Unity 为NGUI增加体感输入方式

背景

NGUI在处理UI和输入方面确实做的不错，但是现在的问题是公司引入体感之后，是通过手的位置来实现按钮的点击操作，前提我不想改变原先设计好的NGUI界面和机制，怎么破？

NGUI的输入底层机制

NGUI对鼠标或者触摸的位置是通过Camera对NGUI层进行射线检测来获得，然后检测按钮事件、触摸屏Press事件来实现UI的操作，从事件机制上而言，NGUI虽然提供了几种事件机制，但底层还是通过Camera的SendMessage来通知被检测到的控件完成某个事件，OK ，看看代码。

NGUI UICamera中进行射线检测的片段 ，利用当前坐标位置转换到世界坐标系射线

static public bool Raycast (Vector3 inPos)
    {
        for (int i = ; i < list.size; ++i)
        {
            UICamera cam = list.buffer[i];

            // Skip inactive scripts
            if (!cam.enabled || !NGUITools.GetActive(cam.gameObject)) continue;

            // Convert to view space
            currentCamera = cam.cachedCamera;
            Vector3 pos = currentCamera.ScreenToViewportPoint(inPos);

            ..

..

        }

     }

 NGUI UICamera 中ProcessMouse()方法利用射线检测当前响应的按钮，并通过Notify（）通知控件。

// No need to perform raycasts every frame
        if (isPressed || posChanged || mNextRaycast < RealTime.time)
        {
            mNextRaycast = RealTime.time + 0.02f;
            if (!Raycast(Input.mousePosition)) hoveredObject = fallThrough;
            if (hoveredObject == null) hoveredObject = genericEventHandler;
            for (int i = ; i < ; ++i) mMouse[i].current = hoveredObject;
        }

…

…

…
        // The button was released over a different object -- remove the highlight from the previous
        if ((justPressed || !isPressed) && mHover != null && highlightChanged)
        {
            currentScheme = ControlScheme.Mouse;
            if (mTooltip != null) ShowTooltip(false);
            Notify(mHover, "OnHover", false);
            mHover = null;
        }

NGUI UICamera中封装的通知方法，Camera通过逻辑判断发送不同类型，如在点击的时候发送 ：Notify(currentTouch.pressed, "OnClick", null);，在按下时候发送：Notify(currentTouch.current, "OnHover", true);

static public void Notify (GameObject go, string funcName, object obj)
    {
        if (mNotifying) return;
        mNotifying = true;

        if (NGUITools.GetActive(go))
        {
            go.SendMessage(funcName, obj, SendMessageOptions.DontRequireReceiver);

            if (genericEventHandler != null && genericEventHandler != go)
            {
                genericEventHandler.SendMessage(funcName, obj, SendMessageOptions.DontRequireReceiver);
            }
        }
        mNotifying = false;
    }

NGUI 中 UIButton中Onclick处理OnClick （）与OnDragOver（）事件，是对Notify(currentTouch.pressed, "OnClick", null)的响应，同时通过EventDelegate.Execute(onClick); 来实现委托。

/// <summary>
    /// Call the listener function.
    /// </summary>

    protected virtual void OnClick ()
    {
        if (current == null && isEnabled)
        {
            current = this;
            EventDelegate.Execute(onClick);
            current = null;
        }
    }

OK, 说到底，NGUI底层还是通过SendMessage来实现，那增加一种输入方式怎么破？

解决方案：引入新的射线检测

好的方式是直接修改NGUI底层UICamera代码逻辑，增加一种体感输入，不过涉及到的太多，倒不如再UIROOT或者UICamera增加一个专门用于体感输入方法，把手的位置作为鼠标，增加一个对NGUI层的射线检测机制，对检测到的按钮发送SendMessage消息，当然发送的内容和NGUI中的一样，就可以保证不修改NGUI的UI脚本等等，实现体感输入。代码示例：

/// <summary>
    /// 作者：细雨淅淅，地址：http://www.cnblogs.com/zsb517/
    /// 摄像机射线检测，对检测到的控件重新设置状态，但不影响鼠标状态
    /// </summary>
    private void OnRayCollision()
    {

        if (UICamera == null || CursorGrp == null)
        {
            return;
        }

        Vector3 realPos     = ScreentoWorldPoint();
        Ray     rayCamera   = UICamera.ScreenPointToRay(realPos);
        RaycastHit hit;

        if (Physics.Raycast(rayCamera, out hit, 1000f, LayMaskCollis.value))
        {
            if (hit.collider == null || hit.collider.gameObject == null)
            {
                return ;
            }

            if (curButton == null )
            {
               // Debug.Log(hit.collider.name);
                curButton = hit.collider.gameObject.GetComponent<UIButton>();
                if (curButton != null)
                {
                    curButton.SetState(UIButtonColor.State.Hover,false);
                  //  curButton.SendMessage("OnHover", true, SendMessageOptions.DontRequireReceiver);
                }

                return;
            }
            else if (curButton != null)
            {
                if (curButton != hit.transform.gameObject.GetComponent<UIButton>())
                {
                    //Restore previous button ,and make new button to hover status
                    //curButton.SendMessage("OnPress", false, SendMessageOptions.DontRequireReceiver);
                    curButton.SetState(UIButtonColor.State.Normal, false);
                    curButton = hit.transform.gameObject.GetComponent<UIButton>();
                    if (curButton)
                    {
                        curButton.SetState(UIButtonColor.State.Hover, false);
                        //curButton.SendMessage("OnHover", true, SendMessageOptions.DontRequireReceiver);
                    }

                    return;
                }
                else
                {

                }
            }
        }
        else
        {
            if (curButton != null)
            {
                curButton.SetState(UIButtonColor.State.Hover, false);

                //curButton.SendMessage("OnPress", false, SendMessageOptions.DontRequireReceiver);
                //curButton.SendMessage("OnHover", false, SendMessageOptions.DontRequireReceiver);
                curButton = null;
                return;
            }
        }

        OnReset();
    }

    private Vector3 ScreentoWorldPoint()
    {
        Vector3 wPos = Vector3.zero;
        if (IsMouseOrKinect)
        {
            wPos = new Vector3(Input.mousePosition.x, Input.mousePosition.y, );
        }
        else
        {
            if (NIHandInput.GetInstance() != null)
            {
                Vector2 pos = NIHandInput.GetInstance().ScreenPos;  //体感输入中手的位置
                wPos = new Vector3(pos.x, pos.y, );
            }
        }

        Debuger.Log(wPos);
        return wPos;
    }

结论

对于输入问题，Unity Input自身还提供了一种机制，不过没做太多研究，但是想把各种输入柔和在一起，确实是一件很纠结的事情，还要多考虑，不仅仅是代码问题，而是改变了整个游戏的体验。

细雨标记: unity