using class ? c++ & sfml and 15 characters in the title - c++

Question

When using class ?
This question is probably evident for you but i know and learned how to create class and how it's working but i don't know when using it.
I think i can cut my program with class but i don't know when i must do it.
I would like to share my code but i can't paste it completely in the topic post.
Include
#include <SFML/Graphics.hpp>
#include <string>
#include <iostream>
#include <cstdlib>
#include <cmath>
Constantes
//Constantes ecran
int tailleEcranX = 1280;
int tailleEcranY = 720;
//Constantes perso
int scalePerso = 3;
int tailleSpriteX = 32;
int tailleSpriteY = 48;
int speed(4);
int speedSprinte(20);
int milieuSpriteX = (tailleSpriteX/2)*scalePerso;
int milieuSpriteY = (tailleSpriteY/2)*scalePerso;
int pv = 100;
unsigned int pvMax = 100;
Initialisation
//Initiation des dessins
sf::RenderWindow window;
sf::RectangleShape rect;
sf::Texture perso;
sf::Sprite sprite_perso;
sf::View view;
sf::RectangleShape rectCol;
sf::RectangleShape pvBar;
sf::RectangleShape pvMaxBar;
enum Dir{Down,Left,Right,Up};
sf::Vector2i anim (1,Down);
#include "gestion_clavier.h"
Main
int main()
{
//{ Positionnement des objets
window.create(sf::VideoMode(tailleEcranX , tailleEcranY), "The Game I");
window.setPosition(sf::Vector2i(500,250));
window.setFramerateLimit(60);
rect.setFillColor(sf::Color(255,0,0));
rect.setSize(sf::Vector2f(tailleEcranX-10,tailleEcranY-10));
rect.setPosition(5,5);
rect.setOutlineColor(sf::Color(255,255,255));
rect.setOutlineThickness(3);
rectCol.setFillColor(sf::Color(0,0,200));
rectCol.setSize(sf::Vector2f(50,50));
rectCol.setPosition(400,500);
rectCol.setOutlineColor(sf::Color(255,255,255));
rectCol.setOutlineThickness(1);
sf::Clock time;
//}
//{Chargement des Sprites
if (!perso.loadFromFile("link/srpite.png",sf::IntRect(0,0,96,192)))
{
std::cout<<"erreur chargement player image"<<std::endl;
}
sprite_perso.setTexture(perso);
sprite_perso.setPosition(tailleEcranX/2-milieuSpriteX,tailleEcranY/2-milieuSpriteY);
//}
//{ Game Loop
while (window.isOpen())
{
sf::Event event;
while (window.pollEvent(event))
{
if (event.type == sf::Event::Closed)
window.close();
}
ProcessInput();
//gestion_clavier();
if(sf::Keyboard::isKeyPressed(sf::Keyboard::Z)||sf::Keyboard::isKeyPressed(sf::Keyboard::S)||sf::Keyboard::isKeyPressed(sf::Keyboard::D)||sf::Keyboard::isKeyPressed(sf::Keyboard::Q))
{
if (time.getElapsedTime().asMilliseconds()>= 50)
{
anim.x++;
if(anim.x*tailleSpriteX >= perso.getSize().x)
anim.x=0;
time.restart();
}
}
sprite_perso.setTextureRect(sf::IntRect(anim.x*tailleSpriteX,anim.y*tailleSpriteY,tailleSpriteX,tailleSpriteY));
sprite_perso.setScale(scalePerso,scalePerso);
pvBar.setFillColor(sf::Color(20,255,30));
pvBar.setSize(sf::Vector2f(4*pv,10));
pvBar.setPosition(20,20);
pvMaxBar.setFillColor(sf::Color(0,0,0));
pvMaxBar.setSize(sf::Vector2f(4*pvMax,10));
pvMaxBar.setPosition(20,20);
pvMaxBar.setOutlineColor(sf::Color(255,255,255));
pvMaxBar.setOutlineThickness(2);
if(pv>=pvMax)
{
pv=pvMax;
}
if(pv<=0)
{
pv=0;
}
if(
(std::abs((sprite_perso.getPosition().x+milieuSpriteX)-(rectCol.getPosition().x+50/2))<50)
&&
(std::abs((sprite_perso.getPosition().y+milieuSpriteY)-(rectCol.getPosition().y+50/2))<50)
)
{
std::cout<<"collision"<<std::endl;
pv--;
std::cout<<pv<<std::endl;
}
//Dessinage
window.draw(rect);
window.draw(rectCol);
window.draw(sprite_perso);
window.draw(pvMaxBar);
window.draw(pvBar);
window.display();
window.clear();
}
//}
return 0;
}
//}

Answer 1

I struggled to understand a lot of your code, since the variables seem to be written in french, but I can give you some general advice.
When deciding when to use a class, you should first think about the contents of your program in terms of objects. For example: if there is a character/person, or something you would consider an independent object in the real world, there's a reasonable chance that should be its own class. The class should contain whatever is needed to describe the object, like it's position, scale, sprite, etc. Here's an example:
class Person
{
private:
sf::Vector2f position;
sf::Sprite sprite;
public:
void setPosition(sf::Vector2f newPosition);
sf::Vector2f getPosition();
void draw(sf::RenderWindow& window);
}
If you want your class to have position and such, I recommend deriving from the sf::Transformable class to save you some time and effort. You can also derive from sf::Drawable if you want to draw your object using window.draw(person) like many of the default SFML objects can.
class Person : public sf::Transformable, public sf::Drawable
{
private:
sf::Sprite sprite;
void draw(sf::RenderTarget& target, sf::RenderStates states) const
{
states.transform *= getTransform();
target.draw(sprite, states);
}
}
You have a number of variables you say are constants. If that is the case, then it's considered good practice to mark them as such using const. For example:
const int scalePerso = 3;
This stops you from changing the value in the future and causing bugs you need to track down.
I also noticed you are defining seperate variables for X and Y positions. This is unnecessary, since SFML defines the sf::Vector2 type. For example:
const sf::Vector2f tailleSpritePosition(32.f, 48.f);
instead of
int tailleSpriteX = 32;
int tailleSpriteY = 48;
You're also mixing types quite a bit by passing integers into functions that accept float values. This is also considered bad practice. If you want to change an integer into a float, you should explicitly cast it using (float) like so:
rect.setSize(sf::Vector2f((float)tailleEcranX - 10.f, (float)tailleEcranY - 10.f));
making sure to add the letter f to the end of your literals to tell the compiler those are floating point values too. Alternatively, you could just define your variables to be of type float to start off with in some cases.
Feel free to ask if there is anything you don't understand.