Remove hard dependency on Browser from Asyncify

[curiousdannii]

Split out from #12165

Remove the hard dependency on Browser from Asyncify. Could save 13KB from a minified build.

[curiousdannii]

I'm not sure if something similar should be done for FS. It doesn't actually declare a dependency, so I guess if Browser doesn't exist for any other reason then there will just be some hanging references, so you better not call createPreloadedFile?

emscripten/src/library_fs.js

Lines 1900 to 1930 in 5fb249c

	createPreloadedFile: function(parent, name, url, canRead, canWrite, onload, onerror, dontCreateFile, canOwn, preFinish) {
	Browser.init(); // XXX perhaps this method should move onto Browser?
	// TODO we should allow people to just pass in a complete filename instead
	// of parent and name being that we just join them anyways
	var fullname = name ? PATH_FS.resolve(PATH.join2(parent, name)) : parent;
	var dep = getUniqueRunDependency('cp ' + fullname); // might have several active requests for the same fullname
	function processData(byteArray) {
	function finish(byteArray) {
	if (preFinish) preFinish();
	if (!dontCreateFile) {
	FS.createDataFile(parent, name, byteArray, canRead, canWrite, canOwn);
	}
	if (onload) onload();
	removeRunDependency(dep);
	}
	var handled = false;
	Module['preloadPlugins'].forEach(function(plugin) {
	if (handled) return;
	if (plugin['canHandle'](fullname)) {
	plugin['handle'](byteArray, fullname, finish, function() {
	if (onerror) onerror();
	removeRunDependency(dep);
	});
	handled = true;
	}
	});
	if (!handled) finish(byteArray);
	}
	addRunDependency(dep);
	if (typeof url == 'string') {
	Browser.asyncLoad(url, function(byteArray) {

Edit: actually I think that might be happening in the test_wget test that failed. emscripten_wget doesn't directly use Browser, so I didn't add it to its dependencies list. But it does call FS.createPreloadedFile, which uses Browser as I just showed.

I'm not sure what the best way to organise these dependencies is, probably adding Browser to the FS dependencies would be most consistent at present? However if we do that then I'll want to fix it somehow in the future, because it's a lot of needless code for anyone like me who uses FS but doesn't want or need Browser ;). Or maybe just add the dependency to emscripten_wget?

Edit: Okay, I've added a Browser dependency to emscripten_wget, not perfect, but good enough for now IMO.

[kripken]

Nice!

I think this is a good first step.

The planned work to split out the main loop handling could be even better. We could in that case emit different code when emscripten_set_main_loop (the function that shows the main loop is actually used) is called by the user or not, maybe.