Skip to content
A simple framework for managing entitys, components, and live-at-runtime scripts in haxeflixel
Haxe Shell
Branch: dev
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


IceEntity Demo

A simple framework for managing entities and components in HaxeFlixel

[ATTENTION ANY/ALL USERS] Please move to the development branch (relies on dev flixel), as I can no longer guarantee updates to anything else, sorry.

[NEW] sample project here [NEW]

(The above may be a much simpler way to learn much of what is included in this readme, as well as some smaller, but still useful features)


[NEW v1.2.0]

Added an api for accesing variables on scripts, while retaining type-safety (Entity.getVar())

[NEW v1.1.0]

Allow use of % in entity declarations, referring to game width or height

[NEW v1.0.0]

Rewrote entire scripting system, allowing proper Haxe classes to be parsed as scripts

Added templates and instance creation to XML parser


Added syntax for exposing and requesting classes inside of scripts


Added simple FSM system (sorry no write-up yet)


Added "noreload" option for script elements

Made scripts remove after calling init, unless they have update or destroy logic

[v0.7.0] Allow access to instance via a property

[v0.6.1] Made live-reloading fully automatic

[v0.6.0] Added live-reloading of scripts

[v0.5.0] Added hscript integration


See Changelog


[1] Run haxelib git iceentity dev in a terminal with access to git

[2] Add <haxelib name="iceentity"/> to your Project.xml file, directly under <haxelib name="flixel"/>



Call add(EntityManager.instance); in the create function of your playstate.

Use EntityManager.instance.AddEntity(); to add objects or groups to the manager.

Use AddComponent(); on a class extending entity to add a new component.

init() will run on the first update cycle, useful for getting and stashing references to other entities.

Entity Parser:

IceEntity includes an xml parser, which can build entities from simple xml files. In order to use this system, you must:

[1] Create an xml file with this structure:

<?xml version="1.0" encoding="utf-8" ?>
    <entity tag="myTag" x="50%" y="100"> //"%" is new as of V1.1.0, it allows you to specify a percentage of game width or height, instead of an exact value
	    <art width="32" height="32" path="assets/images/myimage.png">
		    <animation name="idle" frames="0,1,2,3-6,6-0" framerate="10" looped="true" autorun="true"/>
			//ANIMATIONS: all parameters must be supplied, except "autorun" which mearly starts the animation as soon as it is loaded
			//FRAMES: as of v0.3.1, "frames" can now use hyphens, allowing you to do things such as: "0-10", or "10-0", note that commas are still allowed, and this is valid: "0-10,10-0,0,1,2,3,4,5,6,7,8,9,10"
	    <component type="">
		    <param name="speed" type="int" value="10"/>
		//CHECK NEXT SECTION FOR SCRIPTING INFORMATION (yes, you can write code here!!)

	<entity tag="templateTest" template="test" x="0" y="0"/> [NEW as of 1.0.0]
	/*This is a "template" it can be used to create many instances of a premade entity, 
	simply declare the entity as usual, and add an identifier: template="identifier".
	Note that by default, entities with a template attribute are not built at startup,
	however you can do so by adding instance="true" to the declaration.
	Note this is a single line for clarity, but this "template" would in practice look
	the same as the above entity declaration, with art and all*/

	<instance template="test" tag="overwrittenTag" x="10" y="15"/>
	/*This is an "instance", you can instantiate a template as many times as you wish, 
	while altering tag and position.*/


Important information:

Parameters for components must be in the same order as specified in their constructor, passing the entity's GID is not required, it will be auto set.

Param "name" attributes are not required, but are strongly recommended for organization.

Allowed types for parameters are: "int" "float" and "bool" anything else will be treated as a plain string, including a lack of the "type" attribute. Capitalization on param types does not matter.

MOST IMPORTANTLY: any component you wish to add in a xml file MUST be referenced somewhere in your basecode, even adding an import; to your playstate will work. xml parsing will not work without this, as the component will not be compiled.

[2] Call EntityManager.instance.BuildFromXML("assets/data/MyEntities.xml"); in your playstates create function (or wherever really).

HScript Integration:

As of v1.0.0, IceEntity's scripting system has been completely rewritten, with work from both me and @Ohmnivore, making the act of scripting at runtime much more user friendly. Note this is a new feature, relatively untested, and may not be as efficient as standard components. If you find an issue, or are confused, tweet to me at: @nico_m__, or email me: nico(dot)may99(at)gmail(dot)com. These features allow you, the developer, to program your game in standard Haxe code, without recompiling, at runtime. I'll let you think of the possibilities:). It also means, that with little or no work, modding can be integrated into your game! Here are the steps to getting this to work in you game:

[1] In you entity xml file (described above), there are two places you can put scripts. One being inside of an entity declaration, like so:

<?xml version="1.0" encoding="utf-8" ?>
    <entity tag="myTag" x="0" y="0">

The other is outside of an entity declaration, like so:

<?xml version="1.0" encoding="utf-8" ?>

Before I describe how to actually write a script, let me make sure you know the difference between the two places. A script placed inside an entity declaration has two features that orphan (outside) scripts do not. First: a "owner" variable is automatically created, allowing you to access the entity that the script resides in, and second: they scripts "destroy" function will automatically be called in the event of its parent entity's destruction.

[2] In order to write your script, this is the format you must use:

<script path="" id="-1"> 
<!--if a path is specified, that is loaded instead of the <text> node-->
<!--id can be left at -1, to auto-assign, but I recommend setting it to any int greater than 0, to allow referencing this script-->
	<expose name="FlxG" path="flixel.FlxG"/> <!--Gets a static class-->
	<request name="Player"/>	<!--requests a class instance from a global pool-->
		import flixel.FlxG; //these can be used instead of the above xml tags
		request Player;
		var test:Entity = new Entity("test");

		function init
			var i = 10; //variables declared in init are accessible from other functions
		function reload
			<!--described in next section-->
		function update
			trace(owner.x); <!--if this script was inside an entity declaration, you can reference that entity with "owner"-->
		function destroy
			Player = null;

As you can see, the "expose" tag allows the script to gain access to a static class, and reference it as whatever is in the "name" attribute. The "request" tag allows the script to get access to a class instance (or, truthfully, a static class will also work), from the ScriptHandlers global pool. You can add to the pool in your code with ScriptHandler.AddModule(name, value); Note that this must be done BEFORE you parse the entity file, or you will get a nasty error message.

[3] You may be (rightly) thinking: "Hey, you said we were using real Haxe, that doesn't look like real Haxe to me?" The reason for this: IceEntity's parser has been written to be flexible with how you want to write your scripts, if all you need is a few lines doing something really simple, the above is easier to type, however, if you wish to get full IDE support (auto-completion), you can program your scripts like this (or pretty much anywhere in-between the two styles):

package ;

import flixel.FlxG;
import ice.entity.Entity;

class Test
	var owner:Entity;
			This is a special syntax used in scripts.
			It allows you to create variables that are not compiled when the script is parsed.
			This way, if you have used a request tag, or wish to have completion for the scripts owner, 
			you can create a variable of the proper type, 
			allowing full completion, without fearing issues in your scripts
			You may only use one of these blocks per script, simply open with "//#" and repeat to close.
	public static var testEntity = new Entity("myTag"); /*
														These variable are live-reloaded only if they
														did not previously exist, changing a value will not
														reload the variable, however if you add a new
														variable, it will be detected.
	public function init() 
		var p:Entity = new Entity();

	public function update()
		test("hello world");

	public function reload()


	function test(t)
		This is another special syntax, any function placed inside of this block (one per script),
		acts how you would expect a standard haxe function to act, and is live-reloaded.

[4] The scripting system relies on hscript, which is basically interpreted Haxe. Unfortunately, I do not know enough about hscript yet to explain everything you can and can't do, but an important item to note is to not use properties in scripts, you must instead just use methods (as far as I can tell...). If you are more experienced in hscript, please submit a pull request with a fuller description:)

[5] As a developer, you may not want scripts, specificity mods, to have access to sensitive areas of your game. There are two ways to achieve this. The broad stroke way is to completely disallow access to the expose tag, ensuring scripts have no access to anything unless you specificity add it to the ScriptHandlers modules list. This can be done with: ScriptHandler.allowExpose = false;. The second, more specific way is to "blacklist" classes with ScriptHandler.Blacklist("");, this will warn the user they can not access this package.

Note: this section has been converted for the changes in IceEntity's scripting system, however, it is possible some items are incorrect, if something isn't working how it says, or how you feel it should, please let me know @nico_m__ ,thanks:D

Live Scripting

Live-Scripting Demo IceEntity allows you, with minimal effort, to literally code your game while it is running. This currently only works for CPP and Neko builds, and only for external script files (ie scripts created inside of your entities.xml file will not be editable at runtime). Here is what you need to do to make use of this new feature:

[1] Any scripts you want to be able to edit at runtime must be taken out of your xml file, placed in their own file, and then declared in your script element's "path" attribute.

[2] You can now import and request classes inside your scripts, similar to how imports are handled in regular haxe:

import flixel.FlxG; //this is identical to the expose tag in xml, except you do not control the name of the class, the last section is used, in this case: "FlxG".
request MyClass; //this is identical using a request tag in your xml

function init


This also means the live-reloader will update your imports, so you can add things as you need them. This, however, may slow the reloader, if you find that to be the case, you can add:

<haxedef name="ICE_NO_RELOAD_IMPORTS"/>

to your project.xml file.

[3] To help with testing, a new scripting function has been added:

function reload
	//This runs every time the script is updated, you can use it to edit variables for testing.
	//Note that it is ONLY for testing, you should not use this for any real game code, just for playing with values for speed and so on. THIS NEVER RUNS IN A FINAL BUILD

[4] The files you edit while making use of live reloading are not your main files. The files you want to edit are in: yourProject\export\windows\ (neko or cpp)\bin\assets\data. If you wish to use the logic you've created, copy these files back into your main folder when you are done.

[5] Reloading can be costly; you may not wish to reload every script when you only want to tweak a couple. To stop certain scripts from reloading, a "noreload" attribute has been added for script elements:

<script noreload="true">

Also, you may wish to do setup on an entity after the game begins, but not run any actual update logic. This means a script object is created, and updated every frame for no reason. To combat this, automatic script "cleaning" has been added, meaning that if only an "init" function is specified, the script will be deleted after running said "init" function. It is possible that during live-scripting, this behaviour may be annoying, if you wish to add other functions while running. In that case, you can specify:

<script noclean="true">

Note that cleaning only happens if the script only had an "init" function at startup, the noclean option is only meant to be used if you really have a need for it, in normal use you should never specify this option.

[6] All you need to do to use this system is to add: <haxedef name="ICE_LIVE_RELOAD"/> to your project.xml file. Thats it!!!! If you want control over the exact time between reloads, you can add a reloaddelay attribute to the root element of your entities.xml file.

[6] Important Notes:

This doesn't work for scripts specified in the "text" elements of your xml file

Make sure to remove <haxedef name="ICE_LIVE_RELOAD"/> before releasing your game, it is a performance waster, and has no use in a final build.

For more responsive use I recommend adding <haxedef name="FLX_NO_FOCUS_LOST_SCREEN"/> to your project.xml, and setting FlxG.autoPause to false. This way, you do not need to give the game focus to see your changes in action.

As a rule of thumb, any script that you wish to use the "update" function in should be contained in its own file, not in your xml file. This does two things: keeps your xml file readable, and combats the annoyance of live-reloading not working for internal scripts.

Message System:

IceEntity now includes a simple message broadcasting system. It is a useful way of quickly sending information or data between objects, without needing to store a reference. Simply call SendMessage() on an entity, with any info you need to send (explained in the method details). To receive messages, simply override the ReceiveMessage() function on an entity, and use that as a simple way to do whatever you want with the messages data, no need to manage a complicated event listener setup:)

Contact/Extra Info:

It should be fairly self explanatory, but if not, you can get in touch with me on twitter: @nico_m__.

You can’t perform that action at this time.