search for document(s) meeting
  some specific criteria.
• The user enters a query.
• The user requests the search to start.
• The EFP returns the names of the documents
  satisfying the search criteria.
• The user tries to view a document satisfying the
  search criteria.
• An error message is displayed indicating that a
  document must first be selected.
• The user ends the request for a search operation.

94
Deleting a Document

• Only documents that have been filed by the EFP
  may be deleted. The EFP deletes only the index
  for the document and not the document itself.
• Scenario 7. Normal scenario for deleting a
  document.
• The user wishes to delete a document.
• The user specifies a document.
• The user requests the document to be deleted.
• The document index is deleted.
• Scenario 8. Delete scenario with exceptions.
• The user wishes to delete a document.
• The user requests the document to be deleted.
• An error message is displayed requesting the user
  to specify a document.
• The user cancels the operation.

95
Deleting a Document (2)

Delete a Document
File Name
path
Files
Directories
jim.wri susan.wri article.doc computer.rpt
OK
windows borland (..)
(a) (b) (c)
Cancel
96
Printing a Document

• A user may only print a document that has been
  retrieved as part of a search. Scenarios 3 and 5
  may be used to print a document by simply
  replacing views with prints.

97
Changing the Character Set

• As stated in the problem statement, the user can
  specify which alphanumeric characters will be
  indexed and searchable, thereby limiting the
  search and index to only portions of a
  document(s).
• Scenario 9. Normal scenario for changing the
  character set.
• The user wishes to add a character to the filing
  character set.
• The user specifies a character to include in the
  filing character set.
• The user requests the character to be included.
• The character is included in the character set.
• Scenario 10. Character set scenario with
  exceptions.
• The user wishes to add a character to the filing
  character set.
• The user specifies a character to include in the
  filing character set.
• The user cancels the operation.

98
Changing the Character Set (2)

Changing Filing Characters
Characters
Character
c d _at_ A
OK
Discard
Include
Cancel
99
Changing the Junk Words

• As stated in the problem statement, the user can
  specify any junk words, which will not be
  searched or indexed - for example, and, or, not,
  the, if.
• Scenario 11. Normal scenario for changing junk
  words.
• The user wishes to add a new junk word to the
  programs set of junk words.
• The user enters a junk word.
• The user requests the word to be added.
• The word is added to the junk word set.
• Scenario 12. Junk word scenario with exceptions.
• The user wishes to add a new junk word to the
  programs set of junk words.
• The user enters a junk word.
• The user cancels the operation.

100
Changing the Junk Words (2)

Changing Junk Words
Junk Words
Junk Word
and the if or him then was
Add
Delete
OK
Cancel
101
Preparing Sequence Diagrams

• Sequence diagrams describe how groups of objects
  interact and collaborate, emphasize the time
  order of messages
• Sequence diagrams are constructed by
• identifying events (or messages) in scenarios
• identifying the actor/object that sends an event
• identifying the actor/object that receives an
  event

102
Sequence Diagram for Filing a Document
An User
A User Interface
A Text Document

request file operation
specify directory file name
enter abstract, keywords, and/or authors
request document to be filed
process filing request
document filed
document filed
103
Sequence Diagram for Searching a Document
An User
A User Interface
A Text Document

request search operation
enter abstract, content, keyword, and/or author
queries
request search to start
find document(s)
list of documents meeting search criteria
list of documents meeting search criteria
select a document
view a document
Note that this is also the sequence diagram for
viewing (printing) a document
end request for search
104
Sequence Diagram for Deleting a Document

An User
A User Interface
A Text Document
request delete operation
specify directory file name
request document index deletion
delete document index
document index deleted
document index deleted
105
Sequence Diagram for Changing Character Set

An User
A User Interface
request filing character operation
specify filing character to include
character included in set
106
Sequence Diagram for Adding Junk Words

An User
A User Interface
request junk word operation
enter junk word
request junk word to be added
junk word is added
107
State Diagrams

• State Diagram (SD) - a set of states and events
  and their relationships to describe the behavior
  of a system (usually one state diagram for each
  object class with nontrivial dynamic behavior).
• An event is something that happens at some point
  in time, e.g. left-mouse-button is clicked.
• The attribute values held by an object represent
  the state of the object, e.g. the state of a
  person is healthy or unhealthy.
• Event has no duration state occupies an interval
  of time.

108
State Diagrams (2)

• States and events are dual of one another (i.e.
  an event separates two states and a state
  separates two events).
• State diagram for light switch

Flip switch
off
on
Flip switch
109
State Diagrams (3)

• The state of the system is the product of the
  states of all the objects (not represented by a
  single state of a single object).
• General form

Event guard /action
State 1 entry/action do/activity exit/action
State 2 entry/action do/activity exit/action
110
User Interface Class State Diagram

Start
selection
User Interface
File Document
Delete Document
Search
Filing Character
Junk Word
111
File Document State Diagram
File Document

Document Data Input
Path Input
dorequest file name
file name entered/record file name
cancel operation selected /cancel operation
dorequest directory name
directory name entered/record directory
name
document reference information requested
Document Reference Information
File dofile
document
dorequest keywords
keywords entered/record keywords
dorequest authors
authors entered/record authors
file document operation selected directoryfile
name entered
dorequest abstract
abstract entered/record abstract
112
Delete Document State Diagram
Delete Document

Path Input
dorequest file name
file name entered/record file name
dorequest directory name
directory name entered/record directory
name
delete operation selected directoryfile name
entered
cancel operation selected /cancel operation
Delete dodelete
document
113
Search State Diagram
Search

Query Input
dorequest content query
content query entered/record query
cancel operation selected /cancel operation
dorequest keyword query
keyword query entered/record query
dorequest abstract query
abstract query entered/record query
end search
start search query entered
dorequest author query
author query entered/record query
Find Documents dolook for
document
list of document names generated
complete examining search results
Search Results Visible
view document
dopresent document in window
dorequest file name
file name entered/record file name
viewing done
print document
dorequest directory name
directory name entered/record
directory name
doprint text document
print done
114
Junk Word State Diagram

Junk Word
dorequest junk word
junk word entered/record junk word
cancel operation selected /cancel operation
delete
junk word operation complete
doremove junk word
add
doinclude junk word
115
Filing Character State Diagram

Filing Character
dorequest filing character
filing character entered/record
filing character
cancel operation selected /cancel operation
delete
filing character operation complete
doremove filing character
include
doadd filing character
116
Updating the Object Model

• A user interface is an application object. The
  application object does not need to be included
  in the initial object model, but should be added
  to the object model after making up the dynamic
  model.

117

Revision of EFP Object Model
User interface
file document, search for document, delete a
document, add/delete junk word, include/discard
filing character, view document, print document

references
identifies

Text document

Author
author name
Index
index file name
identifies
document file name creation date
Keyword
keyword name

1..
describes
Page
0..1
Abstract
page number
description
1..


Line
Junk word
1..
line number
word


Ascii character

Word
1..
string
characterfiling or nonfiling
118
Summary of Deliverables

• The deliverables for developing the dynamic model
  are
• scenarios of typical interaction sequences
• sequence diagrams for each scenario
• state diagrams for those classes with important
  dynamic behavior
• a revised object model (if necessary)

119
Key Operations

• File and find are the two major functions of the
  EFP system.
• For the file operation, the processes include
• validate document type
• create text document index
• For the find operation, the processes include
• build search vector (for content query)
• check for document matches (for content query)
• check for document reference information matches
  (for abstract, keyword, author queries)
• build document name list, i.e., the searching
  result

120
Missing Classes

• Hash engine
• Hash document record process has significant
  computation. We create a class, called hash
  engine, to collect all of the hashing functions
  together in one place.
• This will minimize the impact on the system as
  these hashing algorithms are optimized.
• Hash engine creates an index of a document.

121
Missing Classes (2)

• Search engine
• The search engine class collects the search
  functions together in one place to facilitate a
  search.
• The operations of search engine operate on the
  content query, search vector, filing directory,
  and document name list object classes.
• Search vector
• Build search vector is a top-level process that
  is implemented by several low-level processes.
• A search vector is produced by a search engine.

122
Missing Classes (3)

• Content query
• A content query is a collection of phrase(s) and
  query operator(s). It allows users to check for
  document matches. It has features of its own and
  is modeled as a class.
• A content query will be processed by a search
  engine.
• Document name list
• Document name list class has an attribute that is
  a name list, generated as the result of a search.
• Document name list is produced from search engine.

123
Missing Classes (4)

• Document reference information
• Document reference information will be used to
  check for document reference information matches
  (in a find operation).
• A document reference information class consists
  of zero or more author objects, zero or more
  keyword objects, and zero or more abstract. So it
  has an aggregate relationship with the author,
  keyword, and abstract objects.

124
Adding Operations to the OM

• Operation Object Class
• file document text document
• find document search engine
• create text document index document index
• build search vector search engine
• check for document matches content query
• check for doc ref info matches doc ref
  information
• hash document record hash engine
• build document name list search engine

125
Adding Operations to the OM (2)

• Operation Object Class
• view document text document
• print text document text document
• delete document index filing directory

126
Another Missing Classes

• Filing directory
• filing directory object class has zero or more
  index object classes

User interface

Filing directory
file document search for
document delete a document
add/delete junk word include/discard filing
character view document
print document
index file name
references
Index

index file name
Text document
127

Revision 2 of EFP Object Model
Document name list

0..1

Search engine
User Interface
Filing directory
1..
build search vector build doc name list find doc

delete index
Operates on
Index file name
Index file name


Text document
yield
doc name dir name creation date
describes
Document reference information
utilizes
0..1

references
Index file name
1..
0..1
Index
Search vector
file doc view doc print doc
check for document ref info matches
vector
create text doc index

build search vector


1..

1..


Page
1..
Author
Keyword
0..1
Hash engine
Query
1..
1..
Abstract
hash doc record
Line

Word

String
ASCII char
1..
1..
Charfiling or nonfiling
Junk word
Nonjunk word

128
Designing the Application Interface

• The steps in designing the interactive interface
  include
• Isolate user interface objects from the objects
  that define the semantics of the application.
• Use the dynamic model as the structure of the
  program.
• Fully specify the application functions invoked
  by the user interface.

129
Menu Bar Pull-Down Menus

File Document Storage Document Retrieval
Utilities Help
Search...
Filing characters Junk words...
Exit
File Delete...
Contents Using Help Commands
Dialog Boxes Junk Words Filing
Characters
About Electronic Filing
130
User Interface Class Diagram
Window
exit file
delete search
filing characters junk words help

Dialog

File a Doc
Doc Ref Info
Search
Delete
Junk Word
OK cancel reference info
OK cancel
search cancel done
OK cancel
add delete OK