So far I've come up with 3 different ways that I could go about storing the maps in memory:
- A 3-Dimensional vector
A 1-Dimensional vector with an algorithm that finds the appropriate index for any particular tile
A 1-Dimensional vector storing a Layer class object for each layer, each of which contains a 1D (with index algorithm) vector storing tile data.
Code: Select all
//declaration
vector<vector<vector<Tile*>>> mapData;
//accessing contents
mapData[layer][x][y]->getTileType();
Code: Select all
//declaration
vector<Tile*> mapData;
//accessing contents
int tempTileNumber = GetTileNumber(layer, x, y);
mapData[tempTileNumber]->getTileType();
//Index finding function
int GetTileNumber(int layer, int x, int y)
{
int tempa = x / tileWidth;
int tempb = y / tileHeight;
//Note: first layer is 0
int tilenum = layer * tilesPerLayer + tempa + tempb * tilesPerRow;
return tilenum;
}
Level.cpp
Code: Select all
//declaration
vector<Layer*> layerData;
//accessing contents
mapData[layer]->layerGetTileType(x, y);
Code: Select all
//declaration
vector<Tile*> tileData;
//Index finding function
int GetTileNumber(int layer, int x, int y)
{
int tempa = x / tileWidth;
int tempb = y / tileHeight;
//Note: first layer is 0
int tilenum = layer * tilesPerLayer + tempa + tempb * tilesPerRow;
return tilenum;
}
//returns tile type
int layerGetTileType(int x, int y)
{
int tempTileNumber = GetTileNumber( x, y);
return tileData[tempTileNumber]->getTileType();
}
Any opinions regarding ease of use later in the project, performance hits, etc. are much appreciated.
). Think about it, every time that you are looping through your vectors to render, or even fetch an index from an array, you are doing a calculation rather than just indexing straight up.
