DOCSLIB.ORG

Lecture 40: Design Patterns

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Lecture 40: Design Patterns What is a Pattern? Lecture 40: "Each pattern describes a problem which occurs over and over again in our environment, and then describes the core of the solution to that problem, in such a Design Patterns way that you can use this solution a million times over, without ever doing it the same way twice" "Each pattern is a three-part rule, which expresses a relation between a certain CS 62 context, a problem, and a solution" Fall 2017 Christopher Alexander on architecture patterns Kim Bruce & Alexandra Papoutsaki "Patterns are not a complete design method; they capture important practices of existing methods and practices uncodified by conventional methods" James Coplien What are design patterns? Elements of Design Patterns • Pattern Name • Design pattern is a problem & solution in context • Problem statement - context where it might be • Design patterns capture software architectures and applied designs • Solution - elements of the design, their relations, - Not code reuse responsibilities, and collaborations. - Instead solution/strategy reuse - Template of solution - Sometimes interface reuse • Consequences: Results and trade-offs Example: Iterator Pattern Iterator Pattern • Consequences: • Name: Iterator or Cursor - Support different and simultaneous traversals • Problem statement - Multiple implementations of Iterator interface How to process elements of an aggregate in an - - One traversal per Iterator instance implementation independent manner • requires coherent policy on aggregate updates • Solution - Invalidate Iterator by throwing an exception, or - Aggregate returns an instance of an implementation of Iterator interface to control iteration. - Iterator only considers elements present at the time of its creation Goals of Patterns Taxonomy of Patterns • To support reuse, of • Creational patterns - Successful designs - concern the process of object creation - Existing code (though less important) • Structural patterns • To facilitate software evolution - deal with the composition of classes or objects - Add new features easily, without breaking existing ones • Behavioral patterns • Design for change! - characterize the ways in which classes or objects interact • Reduce implementation dependencies between and distribute responsibility. elements of software system. Creational Patterns Structural Patterns • Singleton • Adapter - Ensure a class only has one instance, and provide a global - Convert the interface of a class into another interface point of access to it. clients expect. Adapter lets classes work together that - Ofen used in recursively defined classes (e.g., lists & trees) where couldn't otherwise because of incompatible interfaces don’t have public constructor, just public constant defined using • Proxy private constructor Provide a surrogate or placeholder for another object to Abstract Factory - • control access to it - Provide an interface for creating families of related or • Decorator dependent objects without specifying their concrete classes. - Attach additional responsibilities to an object dynamically - Alows hiding actual constructor cal in method definition Behavioral Patterns • Template - Define the skeleton of an algorithm in an operation, deferring some steps to subclasses • Abstract superclass Creational Patterns • State - Allow an object to alter its behavior when its internal state changes. The object will appear to change its class • Observer - Define a one-to-many dependency between objects so that when one object changes state, all its dependents are notified and updated automatically Abstract Factory Abstract Factory (cont.) • Solution: • Context: - Create interface w/ operations to create new products of different kinds - System should be independent of how pieces created and represented - Multiple concrete classes implement operations to create concrete product objects. - Different families of components - Products also specified w/interface - Must be used in mutually exclusive and consistent way - Concrete classes for each interface and family of - Hide existence of different families from clients products. - Client uses only interfaces Abstract Factory (cont.) Abstract Factory Consequences • Isolate instance creation and handling from clients • Examples: • Can easily change look-and-feel standard - GUI Interfaces: - Reassign a global variable • Mac • Enforce consistency among products in each family • Windows XP Adding to family of products is difficult • Unix • - Have to update factory abstract class and all concrete classes Decorator Pattern • Motivation - Want to add responsibilities/capabilities to individual objects, not to an entire class. Structural Patterns - Inheritance requires a compile-time choice of parent class. • Solution - Enclose the component in another object that adds the responsibility/capability • The enclosing object is called a decorator. Decorator Pattern Decorator Pattern Consequences • Advantages - fewer classes than with static inheritance • A decorator forwards requests to its encapsulated - dynamic addition/removal of decorators component and may perform additional actions before or after forwarding. - keeps root classes simple • Disadvantages • Can nest decorators recursively, allowing unlimited added responsibilities. - proliferation of run-time instances abstract Decorator must provide common interface • Can add/remove responsibilities dynamically - • Tradeoffs: - useful when components are lightweight Decorator Example FileReader frdr= new FileReader(filename); LineNumberReader lrdr = new LineNumberReader(frdr); Behavioral Patterns String line; line = lrdr.readLine() while (line != null){ System.out.print(lrdr.getLineNumber() + ":\t" + line); line = lrdr.readLine() } Observer Pattern Observer Pattern • Solution structure • Problem - Subject is aware of its observers (dependents) - Objects that depend on a certain subject must be made - Observers are notified by the subject when something aware of when that subject changes changes, and respond as necessary • E.g. receives an event, changes its local state, etc. - Examples: Java event-driven programming - These objects should not depend on the implementation • Subject details of the subject - Maintains list of observers; defines a means for notifying • They just care about how it changes, not how it’s them when something happens implemented. • Observer - Defines the means for notification (update) Observer Pattern Observer Pattern Consequences class Subject { private Observer[] observers; public void addObserver(Observer newObs){... } • Low coupling between subject and observers public void notifyAll(Event evt){ - Subject indifferent to its dependents; can add or remove forall obs in observers do them at runtime obs.process(this,evt)} • Support for broadcasting } • Updates may be costly class Observer { public void process(Subject sub, Event evt) { - Subject not tied to computations by observers ... code to respond to event ... } } Visitor Pattern Visitor Pattern • Problem: want to implement multiple analyses on • We define each analysis as a separate Visitor class the same kind of object data - Defines operations for each element of a structure Spellchecking and Hyphenating Glyphs - • A separate algorithm traverses the structure, - Generating code for and analyzing an Abstract Syntax applying a given visitor Tree (AST) in a compiler - But, like iterators, objects must reveal their • Flawed solution: implement each analysis as a implementation to the visitor object method in each object • Separates structure traversal code from operations - Follows idea objects are responsible for themselves on the structure But many analyses will occlude the objects' main code - - Observation: object structure rarely changes, but often - Result is classes hard to maintain want to design new algorithms for processing .
Load more

Recommended publications
  • The Future of Embedded Software
    The Future of Embedded Software Edward A. Lee Professor, Chair of EE, and Associate Chair of EECS UC Berkeley ARTEMIS 2006 Annual Conference Graz, Austria May 22-24, 2006 Why Embedded Software? Why Now? “Information technology (IT) is on the verge of another revolution. Driven by the increasing capabilities and ever declining costs of computing and communications devices, IT is being embedded into a growing range of physical devices linked together through networks and will become ever more pervasive as the component technologies become smaller, faster, and cheaper... These networked systems of embedded computers ... have the potential to change radically the way people interact with their environment by linking together a range of devices and sensors that will allow information to be collected, shared, and processed in unprecedented ways. ... The use of [these embedded computers] throughout society could well dwarf previous milestones in the information revolution.” National Research Council Report Embedded Everywhere Lee, Berkeley 2 The Key Obstacle to Progress: Gap Between Systems and Computing | Traditional dynamic systems theory needs to adapt to better account for the behavior of software and networks. | Traditional computer science needs to adapt to embrace time, concurrency, and the continuum of physical processes. Lee, Berkeley 3 The Next Systems Theory: Simultaneously Physical and Computational The standard model: Embedded software is software on small computers. The technical problem is one of optimization (coping with limited resources). The Berkeley model: Embedded software is software integrated with physical processes. The technical problem is managing time and concurrency in computational systems. Lee, Berkeley 4 Obstacles in Today’s Technology: Consider Real Time Electronics Technology Delivers Timeliness… … and the overlaying software abstractions discard it.
    [Show full text]
  • One-Slide Summary Lecture Outline
    Using Design Patterns #1 One-Slide Summary • Design patterns are solutions to recurring OOP design problems. There are patterns for constructing objects, structuring data, and object behavior . • Since this is PL, we’ll examine how language features like (multiple) inheritance and dynamic dispatch relate to design patterns. #2 Lecture Outline • Design Patterns • Iterator • Observer • Singleton • Mediator #3 1 What is a design pattern? • A solution for a recurring problem in a large object-oriented programming system – Based on Erich Gamma’s Ph.D. thesis, as presented in the “gang of four” book • “Each pattern describes a problem which occurs over and over again in our environment, and then describes the core of the solution to that problem, in such a way that you can use this solution a million times over, without ever doing it the same way twice” – Charles Alexander #4 Types of design patterns • Design patterns can be (roughly) grouped into three categories: • Creational patterns – Constructing objects • Structural patterns – Controlling the structure of a class, e.g. affecting the API or the data structure layout • Behavioral patterns – Deal with how the object behaves #5 Iterator design pattern • Often you may have to move through a collection – Tree (splay, AVL, binary, red-black, etc.), linked list, array, hash table, dictionary, etc. • Easy for arrays and vectors • But hard for more complicated data structures – Hash table, dictionary, etc. • The code doing the iteration should not have to know the details of the data structure
    [Show full text]
  • Design Patterns Promote Reuse
    Design Patterns Promote Reuse “A pattern describes a problem that occurs often, along with a tried solution to the problem” - Christopher Alexander, 1977 • Christopher Alexander’s 253 (civil) architectural patterns range from the creation of cities (2. distribution of towns) to particular building problems (232. roof cap) • A pattern language is an organized way of tackling an architectural problem using patterns Kinds of Patterns in Software • Architectural (“macroscale”) patterns • Model-view-controller • Pipe & Filter (e.g. compiler, Unix pipeline) • Event-based (e.g. interactive game) • Layering (e.g. SaaS technology stack) • Computation patterns • Fast Fourier transform • Structured & unstructured grids • Dense linear algebra • Sparse linear algebra • GoF (Gang of Four) Patterns: structural, creational, behavior The Gang of Four (GoF) • 23 structural design patterns • description of communicating objects & classes • captures common (and successful) solution to a category of related problem instances • can be customized to solve a specific (new) problem in that category • Pattern ≠ • individual classes or libraries (list, hash, ...) • full design—more like a blueprint for a design The GoF Pattern Zoo 1. Factory 13. Observer 14. Mediator 2. Abstract factory 15. Chain of responsibility 3. Builder Creation 16. Command 4. Prototype 17. Interpreter 18. Iterator 5. Singleton/Null obj 19. Memento (memoization) 6. Adapter Behavioral 20. State 21. Strategy 7. Composite 22. Template 8. Proxy 23. Visitor Structural 9. Bridge 10. Flyweight 11.
    [Show full text]
  • Framework Overview with UML Diagrams
    Framework Overview with UML Diagrams Learn to Build Robust, Scalable and Maintainable Applications using PureMVC Framework Overview This document discusses the classes and interfaces of the PureMVC framework; illustrating their roles, responsibilities and collaborations with simple UML (Unified Modeling Language) diagrams. The PureMVC framework has a very narrow goal. That is to help you separate your application’s coding concerns into three discrete tiers; Model, View and Controller. In this implementation of the classic MVC design meta-pattern, the application tiers are represented by three Singletons (a class where only one instance may be created). A fourth Singleton, the Façade, simplifies development by providing a single interface for communications throughout the application. The Model caches named references to Proxies, which expose an API for manipulating the Data Model (including data retrieved from remote services). The View primarily caches named references to Mediators, which adapt and steward the View Components that make up the user interface. The Controller maintains named mappings to Command classes, which are stateless, and only created when needed. The Façade initializes and caches the Core actors (Model, View and Controller), and provides a single place to access all of their public methods. AUTHOR: Cliff Hall <[email protected]> LAST MODIFIED: 3/05/2008 Façade and Core The Façade class makes it possible for the Proxies, Mediators and Commands that make up most of our final application to talk to each other in a loosely coupled way, without having to import or work directly with the Core framework actors. When we create a concrete Façade implementation for our application, we are able to use the Core actors ‘out of the box’, incidental to our interaction with the Façade, minimizing the amount of API knowledge the developer needs to have to be successful with the framework.
    [Show full text]
  • Object-Oriented Analysis, Design and Implementation
    Undergraduate Topics in Computer Science Brahma Dathan Sarnath Ramnath Object-Oriented Analysis, Design and Implementation An Integrated Approach Second Edition Undergraduate Topics in Computer Science Undergraduate Topics in Computer Science (UTiCS) delivers high-quality instruc- tional content for undergraduates studying in all areas of computing and information science. From core foundational and theoretical material to final-year topics and applications, UTiCS books take a fresh, concise, and modern approach and are ideal for self-study or for a one- or two-semester course. The texts are all authored by established experts in their fields, reviewed by an international advisory board, and contain numerous examples and problems. Many include fully worked solutions. More information about this series at http://www.springer.com/series/7592 Brahma Dathan • Sarnath Ramnath Object-Oriented Analysis, Design and Implementation An Integrated Approach Second Edition 123 Brahma Dathan Sarnath Ramnath Department of Information and Computer Department of Computer Science Science and Information Technology Metropolitan State University St. Cloud State University St. Paul, MN St. Cloud, MN USA USA Series editor Ian Mackie Advisory Board Samson Abramsky, University of Oxford, Oxford, UK Karin Breitman, Pontifical Catholic University of Rio de Janeiro, Rio de Janeiro, Brazil Chris Hankin, Imperial College London, London, UK Dexter Kozen, Cornell University, Ithaca, USA Andrew Pitts, University of Cambridge, Cambridge, UK Hanne Riis Nielson, Technical University of Denmark, Kongens Lyngby, Denmark Steven Skiena, Stony Brook University, Stony Brook, USA Iain Stewart, University of Durham, Durham, UK A co-publication with the Universities Press (India) Private Ltd., licensed for sale in all countries outside of India, Pakistan, Bhutan, Bangladesh, Sri Lanka, Nepal, The Maldives, Middle East, Malaysia, Indonesia and Singapore.
    [Show full text]
  • The Problem with Threads
    The Problem With Threads Edward A. Lee Robert S. Pepper Distinguished Professor and Chair of EECS UC Berkeley -and - Senior Technical Adviser, director, and co-founder of BDTI Class #: ESC-211 Embedded Systems Conference (ESC) Tuesday, 3 April 2007 Concurrency in Software Practice, As of 2007 | Component technologies z Objects in C++, C#, or Java z Wrappers as service definitions | Concurrency z Threads (shared memory, semaphores, mutexes, …) z Message Passing (synchronous or not, buffered, …) | Distributed computing z Distributed objects wrapped in web services, Soap, CORBA, DCOM, … Lee, Berkeley & BDTI 2 z1 Observations Threads and objects dominate SW engineering. z Threads: Sequential computation with shared memory. z Objects: Collections of state variables with procedures for observing and manipulating that state. Even distributed objects create the illusion of shared memory through proxies. z The components (objects) are (typically) not active. z Threads weave through objects in unstructured ways. z This is the source of many software problems. Lee, Berkeley & BDTI 3 The Buzz “Multicore architectures will (finally) bring parallel computing into the mainstream. To effectively exploit them, legions of programmers must emphasize concurrency.” The vendor push: “Please train your computer science students to do extensive multithreaded programming.” Lee, Berkeley & BDTI 4 z2 Is this a good idea? Lee, Berkeley & BDTI 5 My Claim Nontrivial software written with threads, semaphores, and mutexes are incomprehensible to humans. Lee, Berkeley & BDTI 6 z3 To Examine the Problems With Threads and Objects, Consider a Simple Example “The Observer pattern defines a one-to-many dependency between a subject object and any number of observer objects so that when the subject object changes state, all its observer objects are notified and updated automatically.” Design Patterns, Eric Gamma, Richard Helm, Ralph Johnson, John Vlissides (Addison-Wesley Publishing Co., 1995.
    [Show full text]
  • Design Patterns Design Patterns
    Design Patterns CSC207 – Software Design Design Patterns • Design pattern: – A general description of the solution to a well- established problem using an arrangement of classes and objects. • Patterns describe the shape of code rather than the details. – There are lots of them in CSC 301 and 302. Loop patterns from first year • Loop pattern: – A general description of an algorithm for processing items in a collection. • All of you (hopefully) have some loop patterns in your heads. • You don’t really need to think about these any more; you just use them, and you should be able to discuss them with your fellow students. • Some first-year patterns: – Process List – Counted Loop – Accumulator – Sentinel Process list pattern • Purpose: to process every item in a collection where you don’t care about order or context; you don’t need to remember previous items. • Outline: • Example: • Other example: darken every pixel in a picture Counted loop pattern • Purpose: to process a range of indices in a collection. • Outline: • Example: • Other example: print indices of even-length string Accumulator pattern • Purpose: to accumulate information about items in a collection. • Outline: • Example: • Other examples: sum, min, accumulate a list of items meeting a particular criterion. Sentinel pattern • Purpose: to remove a condition in a loop guard. • Outline: • Example: Sentinel pattern, continued • Here is the code that Sentinal replaces; note that i != list.size() is evaluated every time through the loop, even though it is false only once. Design Pattern
    [Show full text]
  • Dependency Injection in Unity3d
    Dependency Injection in Unity3D Niko Parviainen Bachelor’s thesis March 2017 Technology, communication and transport Degree Programme in Software Engineering Description Author(s) Type of publication Date Parviainen, Niko Bachelor’s thesis March 2017 Language of publication: English Number of pages Permission for web publi- 57 cation: x Title of publication Dependency Injection in Unity3D Degree programme Degree Programme in Software Engineering Supervisor(s) Rantala, Ari Hämäläinen, Raija Assigned by Psyon Games Oy Abstract The objective was to find out how software design patterns and principles are applied to game development to achieve modular design. The tasks of the research were to identify the dependency management problem of a modular design, find out what the solutions offered by Unity3D are, find out what the dependency injection pattern is and how it is used in Unity3D environment. Dependency management in Unity3D and the dependency injection pattern were studied. Problems created by Unity3D’s solutions were introduced with examples. Dependency in- jection pattern was introduced with examples and demonstrated by implementing an ex- ample game using one of the available third-party frameworks. The aim of the example game was to clarify if the use of dependency injection brings modularity in Unity3D envi- ronment and what the cost of using it is. The principles of SOLID were introduced with generic examples and used to assist depend- ency injection to further increase the modularity by bringing the focus on class design. Dependency injection with the help of SOLID principles increased the modularity by loosely coupling classes even though slightly increasing the overall complexity of the architecture.
    [Show full text]
  • Facet: a Pattern for Dynamic Interfaces
    Facet: A pattern for dynamic interfaces Author: Eric Crahen SUNY at Buffalo CSE Department Amherst, NY 14260 201 Bell Hall 716-645-3180 <[email protected]> Context: Wherever it is desirable to create a context sensitive interface in order to modify or control the apparent behavior if an object. Problem: How can I modify the behavior of an existing object so that different behaviors are shown under different circumstances; and yet still maintain a clean separation between the policy (when each behavior is available) and implementation of each behavior allowing them to be developed independently of one another? Forces: Interfaces provide an essential level of abstraction to object oriented programming. Generally, objects are able define aspects of their function very well using interfaces. At times, this may not be enough. When dealing with two or more classes whose responsibilities are distinctly separate, but whose behavior is closely related, classes can begin to resist modularization. For example, adding security to an application means inserting code that performs security checks into numerous locations in existing classes. Clearly, these responsibilities are distinct; the original classes being responsible for the tasks they were designed to perform, and the security classes being responsible for providing access control. However, making those original classes secure means intermingling code that deals with security. When the classes dealing with security are changed many classes are going to be impacted. One method of adding new behavior to an existing class might be to simply create a subclass and embed that new behavior. In the security example, this would mean creating a subclass for each class that needs to be secure and adding calls to the security code.
    [Show full text]
  • Dependency Injection with Unity
    D EPEN DEPENDENCY INJECTION WITH UNITY Over the years software systems have evolutionarily become more and more patterns & practices D ENCY complex. One of the techniques for dealing with this inherent complexity Proven practices for predictable results of software systems is dependency injection – a design pattern that I allows the removal of hard-coded dependencies and makes it possible to Save time and reduce risk on your NJECT assemble a service by changing dependencies easily, whether at run-time software development projects by or compile-time. It promotes code reuse and loosely-coupled design which incorporating patterns & practices, I leads to more easily maintainable and flexible code. Microsoft’s applied engineering ON guidance that includes both production The guide you are holding in your hands is a primer on using dependency quality source code and documentation. W I injection with Unity – a lightweight extensible dependency injection TH DEPENDENCY INJECTION container built by the Microsoft patterns & practices team. It covers The guidance is designed to help U software development teams: various styles of dependency injection and also additional capabilities N I of Unity container, such as object lifetime management, interception, Make critical design and technology TY and registration by convention. It also discusses the advanced topics of selection decisions by highlighting WITH UNITY enhancing Unity with your custom extensions. the appropriate solution architectures, technologies, and Microsoft products The guide contains plenty of trade-off discussions and tips and tricks for for common scenarios managing your application cross-cutting concerns and making the most out of both dependency injection and Unity. These are accompanied by a Understand the most important Dominic Betts real world example that will help you master the techniques.
    [Show full text]
  • Design Patterns in Ocaml
    Design Patterns in OCaml Antonio Vicente [email protected] Earl Wagner [email protected] Abstract The GOF Design Patterns book is an important piece of any professional programmer's library. These patterns are generally considered to be an indication of good design and development practices. By giving an implementation of these patterns in OCaml we expected to better understand the importance of OCaml's advanced language features and provide other developers with an implementation of these familiar concepts in order to reduce the effort required to learn this language. As in the case of Smalltalk and Scheme+GLOS, OCaml's higher order features allows for simple elegant implementation of some of the patterns while others were much harder due to the OCaml's restrictive type system. 1 Contents 1 Background and Motivation 3 2 Results and Evaluation 3 3 Lessons Learned and Conclusions 4 4 Creational Patterns 5 4.1 Abstract Factory . 5 4.2 Builder . 6 4.3 Factory Method . 6 4.4 Prototype . 7 4.5 Singleton . 8 5 Structural Patterns 8 5.1 Adapter . 8 5.2 Bridge . 8 5.3 Composite . 8 5.4 Decorator . 9 5.5 Facade . 10 5.6 Flyweight . 10 5.7 Proxy . 10 6 Behavior Patterns 11 6.1 Chain of Responsibility . 11 6.2 Command . 12 6.3 Interpreter . 13 6.4 Iterator . 13 6.5 Mediator . 13 6.6 Memento . 13 6.7 Observer . 13 6.8 State . 14 6.9 Strategy . 15 6.10 Template Method . 15 6.11 Visitor . 15 7 References 18 2 1 Background and Motivation Throughout this course we have seen many examples of methodologies and tools that can be used to reduce the burden of working in a software project.
    [Show full text]
  • HEDGEHOG: Automatic Verification of Design Patterns in Java
    HEDGEHOG: Automatic Verification of Design Patterns in Java Alex Blewitt I V N E R U S E I T H Y T O H F G E R D I N B U Doctor of Philosophy School of Informatics University of Edinburgh 2006 Abstract Design patterns are widely used by designers and developers for building complex systems in object-oriented programming languages such as Java. However, systems evolve over time, increasing the chance that the pattern in its original form will be broken. To verify that a design pattern has not been broken involves specifying the original intent of the design pattern. Whilst informal descriptions of patterns exist, no formal specifications are available due to differences in implementations between programming languages. This thesis shows that many patterns (implemented in Java) can be verified automatically. Patterns are defined in terms of variants, mini-patterns, and artefacts in a pattern description language called SPINE. These specifications are then processed by HEDGEHOG, an automated proof tool that attempts to prove that Java source code meets these specifications. iii Acknowledgements I am indebted to Alan Bundy who has given me the freedom to work on this thesis whilst at the same time guiding me towards the final production and presentation of these results. I not would have been able to achieve this without Alan’s support through a sometimes difficult, but always busy part of my life. This project, and especially the production of this thesis, would not have been possible without the care and attention that Alan provided. Ian Stark has provided invaluable feedback on all aspects of this thesis, from the low-level technical intricacies of Java’s design patterns through to the high-level structure of the thesis as a whole.
    [Show full text]