Robot Game Problem

Author: Tsunghan Lee

Date: Apr 23 2017

Command usage:

Two methods are listed in the following.

Method 1:

Execute execute .jar file:

java -jar GameProblem.jar <option>

Method 2:

Compile:

javac GameProblem.java

Execute method 2:

java GameProblem <option>

Command arguments:

<option>  1 : for game version 1
          2 : for game version 2

Game Running Format:

Location Format: [<x>,<y>] (Note:no space in between)
  ex. Location: [2,3]

Direction faced Format: (Note: either <N> or <S> or <E> or <W>)
  ex. Direction faced: N

Actions Format: <M>,<L>,<R> (Note:no space in between):
  ex. Actions: M,M,M,L,M,R,R,R

Maximum actions allowed Format: (Note: Integer only)
  ex. Maximum actions allowed: 4

Design

Utilized Object-oriented design to achieve the goal of easy maintaining and developing new features. Five classes are included, and each class has its own tester to make sure that they perform correctly. When encountered Invalid input, such as illegal location, exceptions are handled. Class are listed in the following.

Coordinate:

Store coordinate information (x,y) and corresponding operations, i.e. move one step toward each direction.

Agent:

Store an agent's information (current coordinates) and actions, i.e. turn left, turn right, move one step forward.

Game:

Basic game information data, involving an agent, and methods to set initial condition, including location and direction faced.

Game1:

Extending Game class and add action-taking sequence method to form the version 1 game.

Game2:

Extending Game class and add new feature to perform path searching in order to become version 2 game.

Algorithm

In the game version 2, path searching is implemented in the sense of backtracking algorithm, which is similar to DFS with some modification. In each step, it checks whether the agent satisfies the target direction and target location. If it reaches the target, then add the path into the solution ArrayList. If not and doesn't reach the maximum actions allowed, it recursively do the searching. In the last case, the agent doesn't satisfy the targeting condition but reaches the maximum actions allowed, it remove the last step in the path, and return.

duke-leeth/Game