https://github.com/michaelgathara/maze-game

https://github.com/michaelgathara/maze-game

Last synced: 2 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/michaelgathara/maze-game
• Owner: Michaelgathara
• Created: 2023-10-03T03:43:32.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-12-03T07:34:41.000Z (over 1 year ago)
• Last Synced: 2025-01-26T05:16:41.802Z (4 months ago)
• Language: TeX
• Size: 31.2 MB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Maze Game Server

A server that lets us create a maze based on a [text file](./mazepool/0.maze) with co-ordinates of walls and takes a command as an input and writes out the current position of bot and the walls around it to stdin, and reads in co-ordinates for the bot to move towards from stdout. Then captures and visualizes the movements of the bot into a video.

![](maze.gif)

### Set-Up

1. Two ways to run the project:

   * Locally after installing all the dependecies

     * Install all the required dependencies, go to step 3.

   * In a docker container, using the Dockerfile in the repo, skip to step 4 after starting the container.

     * `docker build -t maze-game .`

     * `docker run --name=maze-game --detach maze-game`

     * Then you can use vscode to [connect](https://code.visualstudio.com/docs/devcontainers/containers) to attach to the container, go to step 4.

     * Or run `docker exec -it maze-game bash` to drop into bash inside the container, go to step 4.

2. Dependencies

   * [GraphicsMagick](www.graphicsmagick.org)

   * [FFmpeg](https://ffmpeg.org/)

   * [clang](https://clang.llvm.org/)

   * [boost](https://www.boost.org/)

   * [ghostscript](https://www.ghostscript.com/)

3. After installing the required dependencies, clone the repo.

4. run `make` to make the ./server executable.

5. run `make dfsbot` to create a visualization for a depth first search bot.

## Details

-  server can be run with `./server 'program/agent to run'` and communicates with the program passed to it using std-in/out.

   * example: `./server 'python dfbot.py'`

- For two players

  * It is `./server 'python3 agent1.py' 'python3 agent2.py'`

-  sends the sense data and bot location over std-in to be read and processed by the agent/program.

-  The agent should write the direction for the bot to move towards to std-out, which will be read by the server and bot will be moved, updated location of the bot is sent back to agent.

-  All the movements of the bot are captured into frames and stitched together into a video.

### Format of communication between server and agent

#### Bot Sense Information

- distance from walls in four directions is passed to std-in in this format

  - `wall x0 y0 x1 y1`

  - example: `wall 3.000000 1.000000 3.000000 2.000000`

- updated location of the bot

  - `bot x y num_coins`

  - example: `bot 2.014900 1.510483 3`

#### Adversarial Two Player information

- TWall

  - A temporary wall can be created to block the opponent agent in its track

  - This wall block the path for 15 seconds and costs 3 coins

  - Can only be deployed onto one of the walls in the current cell

  - `block tx ty [u|d|l|r]` example `block 1 2 d` creates bottom wall in the 1,2 cell

- Coins

  - Coins data is sent to the agent in the format

  - `Coin x0 y0`, the agent just needs to pass through it to collect it.

#### Agent to server

- direction to move the bot is sent over stdout(printed) in this format

  - `toward x y`

  - example: `toward 1.5 1.5`

### Server Defaults

- Server defaults to the `./mazepool/0.maze`, you can pick another by renaming the desired file to `0.maze` or changing the filename in `./maze-game-server.cpp`.

- The max number of seconds for the simulation, is set to 99. There is a chance that search wouldn't be complete in those seconds, you can increase|decrease it, but keep it mind, the resource consumption and time to process will change relatively.

- dfsbot.py can act as a template for agents, you can choose to use it as a base to make changes, or write one in another language based on it.

- server writes the mp4 file to out.mp4, it clears the out.mp4 at the start of next run for the new out.mp4.