tank-wars-iit 1.0.2

Tank Wars framework for Scarlet HackNiite

Robot Wars

Engine for the Robot Wars hack night.

Requirements

• Python 3.10
• pygame 2.3.0 (SDL 2.24.2, Python 3.10.2)

How do I run it?

Run init.py in Python. (This is a testing script and will be changed/moved in the future.)

Structure

• engine - Package containing full Robot Wars engine.
  • entity - Package containing all Entity classes supported by the engine.
    • bullet.py - Contains the Bullet class (inherits Entity); represents a bullet that moves in the direction it faces, bounces off walls, and collides with Robots.
    • entity.py - Contains the Entity abstract class, which represents a geometric object in an Arena.
    • robot.py - Contains the Robot class (inherits Entity); represents a robot that can move, turn, turn its turret, and shoot Bullets.
    • wall.py - Contains the Wall class (inherits Entity); represents a static, unmovable barrier.
  • maps - Directory containing map JSON files which can be used to easily instantiate an Arena with Walls and spawn locations.
    • basic.json - A small, wide map with five Walls and two spawns.
    • circles.json - A medium, square map with two concentric circles of Walls and four spawns.
    • stress_test.json - A large, square map with no Walls and 64 spawns, meant for stress testing the engine.
  • util - Package containing engine utility modules.
    • draw.py - Contains helper functions for drawing gradients on screen.
    • geometry.py - Contains helper functions for collision detection between two convex polygons.
  • arena.py - Contains the Arena class, which represents a battle arena for one or more Robots.
  • map.py - Contains type definitions and guards for the map JSON format.
  • quadtree.py - Contains the Quadtree class, which is a data structure used to optimize collision filtering.
• tools - Useful scripts to automate workflows.
  • generate_circles.py - Generates the engine/maps/circles.json map.
  • generate_stress_test.py - Generates the engine/maps/stress_test.json map.
• init.py - Loads an Arena from a map, spawns a user-controlled Robot and several NPC Robots, and runs the simulation with visuals. Used for testing.

To Do

• Map Creation
  • [IMPLEMENTED] 1v1 Map - For now, we can use engine/maps/circles.json.
  • Battle Royale Map - Extra-large map with adequate Walls spread across the map.
  • Zoom-In - Useful for observing the battle royale map!
• Optimization
  • [IMPLEMENTED] Quadtree Collision Filtering - Use a Quadtree data structure to filter which collisions to check for as opposed to checking every pair of Entitys.
  • Parallelize Entity.update - Invoke all calls of Entity.update in parallel since they operate mostly independently (exception is modification of shared Arena state, like spawning Bullets via Robot.shoot).
  • Parallelize Entity.handle_collision - Invoke all calls of Entity.handle_collision in parallel; modify the method to return new position and rotation properties that will be applied in parallel.
• Game Rules
  • [IMPLEMENTED] Robot Destruction - Incorporate health in Robot and damage in Bullet.
  • Win Condition - Destroy the other Robot.
  • Stalemate Condition - Both Robots are still alive after n seconds of simulation.
  • [IMPLEMENTED] Coins - Place Coin entities around the Arena; whichever Robot collects the most wins in case of stalemate.
• Robot Programmability
  • [IMPLEMENTED] Robot Manipulation - Expose API to move and turn the Robot and shoot Bullets.
  • [IMPLEMENTED] Enemy Robot Detection - Expose API to find the nearest Robot to a given Robot.
  • [IMPLEMENTED] Pathfinding - Expose API for Robots to navigate their way to a location around Walls.
  • RobotControl Class - Incorporate all programmable/personalizable Robot features into one class which each team extends with their implementation. For now, it's semi-implemented as the Robot.on_update callback, which is unideal because it doesn't secure the engine from unwanted modifications.
  • Engine Encapsulation - Prevent raw access to any Entitys or the Arena; the only accessible component is the RobotControl class.
• Robot Personalization
  • [IMPLEMENTED] Color - Set Robot colors via properties color and head_color.
  • Size - Set Robot size; larger size increases health and decreases speed.
• Robot Examples
• Unit Tests
• Prepare for Distribution