↓ Ir para o conteúdo principal

← todas as notas

📎 Webclip

Strategy Pattern

The page explains the Strategy, or policy, pattern as a behavioral design pattern that lets an application choose among different algorithms at runtime. This flexibility comes from encapsulating algorithms as separate strategy objects and switching them through a setStrategy method, illustrated with a JavaScript AI game that alternates between aggressive, defensive, and balanced behaviors.

Reading notes
#

  • Strategy pattern is also known as the policy pattern and focuses on selecting an algorithm at runtime.
  • Deferring the choice of algorithm until runtime increases flexibility and reusability of the code.
  • The pattern fits scenarios where an application’s behavior must change dynamically while it is running.
  • Example behaviors mentioned are aggressive, defensive, and balanced modes.
  • The intent is to define multiple algorithms and make them interchangeable through a setStrategy() method.
  • It should be applied when classes differ only in behavior and when several variants of an algorithm are required.
  • Using strategy families allows sharing common parts through inheritance.
  • Applying the pattern helps avoid conditional statements that branch on behavior.
  • Clients of the pattern can pick which behavior they need by choosing the appropriate strategy object.
  • Downsides include an increased number of objects in the system.
  • Another drawback is that strategies may be shared between objects instead of being tightly bound to a single instance.
  • The pattern is highlighted as useful for encapsulating an algorithm.
  • It is also described as suitable for changing the algorithm while the application is running.
  • The JavaScript example defines concrete strategies AggressiveStrategy, BalancedStrategy, and DefensiveStrategy, each exposing a next method.
  • AIGameStrategy acts as the strategy context, with a setStrategy method to assign the current game strategy and a next method delegating to the selected strategy.
  • The main code initializes three strategy instances for aggressive, balanced, and defensive behaviors.
  • An AIGameStrategy instance holds the current strategy and forwards calls to the chosen strategy object.
  • The example demonstrates setting the strategy to defensive, then balanced, then aggressive, with console output reflecting the current behavior.
  • A GitHub link is provided to a repository named Design-Patterns-Strategy-Javascript containing the implementation.