什么情况下会onMeasure会执行?

进入View的measure方法:

void measure(){

    boolean forceLayout = (mPrivateFlags & PFLAG_FORCE_LAYOUT) == PFLAG_FORCE_LAYOUT;

    boolean specChanged = widthMeasureSpec != mOldWidthMeasureSpec

        || heightMeasureSpec != mOldHeightMeasureSpec;

    boolean isSepcExactly = MeasureSpec.getMode(widthMeasureSpec) == MeasureSpec.EXACTLY

        && MeasureSpec.getMode(heightMeasureSpec) == MeasureSpec.EXACTLY;

    boolean matchesSpecSize = getMeasuredWidth() == MeasureSpec.getSize(widthMeasureSpec)

        && getMeasuredHeight() == MeasureSpec.getSize(heightMeasureSpec);

    final boolean needsLayout = specChanged

        && (sAlwaysRemeasureExactly || !isSpecExactly || !matchesSpecSize);

    if(forceLayout || needLayout){

        int cacheIndex = forceLayout ? -1 : mMeasureCache.indexOfKey(key);

        if (cacheIndex < 0 || sIgnoreMeasureCache) {

            onMeasure(widthMeasureSpec, heightMeasureSpec);

            mPrivateFlags3 &= ~PFLAG3_MEASURE_NEEDED_BEFORE_LAYOUT;

        } else {

            long value = mMeasureCache.valueAt(cacheIndex);

            setMeasuredDimensionRaw((int) (value >> 32), (int) value);

            mPrivateFlags3 |= PFLAG3_MEASURE_NEEDED_BEFORE_LAYOUT;

        }

   }

}

什么时候forceLayout=true:

1. 
   
2. 调用requestLayout
   
3. 调用forceRequestLayout
   

什么时候needsLayout=true:

1. 
   
2. 当长宽发生改变
   

什么时候调用了onMeasure>方法:

1. 
   
2. forceLayouy=true
   
3. 或者mMeasureCache没有当前的缓存
   

所以总结:当调用了requestLayout一定会测发重测过程.当forceLayout=false的时候会去判断mMeasureCache值.现在研究下这个mMeasureCache

class View{

    LongSparseLongArray mMeasureCache;

    void measure(widthSpec,heightSpec){

        ---

        long key = (long) widthMeasureSpec << 32 | (long) heightMeasureSpec & 0xffffffffL;

        int cacheIndex = forceLayout ? -1 : mMeasureCache.indexOfKey(key);

        if(cacheIndex<0){

            onMeasure(widthSpec,heightSpec);

        }


        mOldWidthMeasureSpec = widthMeasureSpec;

        mOldHeightMeasureSpec = heightMeasureSpec;

        mMeasureCache.put(key,widhSpec|heightSpec);

        ---

    }

}

这里可以看到oldWidthMeasureSpec和mMeasureCache都是缓存上一次的值,那他们有什么不同呢?不同点就是,oldWidthMeasureSpec>不仅仅缓存了测量的spec模式而且缓存了size.但是mMeasureCache只缓存了size.从这行代码可以看出:

long key = (long) widthMeasureSpec << 32 | (long) heightMeasureSpec & 0xffffffffL;

这里一同运算就为了排除掉spec造成的影响.

//不信你可以试下下面的代码

public class Test {

    public static void main(String[] args) {

        long widthMeasureSpec = makeMeasureSpec(10,0);

        long heightMeasureSpec =  makeMeasureSpec(20,0);

        long ss = widthMeasureSpec << 32 | (long) heightMeasureSpec & 0xffffffffL;

        System.out.println("=========="+ss);

    }


    private static final int MODE_MASK = 0x3 << 30;


    public static int makeMeasureSpec(int size,

                                      int mode) {

        return (size & ~MODE_MASK) | (mode & MODE_MASK);

    }

}

//42949672980

//42949672980

//42949672980

什么时候mPrivateFlags会被赋值PFLAG_FORCE_LAYOUT.

在view viewGrouup的构造函数里面会主动赋值一次,然后在ViewGroup.addView时候会给当前View的mProvateFlags赋值PFLAG_FORCE_LAYOUT.

为什么onMeasure会被执行两次?

void measure(int widthMeasureSpec,int heightMeasureSpec){

    ----

    boolean forceLayout = (mPrivateFlags & PFLAG_FORCE_LAYOUT) == PFLAG_FORCE_LAYOUT; 

    if(forceLayout | needsLayout){

        onMeasure()

    }

    ----

}

public void layout(int l, int t, int r, int b){

    ---

    mPrivateFlags &= ~PFLAG_FORCE_LAYOUT;

    ---

}

在第一次触发到measure方法时,forceLayoyt=true needsLayout=true,但是layout方法还没触发到.

在第二次触发到measure>方法时,forceLayout=true needsLayout=false,所以还是会进入onMeasure方法.这次会执行layout方法.然后我们在下次的时候forceLayout就等于false了.上面的这一段分析是分析的measure内部如何防止多次调用onMeasure.

现在分析外部是如何多次调用measure方法的:

在Activity执行到onResume生命周期的时候,会执行WindowManager.addView操作,WindowManager的具体实现类是WindowManagerImpl然后addView操作交给了代理类WindowManagerGlobal,然后在WindowManagerGlobal的addView里面执行了ViewRootImpl.setView操作(ViewRootImpl对象也是在这个时候创建的),在ViewRootImpl会主动调用一次requestLayout,也就开启了第一次的视图 测量 布局 绘制.

在setView的时候主动调用了一次ViewRootImpl.requestLayout,注意这个requestLayout是ViewRootImpl的内部方法,和view viewGroup那些requestLayout不一样.在ViewRootImpl.requestLayout内部调用了performTraversals方法:

class ViewRootImpl{

    void performTraversals(){

        if(layoutResuested){

        //标记1

            windowSizeMayChanged |= measureHierarchy(host,lp,res,desiredWindowWidth,desiredWindowHeight);

        }

        //标记2

        performMeasure()

        performLayout()

    }

    void measureHierarchy(){

        performMeasure()

    }

}

从ViewRootImpl的执行逻辑你可以看出,在执行performLayout之前,他自己就已经调用了两次performMeasure方法.所以你现在就知道为啥了.