Skip to content

Latest commit

 

History

History
157 lines (101 loc) · 5.97 KB

README.md

File metadata and controls

157 lines (101 loc) · 5.97 KB
Raw

Network Game — Boots & Cats

Ruby: RSpec Ruby: Rubocop

Game Definition

There has been a NodeJS Server and JS Clients written to allow console-based play of the Boots and Cats game, which goes as follows:

  • The game starts as soon as at least two players connected.
  • The first player is then prompted to enter a starting number.
  • Each subsequent player is then required to enter the next number in sequence, except for the cases when
    • that number is either divisible by 5 or contains the digit 5, in which case the player must say "cats"
    • or the number is divisible by 7 or contains the digit 7, in which case the player must say "boots".
    • Finally, if the number satisfies both conditions, the player must say "boots and cats".
    • We will call such words that replace a number "a mapping".
  • If the player says the wrong number, or the wrong mapping, the play instantly loses the game, and as long as there are at least two players remaining the game continues.
  • Players may join or leave at any point in time.
  • If only one player is remaining (either because others left, or lost by saying the wrong word), the remaining player wins the game.

Problem Definition

The client-side of the game play written in JS/NodeJS remains and is functional, however, the server for the game had been lost, and only the binary remains.

The task is to reverse-engineer the server and implement it in Ruby.

Examples

Here are a couple of screenshots showing of the game in action:

Manual Clients (aka cli-client)

Manual Clients

AI Clients (aka ai-client)

AI Clients

Solution

Current Ruby Server works as expected. The only unimplemented feature currently is the timeout.

Here is an example session between two AI clients:

AI Clients Playing

Makefile

There is a top-level Makefile that offers several convenient shortcuts.

To see available commands, run:

❯ make
build         Install Ruby and JS dependencies and run the all the tests and linters
check-port    Verifies that nobody is already attached to the server's port
foreman-ai    Use Foreman to start a server + several AI clients
help          Prints help message auto-generated from the comments.
iterm-ai      Use AppleScript & iTerm2 to start a server + several AI clients
iterm-cli     Use AppleScript & iTerm2 to start a server + several CLI clients
js-install    Setup the JS clients
ruby-all      Installs Ruby Gems, runs RSpec and Rubocop.
ruby-install  Setup the Ruby Server
ruby-lint     Run ruby checks such as rubocop
ruby-lint-fix Auto-fix with Rubocop all fixable errors and regenerate auto-config-file
ruby-rspec    Runs the Ruby RSpec tests for the server

Running the Server and Clients

You can test the functionality in several ways, which all result in running ruby server and several connected clients.

But first, a note about server-side logging:

Logging Level on the Server

The server process uses Ruby Logger class to log messages to STDOUT. To emulate the silent NodeJS server, set the LOG_LEVEL=error before running the ruby server, and will be completely quiet.

For example:

# By directly invoking the Ruby Server
LOG_LEVEL=error cd boots-cats-ruby && ./bin/server

# Or by using Foreman from the root of the project
LOG_LEVEL=error make foreman-ai

1. Running All Commands Manually

You need to run the following command first, to ensure all Ruby and JS libraries are setup:

make build

Then open 3 windows:

  1. In the first window, run cd boots-cats-ruby && LOG_LEVEL=debug ./bin/server
  2. In the second window, run cd boots-cats-js && ./bin/cli-client
  3. In the second window, run cd boots-cats-js && ./bin/cli-client

And so on.

To run AI clients, replace cli-client with ai-client.

The following server log levels produce various output:

# No logging
LOG_LEVEL=error ./bin/server

# Minimal logging
LOG_LEVEL=info  ./bin/server

# Debug logging, includes network packages sent to all clients
LOG_LEVEL=debug ./bin/server

2. Automatic: Using Foreman for AI Clients

This method works on all operating systems and uses the Procfile.ai to start the server and the AI clients.

However, when using Foreman, all of the processes will appear in a single window, making it difficult to debug.

3. Automatic: Using iTerm2 and ApppleScript to run either AI or CLI Clients in iTerm2 panels.

If you are an iTerm2 user, this should be the most flexible and easy to use method.

Because Foreman does not make it easy for a sub-process to receive input from STDIN, we do not offer a way to start CLI clients via Foreman.

Instead, we offer you to use AppleScript to drive iTerm2: to open a new window, split it into panels, and run a specific command in each panel.

If you have iTerm2 on OS-X installed, you should be able to use the commands:

# Start a new iTerm2 window, with multiple panels dedicated to the server
# and the AI clients.
❯ make iterm-ai

# OR

# Start a new iTerm2 window, with multiple panels dedicated to the server
# and the CLI clients.
❯ make iterm-cli

This should start a new iTerm2 window and split it into panes where the bottom pane is the server, and the middle-top and right-top are the clients. The left-most top panel is the open shell prompt.

Copyright

  • The Problem Definition: © 2024 Academia.edu, All Rights Reserved.
  • The Solution © 2024 Konstantin Gredeskoul, All Rights Reserved.