kex-io

Async asset handling for Kha. The code is heavily based on iron's Data class, but uses promises instead of callbacks. It also features a simple scoping (i.e. reference counting) mechanism. All assets are cached, whether already loaded or still in progress.

dependencies

• tink_core

optional dependencies

• tink_core_ext for Promises.multi()
• tink_await if you prefer to use javascript style @:await instead of manually handling the promises

usage

using tink.CoreApi;
var blobs = new kex.io.BlobIO();

load a single file

blobs.get('a.txt').handle(function( o ) switch o {
	case Success(blob):
		trace(blob.toString());
	case Failure(err):
});

load 2 files

(blobs.get('foo.txt') && blobs.get('bar.txt')).handle(function( o ) switch o {
	case Success(data):
		trace(data.a.toString()); // foo
		trace(data.b.toString()); // bar
	case Failure(err);
});

load multiple files

Promise.inParallel([
	blobs.get('a.txt'),
	blobs.get('b.txt'),
	blobs.get('c.txt'),
]).handle(function( o ) switch o {
	case Success(datas):
		for (data in datas) {
			trace(data.toString());
		}
	case Failure(err);
});

load some json

typedef Foo = {
	final someInt: Int;
	final someString: String;
}

blobs.get('foo.json').next(function( b ) {
	var foo: Foo = tink.Json.parse(b.toString());
	return foo;
}).handle(function( o ) switch o {
	case Success(foo):
		trace(foo.someInt);
		trace(foo.someString);
	case Failure(err):
});

scopes

Promise.inParallel([
	blobs.get('a.txt', { scope: 'scope-a' }),
	blobs.get('b.txt', { scope: 'scope-a' }),
	blobs.get('b.txt', { scope: 'scope-b' }), // uses cached b.txt
	blobs.get('c.txt', { scope: 'scope-c' }),
]).handle(function( o ) switch o {
	case Success(datas):
		blobs.unloadScope('scope-a');
		// will unload `a.txt`
		// keeps `b.txt` around, as it is also required in `scope-b`
	case Failure(err);
});

named fields (using tink_core_ext's Promises)

Promises.multi({
	hello: blobs.get('a.txt'),
	world: blobs.get('b.txt'),
]).handle(function( o ) switch o {
	case Success(datas):
		trace(datas.hello);
		trace(datas.world);
	case Failure(err);
});

logging

Use the initialization macro --macro kex.io.AssetLog.level(X) to enable more verbose logging (default is errors). If you want to use some custom logging mechanism, define kex_io_custom_log to disable the provided logger, add a kex.io.AssetLog class somewhere in your classpath and define the 4 required macros (asset_debug, asset_info, asset_warn, asset_err).

X	name
0	no logging code will be generated
1	errors
2	warnings
3	info
4	debug

scoping

In case you don't supply a scope for your assets, the default * is used.