# Processors An Input Processor takes a value and returns a processed result for it. The received value and result value must be of the same type. For example, you can use a [clamp](#clamp) Processor to clamp values from a control to a certain range. >__Note__: To convert received input values into different types, see [composite Bindings](ActionBindings.md#composite-bindings). * [Using Processors](#using-processors) * [Processors on Bindings](#processors-on-bindings) * [Processors on Actions](#processors-on-actions) * [Processors on Controls](#processors-on-controls) * [Predefined Processors](#predefined-processors) * [Axis deadzone](#axis-deadzone) * [Clamp](#clamp) * [Invert](#invert) * [Invert Vector 2](#invert-vector-2) * [Invert Vector 3](#invert-vector-3) * [Normalize](#normalize) * [Normalize Vector 2](#normalize-vector-2) * [Normalize Vector 3](#normalize-vector-3) * [Scale](#scale) * [Scale Vector 2](#scale-vector-2) * [Scale Vector 3](#scale-vector-3) * [Stick deadzone](#stick-deadzone) * [Writing custom Processors](#writing-custom-processors) ## Using Processors You can install Processors on [bindings](ActionBindings.md), [actions](Actions.md) or on [controls](Controls.md). ### Processors on Bindings When you create Bindings for your [actions](Actions.md), you can choose to add Processors to the Bindings. These process the values from the controls they bind to, before the system applies them to the Action value. For instance, you might want to invert the `Vector2` values from the controls along the Y axis before passing these values to the Action that drives the input logic for your application. To do this, you can add an [Invert Vector2](#invert-vector-2) Processor to your Binding. If you're using [Input Action Assets](ActionAssets.md), you can add any Processor to your Bindings in the Input Action editor. Select the Binding you want to add Processors to so that the right pane of the window displays the properties for that Binding. Select the Add (+) icon on the __Processors__ foldout to open a list of all available Processors that match your control type, then choose a Processor type to add a Processor instance of that type. The Processor now appears under the __Processors__ foldout. If the Processor has any parameters, you can edit them in the __Processors__ foldout. ![Binding Processors](Images/BindingProcessors.png) To remove a Processor, click the Remove (-) icon next to it. You can also use the up and down arrows to change the order of Processors. This affects the order in which the system processes values. If you create your Bindings in code, you can add Processors like this: ```CSharp var action = new InputAction(); action.AddBinding("/leftStick") .WithProcessor("invertVector2(invertX=false)"); ``` ### Processors on Actions Processors on Actions work in the same way as Processors on Bindings, but they affect all controls bound to an Action, rather than just the controls from a specific Binding. If there are Processors on both the Binding and the Action, the system processes the ones from the Binding first. You can add and edit Processors on Actions in [Input Action Assets](ActionAssets.md) the [same way](#processors-on-bindings) as you would for Bindings: select an Action to edit, then add one or more Processors in the right window pane. If you create your Actions in code, you can add Processors like this: ```CSharp var action = new InputAction(processors: "invertVector2(invertX=false)"); ``` ### Processors on Controls You can have any number of Processors directly on an [`InputControl`](../api/UnityEngine.InputSystem.InputControl.html), which then process the values read from the Control. Whenever you call [`ReadValue`](../api/UnityEngine.InputSystem.InputControl-1.html#UnityEngine_InputSystem_InputControl_1_ReadValue) on a Control, all Processors on that Control process the value before it gets returned to you. You can use [`ReadUnprocessedValue`](../api/UnityEngine.InputSystem.InputControl-1.html#UnityEngine_InputSystem_InputControl_1_ReadUnprocessedValue) on a Control to bypass the Processors. The Input System adds Processors to a Control during device creation, if they're specified in the Control's [layout](Layouts.md). You can't add Processors to existing Controls after they've been created, so you can only add Processors to Controls when you're [creating custom devices](Devices.md#creating-custom-devices). The devices that the Input System supports out of the box already have some useful Processors added on their Controls. For instance, sticks on gamepads have a [Stick Deadzone](#stick-deadzone) Processor. If you're using a layout generated by the Input System from a [state struct](Devices.md#step-1-the-state-struct) using [`InputControlAttributes`](../api/UnityEngine.InputSystem.Layouts.InputControlAttribute.html), you can specify the Processors you want to use via the [`processors`](../api/UnityEngine.InputSystem.Layouts.InputControlAttribute.html#UnityEngine_InputSystem_Layouts_InputControlAttribute_processors) property of the attribute, like this: ```CSharp public struct MyDeviceState : IInputStateTypeInfo { public FourCC format => return new FourCC('M', 'Y', 'D', 'V'); // Add an axis deadzone to the Control to ignore values // smaller then 0.2, as our Control does not have a stable // resting position. [InputControl(layout = "Axis", processors = "AxisDeadzone(min=0.2)")] public short axis; } ``` If you [create a layout from JSON](Layouts.md#layout-from-json), you can specify Processors on your Controls like this: ```CSharp { "name" : "MyDevice", "extend" : "Gamepad", // Or some other thing "controls" : [ { "name" : "axis", "layout" : "Axis", "offset" : 4, "format" : "FLT", "processors" : "AxisDeadzone(min=0.2)" } ] } ``` ## Predefined Processors The Input System package comes with a set of useful Processors you can use. ### Axis Deadzone |__Name__|`AxisDeadzone`| |---|---| |__Operand Type__|`float`| |__Parameters__|`float min`
`float max`| An axis deadzone Processor scales the values of a Control so that any value with an absolute value smaller than `min` is 0, and any value with an absolute value larger than `max` is 1 or -1. Many Controls don't have a precise resting point (that is, they don't always report exactly 0 when the Control is in the center). Using the `min` value on a deadzone Processor avoids unintentional input from such Controls. Also, some Controls don't consistently report their maximum values when moving the axis all the way. Using the `max` value on a deadzone Processor ensures that you always get the maximum value in such cases. ### Clamp |__Name__|`Clamp`| |---|---| |__Operand Type__|`float`| |__Parameters__|`float min`
`float max`| Clamps input values to the [`min`..`max`] range. ### Invert |__Name__|`Invert`| |---|---| |__Operand Type__|`float`| Inverts the values from a Control (that is, multiplies the values by -1). ### Invert Vector 2 |__Name__|`InvertVector2`| |---|---| |__Operand Type__|`Vector2`| |__Parameters__|`bool invertX`
`bool invertY`| Inverts the values from a Control (that is, multiplies the values by -1). Inverts the x axis of the vector if `invertX` is true, and the y axis if `invertY` is true. ### Invert Vector 3 |__Name__|`InvertVector3`| |---|---| |__Operand Type__|`Vector3`| |__Parameters__|`bool invertX`
`bool invertY`
`bool invertZ`| Inverts the values from a Control (that is, multiplies the values by -1). Inverts the x axis of the vector if `invertX` is true, the y axis if `invertY` is true, and the z axis if `invertZ` is true. ### Normalize |__Name__|`Normalize`| |---|---| |__Operand Type__|`float`| |__Parameters__|`float min`
`float max`
`float zero`| Normalizes input values in the range [`min`..`max`] to unsigned normalized form [0..1] if `min` is >= `zero`, and to signed normalized form [-1..1] if `min` < `zero`. ### Normalize Vector 2 |__Name__|`NormalizeVector2`| |---|---| |__Operand Type__|`Vector2`| Normalizes input vectors to be of unit length (1). This is the same as calling `Vector2.normalized`. ### Normalize Vector 3 |__Name__|`NormalizeVector3`| |---|---| |__Operand Type__|`Vector3`| Normalizes input vectors to be of unit length (1). This is the same as calling `Vector3.normalized`. ### Scale |__Name__|`Scale`| |---|---| |__Operand Type__|`float`| |__Parameters__|`float factor`| Multiplies all input values by `factor`. ### Scale Vector 2 |__Name__|`ScaleVector2`| |---|---| |__Operand Type__|`Vector2`| |__Parameters__|`float x`
`float y`| Multiplies all input values by `x` along the X axis and by `y` along the Y axis. ### Scale Vector 3 |__Name__|`ScaleVector3`| |---|---| |__Operand Type__|`Vector3`| |__Parameters__|`float x`
`float y`
`float x`| Multiplies all input values by `x` along the X axis, by `y` along the Y axis, and by `z` along the Z axis. ### Stick deadzone |__Name__|`StickDeadzone`| |---|---| |__Operand Type__|`Vector2`| |__Parameters__|`float min`
`float max`| A stick deadzone Processor scales the values of a Vector2 Control, such as a stick, so that any input vector with a magnitude smaller than `min` results in (0,0), and any input vector with a magnitude greater than `max` is normalized to length 1. Many Controls don't have a precise resting point (that is, they don't always report exactly 0,0 when the Control is in the center). Using the `min` value on a deadzone Processor avoids unintentional input from such Controls. Also, some Controls don't consistently report their maximum values when moving the axis all the way. Using the `max` value on a deadzone Processor ensures that you always get the maximum value in such cases. ## Writing custom Processors You can also write custom Processors to use in your Project. Custom Processors are available in the UI and code in the same way as the built-in Processors. Add a class derived from [`InputProcessor`](../api/UnityEngine.InputSystem.InputProcessor-1.html), and implement the [`Process`](../api/UnityEngine.InputSystem.InputProcessor-1.html#UnityEngine_InputSystem_InputProcessor_1_Process__0_UnityEngine_InputSystem_InputControl_) method: ```CSharp public class MyValueShiftProcessor : InputProcessor { [Tooltip("Number to add to incoming values.")] public float valueShift = 0; public override float Process(float value, InputControl control) { return value + valueShift; } } ``` Now, you need to tell the Input System about your Processor. Call [`InputSystem.RegisterProcessor`](../api/UnityEngine.InputSystem.InputSystem.html#UnityEngine_InputSystem_InputSystem_RegisterProcessor__1_System_String_) in your initialization code. You can do so locally within the Processor class like this: ```CSharp #if UNITY_EDITOR [InitializeOnLoad] #endif public class MyValueShiftProcessor : InputProcessor { #if UNITY_EDITOR static MyValueShiftProcessor() { Initialize(); } #endif [RuntimeInitializeOnLoadMethod] static void Initialize() { InputSystem.RegisterProcessor(); } //... } ``` Your new Processor is now available in the [Input Action Asset Editor window](ActionAssets.md), and you can also add it in code like this: ```CSharp var action = new InputAction(processors: "myvalueshift(valueShift=2.3)"); ``` If you want to customize the UI for editing your Processor, create a custom [`InputParameterEditor`](../api/UnityEngine.InputSystem.Editor.InputParameterEditor-1.html) class for it: ```CSharp // No registration is necessary for an InputParameterEditor. // The system will automatically find subclasses based on the // <..> type parameter. #if UNITY_EDITOR public class MyValueShiftProcessorEditor : InputParameterEditor { private GUIContent m_SliderLabel = new GUIContent("Shift By"); public override void OnEnable() { // Put initialization code here. Use 'target' to refer // to the instance of MyValueShiftProcessor that is being // edited. } public override void OnGUI() { // Define your custom UI here using EditorGUILayout. target.valueShift = EditorGUILayout.Slider(m_SliderLabel, target.valueShift, 0, 10); } } #endif ```