Android 自定义View及其在布局文件中的使用示例(三):结合Android 4.4.2_r1源码分析onMeasure过程

转载请注明出处
http://www.cnblogs.com/crashmaker/p/3549365.html
From crash_coder linguowu
linguowu0622@gamil.com

前言:

  通过Android 自定义View及其在布局文件中的使用示例Android 自定义View及其在布局文件中的使用示例(二),我们知道了如何使用自定义的View,以及Android绘制View的理论基础,其包含三个过程,测量View大小(通过onMeasure()方法实现),计算View位置(通过onLayout()方法实现),最后开始绘制(通过onDraw()方法实现),本篇,我们将结合Android 4.4.2_r1源码详细分析测量过程的具体实现.

  在第一篇里,我们提供了一个自定义的View的源代码,现在引用一下该代码与测量相关的部分:

  @Override
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
setMeasuredDimension(measureWidth(widthMeasureSpec),
measureHeight(heightMeasureSpec));
} /**
* Determines the width of this view
*
* @param measureSpec
* A measureSpec packed into an int
* @return The width of the view, honoring constraints from measureSpec
*/
private int measureWidth(int measureSpec) {
int result = 0;
int specMode = MeasureSpec.getMode(measureSpec);
int specSize = MeasureSpec.getSize(measureSpec); if (specMode == MeasureSpec.EXACTLY) {
// We were told how big to be
result = specSize;
} else {
// Measure the text
result = (int) mTextPaint.measureText(mText) + getPaddingLeft()
+ getPaddingRight();
if (specMode == MeasureSpec.AT_MOST) {
// Respect AT_MOST value if that was what is called for by
// measureSpec
result = Math.min(result, specSize);
}
} return result;
} /**
* Determines the height of this view
*
* @param measureSpec
* A measureSpec packed into an int
* @return The height of the view, honoring constraints from measureSpec
*/
private int measureHeight(int measureSpec) {
int result = 0;
int specMode = MeasureSpec.getMode(measureSpec);
int specSize = MeasureSpec.getSize(measureSpec); mAscent = (int) mTextPaint.ascent();
if (specMode == MeasureSpec.EXACTLY) {
// We were told how big to be
result = specSize;
} else {
// Measure the text (beware: ascent is a negative number)
result = (int) (-mAscent + mTextPaint.descent()) + getPaddingTop()
+ getPaddingBottom();
if (specMode == MeasureSpec.AT_MOST) {
// Respect AT_MOST value if that was what is called for by
// measureSpec
result = Math.min(result, specSize);
}
}
return result;
}
我们可以看到:protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec)是一个override的方法,它接收两个参数,通过字面意思,我们知道,这两个参数分别为宽度测量规格,高度测量规格,此时,我们会有一个疑问,这两个参数是从哪里来的?这个疑问咱们先记下来,给它编个号:Q01,暂时略过,到本文下一部分,我们就知道它的来龙去脉了.接着,我们来看onMeasure方法在本地的实现:
setMeasuredDimension(measureWidth(widthMeasureSpec),measureHeight(heightMeasureSpec));

我们跟进setMeasuredDimension(int,int)方法,看看它到底都做了些什么事情:

因为我们自定义的View是继承自View,所以我们进入View.java(源码位置:/frameworks/base/core/java/android/view/View.java)去看看有没有这个方法:

16575    /**
16576 * <p>This method must be called by {@link #onMeasure(int, int)} to store the
16577 * measured width and measured height. Failing to do so will trigger an
16578 * exception at measurement time.</p>
16579 *
16580 * @param measuredWidth The measured width of this view. May be a complex
16581 * bit mask as defined by {@link #MEASURED_SIZE_MASK} and
16582 * {@link #MEASURED_STATE_TOO_SMALL}.
16583 * @param measuredHeight The measured height of this view. May be a complex
16584 * bit mask as defined by {@link #MEASURED_SIZE_MASK} and
16585 * {@link #MEASURED_STATE_TOO_SMALL}.
16586 */
16587 protected final void setMeasuredDimension(int measuredWidth, int measuredHeight) {
16588 boolean optical = isLayoutModeOptical(this);
16589 if (optical != isLayoutModeOptical(mParent)) {
16590 Insets insets = getOpticalInsets();
16591 int opticalWidth = insets.left + insets.right;
16592 int opticalHeight = insets.top + insets.bottom;
16593
16594 measuredWidth += optical ? opticalWidth : -opticalWidth;
16595 measuredHeight += optical ? opticalHeight : -opticalHeight;
16596 }
16597 mMeasuredWidth = measuredWidth;
16598 mMeasuredHeight = measuredHeight;
16599
16600 mPrivateFlags |= PFLAG_MEASURED_DIMENSION_SET;
16601 }

果然,我们在View.java中找到了这个方法的具体实现,通过方法说明,得知此方法必须被onMeasure()方法调用 ,来保存测量到的宽度和高度,否则的话,会在测量时引发异常.通过代码主线 ,我们知道它将传进去的两个参数赋给本地的mMeasuredWidth和mMeasuredHeight变量,以便在View类中使用;好了,此时我们该抽离出来,回到我们出发的地方:

setMeasuredDimension(measureWidth(widthMeasureSpec),measureHeight(heightMeasureSpec));

有了上面的分析过程,我们知道这个方法中的measureWidth(widthMeasureSpec)是作为测量到的宽度,measureHeight(heightMeasureSpec)是作为测量到的高度,而这两个是需要我们在自定义的View中去实现的,由于测量宽度与高度的过程类似,我们在此文中仅分析measureWidth()的过程,很自然地,我们看看本地的measureWidth()是如何实现的:

 /**
* Determines the width of this view
*
* @param measureSpec
* A measureSpec packed into an int
* @return The width of the view, honoring constraints from measureSpec
*/
private int measureWidth(int measureSpec) {
int result = 0;
int specMode = MeasureSpec.getMode(measureSpec);
int specSize = MeasureSpec.getSize(measureSpec); if (specMode == MeasureSpec.EXACTLY) {
// We were told how big to be
result = specSize;
} else {
// Measure the text
result = (int) mTextPaint.measureText(mText) + getPaddingLeft()
+ getPaddingRight();
if (specMode == MeasureSpec.AT_MOST) {
// Respect AT_MOST value if that was what is called for by
// measureSpec
result = Math.min(result, specSize);
}
} return result;
}

该方法用来确定我们自定义的这个View的宽度,它接收onMeasure()的widthMeasureSpec参数,接着

         int specMode = MeasureSpec.getMode(measureSpec);
int specSize = MeasureSpec.getSize(measureSpec);
MeasureSpec.getMode(measureSpec),getMode()?我们在上一篇文章中的最后,有如下描述:
MeasureSpec:

  该对象封装了父容器传递给子元素的布局要求,它有三种模式:

1)
UNSPECIFIED:父容器对子元素没有要求,子元素可以得到任意值;
2)
EXACTLY:父窗口决定子元素的大小,子元素将被限定在给定的边界里而忽略它本身大小;
3)
AT MOST:子元素至多达到父窗口指定的大小,子元素不能超过这个边界;

所以我们会想,getMode()方法,应该就是获取上述这三种模式之一吧?我们跟进源码,看看getMode()都做了哪些事情:

18341        /**
18342 * Extracts the mode from the supplied measure specification.
18343 *
18344 * @param measureSpec the measure specification to extract the mode from
18345 * @return {@link android.view.View.MeasureSpec#UNSPECIFIED},
18346 * {@link android.view.View.MeasureSpec#AT_MOST} or
18347 * {@link android.view.View.MeasureSpec#EXACTLY}
18348 */
18349 public static int getMode(int measureSpec) {
18350 return (measureSpec & MODE_MASK);
18351 }

由此方法的文字描述部分,我们得知,该方法从接收的参数measureSpec中,获取到对应的三种模式之一,即返回measureSpec & MODE_MASK,这里的MODE_MASK又是个什么东西呢?在View.java中,我们找到在View这个类中,有个内部类MeasureSpec类

18289    public static class MeasureSpec {
18290 private static final int MODE_SHIFT = 30;
18291 private static final int MODE_MASK = 0x3 << MODE_SHIFT;
        ..............................................................

 18297 public static final int UNSPECIFIED = 0 << MODE_SHIFT;
 18298
  18299 /**
  18300 * Measure specification mode: The parent has determined an exact size
  18301 * for the child. The child is going to be given those bounds regardless
  18302 * of how big it wants to be.
  18303 */
  18304 public static final int EXACTLY = 1 << MODE_SHIFT;
  18305
  18306 /**
  18307 * Measure specification mode: The child can be as large as it wants up
  18308 * to the specified size.
  18309 */
  18310 public static final int AT_MOST = 2 << MODE_SHIFT;

        ................................
      }

所以,MODE_MASK的值为0x3左移了MODE_SHIFT(30)位,那么,用32位的二进制来表示的话,MODE_MASK为:1100 0000 0000 0000 0000 0000 0000 0000;如果非要探究此时的measureSpec & MODE_MASK后的值是多少,那么我们不妨用Debug模式调试一下我们的代码来获取getMode方法中传进来的参数measureSpec是什么值, 首先,从上面的源码中,可以知道三种MeasureSpec三种模式的值:

UNSPECIFIED = 0 << MODE_SHIFT;即:UNSPECIFIED为:0000 0000 0000 0000 0000 0000 0000 0000

其实我们也可以参阅官方文档对此值的定义:

public static final int UNSPECIFIED

Added in API level 1

Measure specification mode: The parent has not imposed any constraint on the child. It can be whatever size it wants.

Constant Value: 0 (0x00000000)

注:只不过官方文档此处用十六进制表示而已,以下两个模式也都用十六进制表示而已.
 
EXACTLY = 1 << MODE_SHIFT;即 EXACTLY为:0100 0000 0000 0000 0000 0000 0000 0000

public static final int EXACTLY


Added in API level 1

Measure specification mode: The parent has determined an exact size for the child. The child is going to be given those bounds regardless of how big it wants to be.

Constant Value: 1073741824 (0x40000000)
 
AT_MOST = 2 << MODE_SHIFT;即 AT_MOST为:1000 0000 0000 0000 0000 0000 0000 0000

public static final int AT_MOST




Added in API level 1



Measure specification mode: The child can be as large as it wants up to the specified size.

Constant Value: -2147483648 (0x80000000)

MODE_MASK为:1100 0000 0000 0000 0000 0000 0000 0000

好,我们来看一下debug前,自定义的View在布局文件中的layout_width的配置及我所调试的设备的屏幕像素为480*800,也就是我的显示屏宽为480像素;

<com.project.summary.customview.CustomView
android:id="@+id/customView"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
app:colorValue="@color/textRed"
app:textSize="20sp"
app:textString="This the Custom View1!!!" />

调试结果出来,此时传入的measureSpec的值是-2147483648,到了这里,我们又会产生一个疑问 ,为什么是它?为什么是这个值?我们先把这个疑问做个标记:Q02;到了文章最后,这个疑问就能解开了,这里先把思路跳出来,继续分析我们的measureWidth()这个本地方法的代码;

 1 /**
2 * Determines the width of this view
3 *
4 * @param measureSpec
5 * A measureSpec packed into an int
6 * @return The width of the view, honoring constraints from measureSpec
7 */
8 private int measureWidth(int measureSpec) {
9 int result = 0;
10 int specMode = MeasureSpec.getMode(measureSpec);
11 int specSize = MeasureSpec.getSize(measureSpec);
12
13 if (specMode == MeasureSpec.EXACTLY) {
14 // We were told how big to be
15 result = specSize;
16 } else {
17 // Measure the text
18 result = (int) mTextPaint.measureText(mText) + getPaddingLeft()
19 + getPaddingRight();
20 if (specMode == MeasureSpec.AT_MOST) {
21 // Respect AT_MOST value if that was what is called for by
22 // measureSpec
23 result = Math.min(result, specSize);
24 }
25 }
26
27 return result;
28 }

上面我们已经分析到第10行,由于第11行是获取传入的measureSpec的大小,过程与获取传入的measureSpec的模式类似,这里暂时先略过,接下来看第13行代码,这里要对获取到的模式进行判断,由上一篇文章,我们知道,如果自定义的View在布局文件中指定固定大小,那么,它的模式就是属于MeasureSepc.EXACTLY,此时,measureWidth()这个本地方法就返回11行所得的大小,否则进入另外一个分支,因为本系列中我们实现的实现上是一个类似于TextView的自定义控件,那么,这个View的大小就应该由它所绘制的文字长度来决定,此时,我们先计算出文字的宽度,然后再对其模式进行判断,如果模式是属于measureSpec.AT_MOST,我们通过数学运算,比较文字长度与通过传入的measureSpec所包含的大小,它们之中更小的那个做为我们控件的宽度.

  文章开头的相关代码中,本地方法:getMeasureHeight()的过程与本地方法getMeasureWidth()类似,在此不再分析.

  在此总结一下,文章开头引用的代码是我们在编写自定义View时,在重写onMeasure()这个方法时的一般步骤,那么,本文中的分析过程中还留有两个疑问:

Q01:
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec)是一个override的方法,它接收两个参数,通过字面意思,我们知道,这两个参数分别为宽度测量规格,高度测量规格,此时,我们会有一个疑问,这两个参数是从哪里来的?
Q02:
调试结果出来,此时传入的measureSpec的值是-2147483648,到了这里,我们又会产生一个疑问 ,为什么是它?为什么是这个值?

要探究这两个疑问,我们在本系列第二篇文章中,曾经提过Android绘制View的理论基础,从那篇文章中,我们明白,Android要绘制View的时候,必须要先遍历View的树形结构,并且先从最顶端的结点开始遍历,通过查找官方文档,我们进入

ViewRootImpl.java(文件位于:/frameworks/base/core/java/android/view/ViewRootImpl.java),一起找出上面的那两个疑问.........

/*********************************友情提醒:开始下面的探究前,最好先休息一下*********************************/

我们先大致浏览一下ViewRootImpl.java,这个文件代码有6707行有没有,不用怕,我们先找到一个叫performtraversals()的方法,看这字面意思,它是要开始遍历的节奏啊,果断跟进去看一下,顺便找找几个有用的干货:

private void performTraversals()
{
.......................................
1122 WindowManager.LayoutParams lp = mWindowAttributes;//详见分析PERFORMTRAVERSALS()点1 .........................................................
1155 Rect frame = mWinFrame;//详见分析PERFORMTRAVERSALS()点2
.......................................................
1563 if (mWidth != frame.width() || mHeight != frame.height()) {
1564 mWidth = frame.width();
1565 mHeight = frame.height();
1566 }
1567
.......................................................................
PERFORMTRAVERSALS()点3:
1634 if (!mStopped) {
1635 boolean focusChangedDueToTouchMode = ensureTouchModeLocally(
1636 (relayoutResult&WindowManagerGlobal.RELAYOUT_RES_IN_TOUCH_MODE) != 0);
1637 if (focusChangedDueToTouchMode || mWidth != host.getMeasuredWidth()
1638 || mHeight != host.getMeasuredHeight() || contentInsetsChanged) {
1639 int childWidthMeasureSpec = getRootMeasureSpec(mWidth, lp.width);//详见getRootMeasureSpec()方法的分析
1640 int childHeightMeasureSpec = getRootMeasureSpec(mHeight, lp.height);
1641
1642 if (DEBUG_LAYOUT) Log.v(TAG, "Ooops, something changed! mWidth="
1643 + mWidth + " measuredWidth=" + host.getMeasuredWidth()
1644 + " mHeight=" + mHeight
1645 + " measuredHeight=" + host.getMeasuredHeight()
1646 + " coveredInsetsChanged=" + contentInsetsChanged);
1647
1648 // Ask host how big it wants to be
1649 performMeasure(childWidthMeasureSpec, childHeightMeasureSpec);
}

/************************************分析PERFORMTRAVERSALS()点1 开始**********************************/

这里的lp用得还挺多,也许对我们有用,

因为

WindowManager.LayoutParams lp = mWindowAttributes;

所以我们分析一下这个mWindowAttributes是何方神圣:
分析PERFORMTRAVERSALS()点1:mWindowAttributes相关代码:

final WindowManager.LayoutParams mWindowAttributes = new WindowManager.LayoutParams();

我们进入WindowManager类的内部类LayoutParams的构造方法

 public LayoutParams() {
super(LayoutParams.MATCH_PARENT, LayoutParams.MATCH_PARENT);
type = TYPE_APPLICATION;
format = PixelFormat.OPAQUE;
}

其中有这么一句:注意两个参数都为LayoutParams.MATCH_PARENT

super(LayoutParams.MATCH_PARENT, LayoutParams.MATCH_PARENT);

因为WindowManager类的内部类LayoutParams继承自ViewGroup.LayoutParams,所以进入ViewGroup的内部类LayoutParams看一下
/frameworks/base/core/java/android/view/ViewGroup.java:

5829 public static class LayoutParams {
5830 /**
5831 * Special value for the height or width requested by a View.
5832 * FILL_PARENT means that the view wants to be as big as its parent,
5833 * minus the parent's padding, if any. This value is deprecated
5834 * starting in API Level 8 and replaced by {@link #MATCH_PARENT}.
5835 */
5836 @SuppressWarnings({"UnusedDeclaration"})
5837 @Deprecated
5838 public static final int FILL_PARENT = -1; .......................................... 5918 public LayoutParams(int width, int height) {
5919 this.width = width;
5920 this.height = height;
5921 }

分析总结:这里的width与height,都被赋为LayoutParams.MATCH_PARENT,所以这里的lp的宽与高,都为LayoutParams.MATCH_PARENT

/************************************分析PERFORMTRAVERSALS()点1 结束**********************************/

######################################################################################################################

/************************************分析PERFORMTRAVERSALS()点2 开始**********************************/

1563 if (mWidth != frame.width() || mHeight != frame.height()) {
1564 mWidth = frame.width();
1565 mHeight = frame.height();
1566 })

此时的mWidth为ViewRootImpl的变量,在这里使它的值为frame.width()的值;
frame又是从哪里来的呢?在performTraversals()方法中,1155行,原来它只是个局部变量,

1155        Rect frame = mWinFrame;

到了这里,关键就是找出mWinFrame了,继续找mWinFrame:
在ViewRootImpl的变量声明中:

256 final Rect mWinFrame; // frame given by window manager.

在ViewRootImpl这个类的构造方法中:

360 mWinFrame = new Rect();

frame given by window manager?那大概就是说mWinFrame是由窗口管理类来赋值的了,那么这么里mWinFrame应该就是屏幕的窗口大小了.我们这里先这么假设,后续文章再进行验证.

/**********************************************分析PERFORMTRAVERSALS()点2 结束**************************************/

/**********************************************分析PERFORMTRAVERSALS()点3开始**************************************/
PERFORMTRAVERSALS()点3:performMeasure(childWidthMeasureSpec, childHeightMeasureSpec);
1,两个参数:childWidthMeasureSpec,childHeightMeasureSpec分析
a)childWidthMeasureSpec:
int childWidthMeasureSpec = getRootMeasureSpec(mWidth, lp.width);
先分析getRootMeasureSpec的两个参数:
1)mWidth:
见分析点2(
1563 if (mWidth != frame.width() || mHeight != frame.height()) {
1564 mWidth = frame.width();
1565 mHeight = frame.height();
1566 })
所以猜想mWidth就是窗口的初始宽度(本文暂未验证)
2)lp.width:这里的lp就是分析点1中的 WindowManager.LayoutParams lp = mWindowAttributes;即:lp.width为LayoutParams.MATCH_PARENT;
由以上1)和2),我们先搞定了getRootMeasureSpec(mWidth,lp.width)这个方法的两个参数的意义,接下来,我们进入getRootMeasureSpec(mWidth,lp.width)这个方法

b)childHeightMeasureSpec:
int childHeightMeasureSpec = getRootMeasureSpec(mHeight, lp.height);
先分析getRootMeasureSpec的两个参数:
1)mHeight:类似上述的猜想,这里的mHeight就是窗口的初始高度
2)lp.height:这里的lp就是分析点1中的 WindowManager.LayoutParams lp = mWindowAttributes;即:lp.height为LayoutParams.MATCH_PARENT;

因为上述a)与b)的调用过程类似,只不过a)是获取宽度的规格,b)是获取高度的规格,所以以下分析只以获取宽度规格的过程来分析
*******************************************************************进入getRootMeasureSpec()方法的分析**********************************************

1924 private static int getRootMeasureSpec(int windowSize, int rootDimension) {
1925 int measureSpec;
1926 switch (rootDimension) {
1927
1928 case ViewGroup.LayoutParams.MATCH_PARENT:
1929 // Window can't resize. Force root view to be windowSize.
1930 measureSpec = MeasureSpec.makeMeasureSpec(windowSize, MeasureSpec.EXACTLY);
1931 break;
1932 case ViewGroup.LayoutParams.WRAP_CONTENT:
1933 // Window can resize. Set max size for root view.
1934 measureSpec = MeasureSpec.makeMeasureSpec(windowSize, MeasureSpec.AT_MOST);
1935 break;
1936 default:
1937 // Window wants to be an exact size. Force root view to be that size.
1938 measureSpec = MeasureSpec.makeMeasureSpec(rootDimension, MeasureSpec.EXACTLY);
1939 break;
1940 }
1941 return measureSpec;
1942 }

此方法接收的第二个参数rootDimension,就是lp.width,通过上面的分析,lp.width=LayoutParams.MATCH_PARENT,所以,进入第一个switch分支
此方法的返回值measureSpec=MeasureSpec.makeMeasureSpec(windowSize, MeasureSpec.EXACTLY);
所以,分析此方法,我们也知道,当我们的自定义View的layout_width/layout_height设置成MATCH_PARENT时,MODE 为MeasureSpec.EXACTLY;当设置成WRAP_CONTENT时,MODE为MeasureSpec.AT_MOST;

接下来我们分析1938行:

1938 measureSpec = MeasureSpec.makeMeasureSpec(rootDimension, MeasureSpec.EXACTLY);

***********************************************************************************************************************************************************

*******************************************************************进入MeasureSpec.makeMeasureSpec()方法的分析**********************************************

17245 /**
17246 * Creates a measure specification based on the supplied size and mode.
17247 *
17248 * The mode must always be one of the following:
17249 * <ul>
17250 * <li>{@link android.view.View.MeasureSpec#UNSPECIFIED}</li>
17251 * <li>{@link android.view.View.MeasureSpec#EXACTLY}</li>
17252 * <li>{@link android.view.View.MeasureSpec#AT_MOST}</li>
17253 * </ul>
17254 *
17255 * @param size the size of the measure specification
17256 * @param mode the mode of the measure specification
17257 * @return the measure specification based on size and mode
17258 */
17259 public static int makeMeasureSpec(int size, int mode) {
17260 return size + mode;
17261 }

此方法在/frameworks/base/core/java/android/view/View.java中的内部类MeasureSpec中的方法,该方法返回两个参数size+mode之和,参数size对应我们传进来的windowSize,即:窗口的初始宽度(当传进来的是mHeight时,为窗口的初始高度);
参数mode对应我们传进来的MeasureSpec.EXACTLY

**************************************************************************************************************************************************************************
有了上面这些分析之后,我们可以进入performMeasure(childWidthMeasureSpec, childHeightMeasureSpec)的分析了:

1913 private void performMeasure(int childWidthMeasureSpec, int childHeightMeasureSpec) {
1914 Trace.traceBegin(Trace.TRACE_TAG_VIEW, "measure");
1915 try {
1916 mView.measure(childWidthMeasureSpec, childHeightMeasureSpec);
1917 } finally {
1918 Trace.traceEnd(Trace.TRACE_TAG_VIEW);
1919 }
1920 }

*************************************mView.measure(childWidthMeasureSpec, childHeightMeasureSpec)的分析************************************************************

16450    /**
16451 * <p>
16452 * This is called to find out how big a view should be. The parent
16453 * supplies constraint information in the width and height parameters.
16454 * </p>
16455 *
16456 * <p>
16457 * The actual measurement work of a view is performed in
16458 * {@link #onMeasure(int, int)}, called by this method. Therefore, only
16459 * {@link #onMeasure(int, int)} can and must be overridden by subclasses.
16460 * </p>
16461 *
16462 *
16463 * @param widthMeasureSpec Horizontal space requirements as imposed by the
16464 * parent
16465 * @param heightMeasureSpec Vertical space requirements as imposed by the
16466 * parent
16467 *
16468 * @see #onMeasure(int, int)
16469 */
16470 public final void measure(int widthMeasureSpec, int heightMeasureSpec) {
         .....................................................
16496 // measure ourselves, this should set the measured dimension flag back
16497 onMeasure(widthMeasureSpec, heightMeasureSpec);
16498 mPrivateFlags3 &= ~PFLAG3_MEASURE_NEEDED_BEFORE_LAYOUT; 16522 }

这里的measure()方法是个final方法,结合该方法的说明,

The actual measurement work of a view is performed in onMeasure()

并且measure的两个参数同时传入onMeasure()中,

所以,才有了文章开头时引用的代码,在自定义的View中,重写onMeasure()方法,那么,本文上部分遗留下来的两个问题,至此就有了答案:

Q01:
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec)是一个override的方法,它接收两个参数,通过字面意思,我们知道,这两个参数分别为宽度测量规格,高度测量规格,此时,我们会有一个疑问,这两个参数是从哪里来的?
通过:1639 int childWidthMeasureSpec = getRootMeasureSpec(mWidth, lp.width);//详见getRootMeasureSpec()方法的分析,onMeasure的第一个参数widthMeasureSpec就是这里的childWidthMeasureSpec,heightMeasureSpec对应 childHeightMeasureSpec;
Q02:
调试结果出来,此时传入的measureSpec的值是-2147483648,到了这里,我们又会产生一个疑问 ,为什么是它?为什么是这个值? 那么这里的measureSpec就是MeasureSpec.makeMeasureSpec()方法的分析中,返回的size+mode;size是手机显示屏的像素宽或者高,文章上半部分中,我调试的手机像素宽是480,而且在自定义的View的布局文件中,layout_width设置成wrap_content,通过上面的分析,当设置成wrap_content时,模式为AT_MOST模式,通过文档描述,它的十进制值是-2147483648,那么size+mode就是480+(-2147483648)=-2147483168,也就是我们调试出来时,所得到的值-2147483648
)

另外,我们或许还会有一个疑问 :为什么MODE_MASK是1100 0000 0000 0000 0000 0000 0000 0000?EXACTLY为:0100 0000 0000 0000 0000 0000 0000 0000?

AT_MOST为:1000 0000 0000 0000 0000 0000 0000 0000?

其实对于这个问题,我们想,既然android规定了MODE必须是EXACTLY,AT_MOST,UNSPECIFIED这三种模式之一,那么,就可以用32位二进制的最高两位来表示,它有00,01,10,11这四种情况,那么它的MODE_MASK取值为

1100 0000 0000 0000 0000 0000 0000 0000就能很方便地取到它的模式了,由getMode()的实现:

return (measureSpec & MODE_MASK);

我们就可以取到它的最高两位,由此来确定它是哪种模式;同理对于getSize():

  public static int getSize(int measureSpec) {
return (measureSpec & ~MODE_MASK);
}

对于屏幕宽度,再大的屏幕也用不了32位二进制来表示其尺寸,所以才有measureSpec & ~MODE_MASK,这样就能取到它的值了.

 转载请注明出处
http://www.cnblogs.com/crashmaker/p/3549365.html 
From crash_coder linguowu
linguowu0622@gamil.com
 
 
上一篇:Android基础_2 Activity线性布局和表格布局


下一篇:jQuery如何追加tr到table中任意位置--向Table中指定位置添加tr或td(jQuery)