feat: LoadTMXFromMemory

[cortexmancer]

Hello. I took the liberty to implement this TODO and notice some things.

Looking at baylej/tmx#58 is kinda hard to understand because it seens like baylej forced pushed on your commit, instead of creating a new one.

But I think you intended to tmx_load_buffer_path to acts like tmx_rcmgr_load_buffer_vpath is acting but without the definition of an resource manager. Due to this, it is impossible for the current function tmx_load_buffer_path to find files in a path of a given fileName.

The tests shows that to use this method correctly as it is, without defining the resource manager, the user must control the current directory that is loading. I don't think that is a great deal anyway.

Let me know what you think?

Was that your idea?
Is this implementation okay?
Do you think we should handle the resource_manager somehow?

• Implementation
• Test
• Example?
• Update README

closes #4 ?

[RobLoach]

This is great, thanks for taking it on. Glad baylej brought in the change. While I haven't looked deep into how it's set up, I think we'd need to figure out the virtual path stuffs within tmx.

Loading a map from a buffer is helpful if you're loading the map from a .zip file, or another virtual file system. Like through https://github.com/RobLoach/raylib-physfs .

#define RAYLIB_PHYSFS_IMPLEMENTATION
#include "raylib-physfs.h"

int main() {
    // Initiatize the file system.
    InitPhysFS();
    MountPhysFS("resources.zip", "res");

    // Load an image through PhysFS directly from resources.zip.
    tmx_map* map = LoadTMXFromPhysFS("res/map.tmx");

    // TODO: Since the "res" basepath isn't sent through to
    // LoadTMXFromPhysFS(), how does the .tmx know where to load its images?

    // Close the file system.
    ClosePhysFS();
}

But again, the buffer loading is only helpful if you're using something other than raylib's stdio file loading.

[cortexmancer]

Loading a map from a buffer is helpful if you're loading the map from a .zip file, or another virtual file system. Like through https://github.com/RobLoach/raylib-physfs .

I was asking myself why this was helpful! Nice input.

So some points:

1. It would be useful to pack everything in a zip file, for better maintaining in a real scenario where resources/assets are too much for just git it?
2. Would physfs be the concrete implementation of the resource manager? I need to take a look at how this manager works. So far I've implemented things as they appeared by my needs and I didn't have a need or reference point to check up the resource manager feature.
3. This PR surely then will be far more complex, and this current implementation then don´t really have a useful meaning to a user as it is then.