Making a C++ map editor (need help)

[Netwatcher]

Well, I've been wanting to make a map editor in C++ for a long time now, (let's ignore the fact that I'm using DirectX for a second)
and I just can't understand how to split the map to 64x64 or any kind of "blocks".
If you can send me some code or tutorial I can look at it would be very appreciated (I don't really care what language, just want to get a hold of the concept)

P.S.
I don't have problem blitting tiles, but when it comes to the user placing them in a "tile" order like in the ES map editor I just get lost.

will post my progress here.

Modulus FTW%!

Thank you in advance.

[Netwatcher]

Instead of letting the user browse files, I will use an algorithm to bring all image files from a folder, you can then switch tiles (added later) using the directional keys (left and right) on the keyboard or some other method I haven't quite decided yet. (move back and forth in the vector, or should I make this a list? don't have anything done so I chose the option that fits anything, i.e. vector)

This is what I got so far

Code: Select all

//Test version(debug) to be moved to ListImg.cpp
#include <windows.h>
#include <iostream> 
#include <vector>


using namespace std;

char* locjpg = "C:\\UNIMPICS\\*.jpg"; //any file; *.jpg- any filename with .jpg extension
char* locbmp = "C:\\UNIMPICS\\*.bmp"; //any file; *.bmp- any filename with .bmp extension
char* locpos = "C:\\UNIMPICS\\";
char* nfilen;

unsigned f=1;


int main()
{
	
	char desOfile[80];
	vector<string> victor;
	
	

	
       WIN32_FIND_DATA findFileData;
	HANDLE hFind = FindFirstFile((LPCSTR)locjpg, &findFileData); 
	cout <<"Picture files in dir:" <<"\n"<<locjpg<<"\n"<<locbmp<<"\n\n"<<endl;
	cout <<findFileData.cFileName<<"\n"<<endl; 
	nfilen = findFileData.cFileName;
	strcpy(desOfile,"C:\\UNIMPICS\\");
	strcat(desOfile,nfilen);
        victor.push_back(desOfile);
	
	
	
	for(f=1/*FindFirstFile is the first file*/;FindNextFile(hFind,&findFileData);++f/*count file for w/e use*/)
	{
		
		strcpy(desOfile,"C:\\UNIMPICS\\");
		strcat(desOfile,nfilen);                  
		cout<<desOfile<<"\n\n"<<endl;    //I had to go and wanted to flush...
		victor.push_back(desOfile);
	}



        for(unsigned v=0;v<f;v++)
	{
		cout<<victor[v].c_str()<<"\n";
	}





    system("PAUSE");
    return 0;
}

The purpose of this algorithm is to get the full path to all the image files in the specific folder, then link them to a LPDIRECT3DTEXTURE(again... DirectX) the user controls

You can also use this algorithm for tons of other stuff, tell me what you think.

[dandymcgee]

Basically the map is just an array (whether it be 1D or 2D) of tiles. Tiles can be as simple as integers, or they can actually be their own class with all sorts of other attributes.

Drawing the map to the screen involved iterating through that array, or vector if you wish, and adding tile width to the drawing position each time until you reach the end of the row where you would add tile height to the drawing position (ie. 0,0 ; 32,0 ; 64, 0 ; 96,0 for 32x32 tiles).

Saving the map can also be done by simple writing integers out to a text file, and reading them back into an array to load the map. More complex methods would have to be created if you plan to store more information than just the image of that particular tile.

Did this help at all?

[Netwatcher]

Basically the map is just an array (whether it be 1D or 2D) of tiles. Tiles can be as simple as integers, or they can actually be their own class with all sorts of other attributes.

Drawing the map to the screen involved iterating through that array, or vector if you wish, and adding tile width to the drawing position each time until you reach the end of the row where you would add tile height to the drawing position (ie. 0,0 ; 32,0 ; 64, 0 ; 96,0 for 32x32 tiles).

Saving the map can also be done by simple writing integers out to a text file, and reading them back into an array to load the map. More complex methods would have to be created if you plan to store more information than just the image of that particular tile.

Did this help at all?

I know how to draw tiles on a 2D plane(modulus ftw) and the mapping method (why would I use vector where I can use an array?)
I use a multi-dimensional array for multiple levels or for special attributes(non-passable,Z-order etc..)


My problem is with the UI
I don't see how I can let the user replace tiles visually;clicking on the tile and changing it to a different one, or how can I detect when the mouse is over a certain tile.

Do you say I need to first make an integer-based tile-editor and then bind the integers to the bitmaps/visual tiles?

[Bakkon]

My problem is with the UI
I don't see how I can let the user replace tiles visually;clicking on the tile and changing it to a different one, or how can I detect when the mouse is over a certain tile.

Basically use point inside rectangle collision detection.

[Netwatcher]

My problem is with the UI
I don't see how I can let the user replace tiles visually;clicking on the tile and changing it to a different one, or how can I detect when the mouse is over a certain tile.

Basically use point inside rectangle collision detection.

Wow, the simplest solution and I didn't even think about it >,<.
The more tiles I get the more rectangles I need to make...
Is there a way to fasten this process of creating rects for each tile?

[thejahooli]

I just use this to determine the x of the tile which the mouse is inside

Code: Select all

((mousex + camera.x) - ((mousex + camera.x) % OBJECTWIDTH)) / OBJECTWIDTH;

and the same for the y.

[Netwatcher]

Code: Select all

((mousex + camera.x) - ((mousex + camera.x) % OBJECTWIDTH)) / OBJECTWIDTH;

So, I've
created a sprite "mouse" object who's position is determined by mickies (if I spelled it right ) so it's easier to locate the mouse and check collision like that(since i've already created my bounding box) ;D

What is the camera object for??? i didn't say i was scrolling :O


#IF nothing_here_makes_sense
#define ME tired
#endif

[dandymcgee]

Code: Select all

mapArray[ layer ][ mouse_x/tile_width ][ mouse_y/tile_height ];

Or something like that..

[thejahooli]

What is the camera object for??? i didn't say i was scrolling :O

sorry. That code is just what is used in mine. It should work exactly the same without the camera.x

Code: Select all

mapArray[ layer ][ mouse_x/tile_width ][ mouse_y/tile_height ];

Or something like that..

That's pretty much mine but with yours it would round it to the nearest integer, this means that even if your mouse was to the right or bottom of the tile's center it would place the tile in the tile to the right or below.

[dandymcgee]

That's pretty much mine but with yours it would round it to the nearest integer, this means that even if your mouse was to the right or bottom of the tile's center it would place the tile in the tile to the right or below.

Fix'd

Code: Select all

mapArray[ layer ][ (mouse_x - (mouse_x % tile_width)) / tile_width ][ (mouse_y - (mouse_y % tile_height)) / tile_height ];

What do ya know, it's EXACTLY the same as your code.

[Netwatcher]

That's pretty much mine but with yours it would round it to the nearest integer, this means that even if your mouse was to the right or bottom of the tile's center it would place the tile in the tile to the right or below.

Fix'd

Code: Select all

mapArray[ layer ][ (mouse_x - (mouse_x % tile_width)) / tile_width ][ (mouse_y - (mouse_y % tile_height)) / tile_height ];

What do ya know, it's EXACTLY the same as your code.

Looks like the code I use to blit tiles O_o just with mouse input.(well not completely like it)


Thank you guys.

[RyanPridgeon]

Take the position of the mouse ( add and subtract here if you want to use a camera object)

Then divide that by the size of the tile to get a number representing the position of the mouse in tile widths

Then floor this number to get the tile that the user is on.

(If you're using a 1d array, the position in the array will be the x tile position, add the y tile position multiplied by how many tiles wide the screen is)

I know others have said this, but just said it in idiot proof steps incase the problem isnt fixed :P

To the OP; i think you need to put [Solved] in the title if your problem is solved

[Netwatcher]

Take the position of the mouse ( add and subtract here if you want to use a camera object)

Then divide that by the size of the tile to get a number representing the position of the mouse in tile widths

Then floor this number to get the tile that the user is on.

(If you're using a 1d array, the position in the array will be the x tile position, add the y tile position multiplied by how many tiles wide the screen is)

I know others have said this, but just said it in idiot proof steps incase the problem isnt fixed :P

To the OP; i think you need to put [Solved] in the title if your problem is solved

I need to take the position of the mouse position to the window right? (If i'm using abs mouse pos)
Won't it only work in full-screen mode?


Think I just might learn SDL instead of fighting with DirectX for the map editor co's mr DiX is pissing me offafa...

[MarauderIIC]

Basically use point inside rectangle collision detection.

This isn't necessary. Simple division will do the trick. Much easier on memory, too.

That's pretty much mine but with yours it would round it to the nearest integer, this means that even if your mouse was to the right or bottom of the tile's center it would place the tile in the tile to the right or below.

No. C doesn't round. It truncates.

I need to take the position of the mouse position to the window right? (If i'm using abs mouse pos)

Take the mouse coordinates relative to the control. Which really just means, take the mouse coordinates. It doesn't matter whether or not the screen is maximized. If you want to "snap" the mouse to a position, for instance, (0, 0) - (31, 31) is one tile, which means TILE_HEIGHT = 32 and TILE_WIDTH = 32, then you can do:

myMapsTiles[mouse.y / TILE_HEIGHT][mouse.x / TILE_WIDTH]

Thanks to the wonders of integer math, this works just fine. No need to floor() anything.

mouse.y / TILE_HEIGHT
0/32 = 0.
1/32 = 0.0322580645 -> which is truncated to 0
31/32 = 0.98675 -> 0

32/32 = 1, which means you're now modifying the next tile down.

To be able to select a tile, you do the same thing, only in your tile selection tool. If all your tiles are output in a straight line, you do
selectedTile = mouse.x / TILE_WIDTH;
And then when they click on the map, you do
myMapsTiles[mouse.y / TILE_HEIGHT][mouse.x / TILE_WIDTH] = selectedTile;

Keep in mind the mouse coordinates I've used are relative to the control that has been clicked. If you have application-wide mouse coordinates, you can do this:
selectedTile = (mouse.x - myTool.x) / TILE_WIDTH;
myMapsTiles[(mouse.y - mapViewport.y) / TILE_HEIGHT][(mouse.x - mapViewport.x) / TILE_WIDTH];