// TODO: Document GLSL Shaders A ShaderJob instance is used to execute a shader operation in stand-alone mode. The shader operation executes and returns its result data. It is up to the developer to determine how to use the result. There are two primary reasons for using a shader in stand-alone mode:

  • Processing non-image data: Using a ShaderJob instance you have control over input values and over how the shader result is used. The shader can return the result as binary data or number data instead of image data.
  • Background processing: Some shaders are complex and require a notable amount of time to execute. Executing a complex shader in the main execution of an application could slow down other parts of the application such as user interaction or updating the screen. Using a ShaderJob instance, you can execute the shader in the background. When the shader is executed in this way, the shader operation runs separate from the main execution of the application.

The shader property (or constructor parameter) specifies the Shader instance representing the shader that is used for the operation. You provide any parameter or input that the shader expects using the associated ShaderParameter or ShaderInput instance.

Before execution a ShaderJob operation, you provide an object into which the result is written, by setting it as the value of the target property. When the shader operation completes the result is written into the target object.

To begin a background shader operation, call the start() method. When the operation finishes the result is written into the target object. At that point the ShaderJob instance dispatches a complete event, notifying listeners that the result is available.

To execute a shader synchronously (that is, not running in the background), call the start() method and pass true as an argument. The shader runs in the main execution thread and your code pauses until the operation completes. When it finishes the result is written into the target object. At that point the application continues running at the next line of code.

Events:

complete

Dispatched when a ShaderJob that executes asynchronously finishes processing the data using the shader. A ShaderJob instance executes asynchronously when the start() method is called with a false value for the waitForCompletion parameter.

Constructor

@:value({ height : 0, width : 0, target : null, shader : null })@SuppressWarnings("checkstyle:Dynamic")new(?shader:Shader, ?target:Dynamic, width:Int = 0, height:Int = 0)

Parameters:

shader

The shader to use for the operation.

target

The object into which the result of the shader operation is written. This argument must be a BitmapData, ByteArray, or Vector. instance.

width

The width of the result data in the target if it is a ByteArray or Vector. instance. The size of the ByteArray or Vector. instance is enlarged if necessary and existing data is overwritten.

height

The height of the result data in the target if it is a ByteArray or Vector. instance. The size of the ByteArray or Vector. instance is enlarged if necessary and existing data is overwritten.

Variables

height:Int

The height of the result data in the target if it is a ByteArray or Vector. instance. The size of the ByteArray or Vector. instance is enlarged if necessary and existing data is overwritten.

read onlyprogress:Float

The progress of a running shader. This property is a value from 0 through 1. Zero is the initial value (0% complete). One indicates that the shader has completed its operation. If the cancel() method is called this property becomes undefined, and its value cannot be used reliably until the shader operation starts again.

shader:Shader

The shader that's used for the operation. Any input or parameter that the shader expects must be provided using the ShaderInput or ShaderParameter property of the Shader instance's data property. An input must be provided using its corresponding ShaderInput even if it is the same as the target object. To process a ByteArray containing a linear array of data (as opposed to image data) set the corresponding ShaderInput instance's height to 1 and width to the number of 32-bit floating-point values in the ByteArray. In that case, the input in the shader must be defined with the image1 data type.

@SuppressWarnings("checkstyle:Dynamic")target:Dynamic

The object into which the result of the shader operation is written. This object must be a BitmapData, ByteArray, or Vector. instance.

width:Int

The width of the result data in the target if it is a ByteArray or Vector. instance. The size of the ByteArray or Vector. instance is enlarged if necessary and existing data is overwritten.

Methods

cancel():Void

Cancels the currently running shader operation. Any result data that is already computed is discarded. The complete event is not dispatched. Calling cancel() multiple times has no additional effect.

@:value({ waitForCompletion : false })start(waitForCompletion:Bool = false):Void

Starts a shader operation in synchronous or asynchronous mode, according to the value of the waitForCompletion parameter. In asynchronous mode (when waitForCompletion is false), which is the default, the ShaderJob execution runs in the background. The shader operation does not affect the responsiveness of the display or other operations. In asynchronous mode the start() call returns immediately and the program continues with the next line of code. When the asynchronous shader operation finishes, the result is available and the complete event is dispatched.

Only one background ShaderJob operation executes at a time. Shader operations are held in a queue until they execute. If you call the start() method while a shader operation is executing, the additional operation is added to the end of the queue. Later, when its turn comes, it executes.

To execute a shader operation in synchronous mode, call start() with a true value for the waitForCompletion parameter (the only parameter). Your code pauses at the point where start() is called until the shader operation completes. At that point the result is available and execution continues with the next line of code.

When you call the start() method the Shader instance in the shader property is copied internally. The shader operation uses that internal copy, not a reference to the original shader. Any changes made to the shader, such as changing a parameter value, input, or bytecode, are not applied to the copied shader that's used for the shader processing. To incorporate shader changes into the shader processing, call the cancel() method (if necessary) and call the start() method again to restart the shader processing.

While a shader operation is executing, the target object's value is not changed. When the operation finishes (and the complete event is dispatched in asynchronous mode) the entire result is written to the target object at one time. If the target object is a BitmapData instance and its dispose() method is called before the operation finishes, the complete event is still dispatched in asynchronous mode. However, the result data is not written to the BitmapData object because it is in a disposed state.

Parameters:

waitForCompletion

Specifies whether to execute the shader in the background (false, the default) or in the main program execution (true).

Throws:

ArgumentError

When the target property is null or is not a BitmapData, ByteArray, or Vector. instance.

ArgumentError

When the shader specifies an image input that isn't provided.

ArgumentError

When a ByteArray or Vector. instance is used as an input and the width and height properties aren't specified for the ShaderInput, or the specified values don't match the amount of data in the input object. See the ShaderInput.input property for more information.

Events:

complete

Dispatched when the operation finishes, if the start() method is called with a waitForCompletion argument of true.

Inherited Variables

Inherited Methods

Defined by EventDispatcher

@:value({ useWeakReference : false, priority : 0, useCapture : false })addEventListener<T>(type:EventType<T>, listener:T ‑> Void, useCapture:Bool = false, priority:Int = 0, useWeakReference:Bool = false):Void

Registers an event listener object with an EventDispatcher object so that the listener receives notification of an event. You can register event listeners on all nodes in the display list for a specific type of event, phase, and priority.

After you successfully register an event listener, you cannot change its priority through additional calls to addEventListener(). To change a listener's priority, you must first call removeListener(). Then you can register the listener again with the new priority level.

Keep in mind that after the listener is registered, subsequent calls to addEventListener() with a different type or useCapture value result in the creation of a separate listener registration. For example, if you first register a listener with useCapture set to true, it listens only during the capture phase. If you call addEventListener() again using the same listener object, but with useCapture set to false, you have two separate listeners: one that listens during the capture phase and another that listens during the target and bubbling phases.

You cannot register an event listener for only the target phase or the bubbling phase. Those phases are coupled during registration because bubbling applies only to the ancestors of the target node.

If you no longer need an event listener, remove it by calling removeEventListener(), or memory problems could result. Event listeners are not automatically removed from memory because the garbage collector does not remove the listener as long as the dispatching object exists(unless the useWeakReference parameter is set to true).

Copying an EventDispatcher instance does not copy the event listeners attached to it.(If your newly created node needs an event listener, you must attach the listener after creating the node.) However, if you move an EventDispatcher instance, the event listeners attached to it move along with it.

If the event listener is being registered on a node while an event is being processed on this node, the event listener is not triggered during the current phase but can be triggered during a later phase in the event flow, such as the bubbling phase.

If an event listener is removed from a node while an event is being processed on the node, it is still triggered by the current actions. After it is removed, the event listener is never invoked again(unless registered again for future processing).

Parameters:

type

The type of event.

useCapture

Determines whether the listener works in the capture phase or the target and bubbling phases. If useCapture is set to true, the listener processes the event only during the capture phase and not in the target or bubbling phase. If useCapture is false, the listener processes the event only during the target or bubbling phase. To listen for the event in all three phases, call addEventListener twice, once with useCapture set to true, then again with useCapture set to false.

priority

The priority level of the event listener. The priority is designated by a signed 32-bit integer. The higher the number, the higher the priority. All listeners with priority n are processed before listeners of priority n-1. If two or more listeners share the same priority, they are processed in the order in which they were added. The default priority is 0.

useWeakReference

Determines whether the reference to the listener is strong or weak. A strong reference(the default) prevents your listener from being garbage-collected. A weak reference does not.

Class-level member functions are not subject to garbage collection, so you can set useWeakReference to true for class-level member functions without subjecting them to garbage collection. If you set useWeakReference to true for a listener that is a nested inner function, the function will be garbage-collected and no longer persistent. If you create references to the inner function (save it in another variable) then it is not garbage-collected and stays persistent.

Weak references are supported on some OpenFL targets only, including html5, cpp, and flash/air. On other targets, this parameter is ignored, and the reference will be strong instead.

Throws:

ArgumentError

The listener specified is not a function.

dispatchEvent(event:Event):Bool

Dispatches an event into the event flow. The event target is the EventDispatcher object upon which the dispatchEvent() method is called.

Parameters:

event

The Event object that is dispatched into the event flow. If the event is being redispatched, a clone of the event is created automatically. After an event is dispatched, its target property cannot be changed, so you must create a new copy of the event for redispatching to work.

Returns:

A value of true if the event was successfully dispatched. A value of false indicates failure or that preventDefault() was called on the event.

Throws:

Error

The event dispatch recursion limit has been reached.

hasEventListener(type:String):Bool

Checks whether the EventDispatcher object has any listeners registered for a specific type of event. This allows you to determine where an EventDispatcher object has altered handling of an event type in the event flow hierarchy. To determine whether a specific event type actually triggers an event listener, use willTrigger().

The difference between hasEventListener() and willTrigger() is that hasEventListener() examines only the object to which it belongs, whereas willTrigger() examines the entire event flow for the event specified by the type parameter.

When hasEventListener() is called from a LoaderInfo object, only the listeners that the caller can access are considered.

Parameters:

type

The type of event.

Returns:

A value of true if a listener of the specified type is registered; false otherwise.

@:value({ useCapture : false })removeEventListener<T>(type:EventType<T>, listener:T ‑> Void, useCapture:Bool = false):Void

Removes a listener from the EventDispatcher object. If there is no matching listener registered with the EventDispatcher object, a call to this method has no effect.

Parameters:

type

The type of event.

useCapture

Specifies whether the listener was registered for the capture phase or the target and bubbling phases. If the listener was registered for both the capture phase and the target and bubbling phases, two calls to removeEventListener() are required to remove both, one call with useCapture() set to true, and another call with useCapture() set to false.

toString():String

willTrigger(type:String):Bool

Checks whether an event listener is registered with this EventDispatcher object or any of its ancestors for the specified event type. This method returns true if an event listener is triggered during any phase of the event flow when an event of the specified type is dispatched to this EventDispatcher object or any of its descendants.

The difference between the hasEventListener() and the willTrigger() methods is that hasEventListener() examines only the object to which it belongs, whereas the willTrigger() method examines the entire event flow for the event specified by the type parameter.

When willTrigger() is called from a LoaderInfo object, only the listeners that the caller can access are considered.

Parameters:

type

The type of event.

Returns:

A value of true if a listener of the specified type will be triggered; false otherwise.