Python-Seed

Environment

This is the environment (language, IDE, tools) recommended for development

• Python 3.8
• Visual Studio Code, Recommended extensions: Python, Docker, LiveShare
• Git
• Docker
  • Install Docker Desktop on Windows Home

Setup

Download code repository

1. Make sure you have Git installed on your machine
2. In Github, click on the green button label "Code" in top right of the repository
3. Copy the HTTPS url
4. Open a terminal in the folder where you want the your code repository to be
5. Run: git clone [HTTPS url]

If you need more help click Here

Install required python packages

1. Make sure to have Python 3.8 installed on your machine
2. Clone code repository (see section above to know how)
3. Open a terminal in the root of the project
4. Run: pip install -r requirements.txt

First Step

After you are all set up, you are finally ready to code! ;)

But where to start? In the GitHub repository you just cloned, you will find a file named Src/brain.py. In this file, you have the function on_next_move(turn_info) and it is in this function that you have to code your AI. It takes as parameter a turn_info which contains all the information given by the GameSever so that your AI can make a decision on its next move. You can find this model in Src/Models/turnInformation.py. The function must return a direction (up, down, left, right). If it fails to do this 3 times in a row, the GameServer will consider your AI to be obsolete. You can find these constants in Src/Models/direction.py (If you change these values, your next moves will be considered invalid). Another function found in brain.py is on_finalized(turn_info). Once a game is finished, this method is triggered with the final state of the map and nothing needs to be returned to the GameServer.

The map you receive is encoded in a list of strings. Some of the encodings are:

• “P0*” : represente the head of the player with the id 0
• “P0” : represente part of the body of the player with id 0
• “p0” : represente part of the neck of the player with id 0
• “W” : A wall

Now you know where to start coding. As a first goal, try to get your AI from point A to point B as simply as possible.

Good luck!

Local game simulation

Download the local GameServer

1. Make sure you have Docker installed on your machine and running
2. Login into docker:
   • With Docker Hub: docker login
   • With Github: docker login https://docker.pkg.github.com
   • You can login using your password or a personal access tokens
   • If using a personal access tokens:
     1. Create your personal access tokens Here
     2. Make sure you give it the following scope: read:packages
     3. Login using the command: docker login -u [GITHUB_USERNAME] -p [TOKEN_ID]
3. Pull Docker images:
   • Windows or MacOs: docker pull docker.pkg.github.com/readyaicode-2020/gslocal/gslocal:win_mac
   • Linux: docker pull docker.pkg.github.com/readyaicode-2020/gslocal/gslocal:linux
4. Run the following command to know if the images was correctly loaded:
   • docker images
   • You should see a repository with the name of your images you just pull

• To know how to clean up your docker environment at the end of the competition, see the last section of this ReadMe

Execution

1. Run code

   1. Open a terminal in the root of the project
   2. Make sure the required packages are installed (see section Install required python packages)
   3. Run: python main.py
      • If it is not starting a game, sometime you need to restart the GameServer Local container
   4. You can stop the execution by doing ctrl-c in the terminal

2. Run the local GameServer in Docker container

   1. Make sure docker image in install (see section Setup GameServer Local)
   2. In a terminal run:
      • Windows or MacOs: docker run --rm -d -p 5001:5001 --name gs_container docker.pkg.github.com/readyaicode-2020/gslocal/gslocal:win_mac
      • Linux: docker run --rm -d --network host -p 5001:5001 --name gs_container docker.pkg.github.com/readyaicode-2020/gslocal/gslocal:linux
      • --rm: Remove the container when stop
      • -d: Hide the outputs, remove if you want see them
   3. You can verify if the container is running with: docker ps

3. Connect the Dashboard UI to the GameServer Local

   1. Make sure the GameServer Local container is running
   2. Go to the section Local debug in the event dashboard
   3. Click on the button Start
      • Attention: If you restart the GameServer Local container you will need to repeat this step to reconnect the UI

4. In case its not working, here’s a few things to try:

   • You can restart the container with: docker restart gs_container
   • You can stop the container with: docker stop gs_container and then repeat the steps from Run the local GameServer in Docker container

Deploy code

1. If you installed and are using new packages, in requirements.txt add the packages names and its version. Otherwise your code will not work in deployment

   • [PACKAGE NAME]==[PACKAGER VERSION]
   • Attention Do not use: pip freeze > requirements.txt

2. Push the code on the Master branch

   1. Open a terminal in the root of the project
   2. Stage all changes: git add .
      • You can also stage on file at a time: git add [file_path]
   3. Commit changes: git commit -m "Descriptive message"
   4. Push code: git push

3. See the status if your deployment

   1. Go to the Actions tab in the Github repository
   2. See on the latest workflow
      • If it is on going it will have a yellow spinning wheel next to it
      • If it is done successfully it will have a green circle next to it
      • If it fails it will have a red cross next to it. Click on the workflow to see the logs and find out why it failed to deploy

4. Wait 2 to 4 minutes after the workflow is done successfully to make sure the deployment is fully completed on are servers

!ATTENTION!: Only push on the Master branch when it's really necessary to avoid problems with the deployment pipeline. Problems may last a while and slow down your developpement process.

Clean up

Clean Docker environment

1. Stop and remove container
   1. docker stop gs_container
   2. If you remove --rm earlier: docker rm gs_container
2. Remove images
   • docker rmi [images_name]
     • Use docker images to see the list of images and find its name or id
3. Logout of docker
   • docker logout