2015-07-28 17:29:19
这一篇主要看看布局过程
一、布局过程肯定要不可避免的涉及到layout()和onLayout()方法，这两个方法都是定义在View.java中，源码如下：
/**
* Assign a size and position to a view and all of its
* descendants
*
* <p>This is the second phase of the layout mechanism.
* (The first is measuring). In this phase, each parent calls
* layout on all of its children to position them.
* This is typically done using the child measurements
* that were stored in the measure pass().</p>
*
* <p>Derived classes should not override this method.
* Derived classes with children should override
* onLayout. In that method, they should
* call layout on each of their children.</p>
*
* @param l Left position, relative to parent
* @param t Top position, relative to parent
* @param r Right position, relative to parent
* @param b Bottom position, relative to parent
*/
@SuppressWarnings({"unchecked"})
public void layout(int l, int t, int r, int b) {
if ((mPrivateFlags3 & PFLAG3_MEASURE_NEEDED_BEFORE_LAYOUT) != 0) {
onMeasure(mOldWidthMeasureSpec, mOldHeightMeasureSpec);
mPrivateFlags3 &= ~PFLAG3_MEASURE_NEEDED_BEFORE_LAYOUT;
} int oldL = mLeft;
int oldT = mTop;
int oldB = mBottom;
int oldR = mRight; boolean changed = isLayoutModeOptical(mParent) ?
setOpticalFrame(l, t, r, b) : setFrame(l, t, r, b);
if (changed || (mPrivateFlags & PFLAG_LAYOUT_REQUIRED) == PFLAG_LAYOUT_REQUIRED) {
/// M: Monitor onLayout time if longer than 3s print log.
long logTime = System.currentTimeMillis();
onLayout(changed, l, t, r, b);
long nowTime = System.currentTimeMillis();
if (nowTime - logTime > DBG_TIMEOUT_VALUE) {
Xlog.d(VIEW_LOG_TAG, "[ANR Warning]onLayout time too long, this =" + this + "time =" + (nowTime - logTime));
}
mPrivateFlags &= ~PFLAG_LAYOUT_REQUIRED; ListenerInfo li = mListenerInfo;
if (li != null && li.mOnLayoutChangeListeners != null) {
ArrayList<OnLayoutChangeListener> listenersCopy =
(ArrayList<OnLayoutChangeListener>)li.mOnLayoutChangeListeners.clone();
int numListeners = listenersCopy.size();
for (int i = 0; i < numListeners; ++i) {
listenersCopy.get(i).onLayoutChange(this, l, t, r, b, oldL, oldT, oldR, oldB);
}
}
} mPrivateFlags &= ~PFLAG_FORCE_LAYOUT;
mPrivateFlags3 |= PFLAG3_IS_LAID_OUT;
} /**
* Called from layout when this view should
* assign a size and position to each of its children.
*
* Derived classes with children should override
* this method and call layout on each of
* their children.
* @param changed This is a new size or position for this view
* @param left Left position, relative to parent
* @param top Top position, relative to parent
* @param right Right position, relative to parent
* @param bottom Bottom position, relative to parent
*/
protected void onLayout(boolean changed, int left, int top, int right, int bottom) {
}
简单的翻译一下layout()方法的那段注释哈～E文不好～
“指定一个view以及它的所有子孙节点的大小和位置，这是布局机制的第二阶段(第一阶段是测量)，在这一阶段，父view调用所有子view的layout()方法以确定他们所在的位置，通常是使用子View存储的自身的尺寸。派生类不应该重写此方法，应该重写onLayout()方法，在派生类重写的onLayout()方法中，应该调用每一个子View的layout方法。”啰嗦一句，int l, int t, int r, int b都是相对于父节点的坐标值。
注意layout方法中的红色代码，调用了onLayout。而onLayout在view中实现为空。现在来看看ViewGroup中的这两个方法。
/**
* {@inheritDoc}
*/
@Override
public final void layout(int l, int t, int r, int b) {
if (!mSuppressLayout && (mTransition == null || !mTransition.isChangingLayout())) {
if (mTransition != null) {
mTransition.layoutChange(this);
}
super.layout(l, t, r, b);
} else {
// record the fact that we noop'd it; request layout when transition finishes
mLayoutCalledWhileSuppressed = true;
}
} /**
* {@inheritDoc}
*/
@Override
protected abstract void onLayout(boolean changed,
int l, int t, int r, int b);
在ViewGroup的layout方法中，mSuppressLayout用来控制是否禁止调用layout()，该值由如下方法来控制：
/**
* Tells this ViewGroup to suppress all layout() calls until layout
* suppression is disabled with a later call to suppressLayout(false).
* When layout suppression is disabled, a requestLayout() call is sent
* if layout() was attempted while layout was being suppressed.
*
* @hide
*/
public void suppressLayout(boolean suppress) {
mSuppressLayout = suppress;
if (!suppress) {
if (mLayoutCalledWhileSuppressed) {
requestLayout();
mLayoutCalledWhileSuppressed = false;
}
}
}
这个方法不是对外公开的，所以不了解它也行。可以简单地理解ViewGroup的layout方法，它直接调用了父类View的layout()方法即可。至于onLayout方法，竟然被搞成了abstract的，这是逼着ViewGroup的子类必须得去实现啊～当然了，你必须得实现啊，你作为一个容器类，如何摆放你的子孙控件，是你义不容辞的责任啊。
至此我们已经明白了几点：
1. 派生类不需要重写layout()，而应该重写onLayout()方法，因为在layout()方法中就调用了onLayout()。
2. 在重写onLayout()方法时，我们需要显式的调用每一个childView的layout方法，把它摆放在合适的位置上。前提是在调用之前，得先计算好该childView的坐标。
3. 如果直接继承自View,那么可以不用重写onLayout()方法，比如ImageView、ImageButton等都没有重写该方法，所以不重写这个方法对于自定义View影响不大，至于TextView比较特殊，它重写了该方法，如下：
@Override
protected void onLayout(boolean changed, int left, int top, int right, int bottom) {
super.onLayout(changed, left, top, right, bottom);
if (mDeferScroll >= 0) {
int curs = mDeferScroll;
mDeferScroll = -1;
bringPointIntoView(Math.min(curs, mText.length()));
}
} /**
* Move the point, specified by the offset, into the view if it is needed.
* This has to be called after layout. Returns true if anything changed.
*/
public boolean bringPointIntoView(int offset) {
if (isLayoutRequested()) {
mDeferScroll = offset;
return false;
}
boolean changed = false; Layout layout = isShowingHint() ? mHintLayout: mLayout; if (layout == null) return changed; int line = layout.getLineForOffset(offset); int grav; switch (layout.getParagraphAlignment(line)) {
case ALIGN_LEFT:
grav = 1;
break;
case ALIGN_RIGHT:
grav = -1;
break;
case ALIGN_NORMAL:
grav = layout.getParagraphDirection(line);
break;
case ALIGN_OPPOSITE:
grav = -layout.getParagraphDirection(line);
break;
case ALIGN_CENTER:
default:
grav = 0;
break;
} // We only want to clamp the cursor to fit within the layout width
// in left-to-right modes, because in a right to left alignment,
// we want to scroll to keep the line-right on the screen, as other
// lines are likely to have text flush with the right margin, which
// we want to keep visible.
// A better long-term solution would probably be to measure both
// the full line and a blank-trimmed version, and, for example, use
// the latter measurement for centering and right alignment, but for
// the time being we only implement the cursor clamping in left to
// right where it is most likely to be annoying.
final boolean clamped = grav > 0;
// FIXME: Is it okay to truncate this, or should we round?
final int x = (int)layout.getPrimaryHorizontal(offset, clamped);
final int top = layout.getLineTop(line);
final int bottom = layout.getLineTop(line + 1); int left = (int) FloatMath.floor(layout.getLineLeft(line));
int right = (int) FloatMath.ceil(layout.getLineRight(line));
int ht = layout.getHeight(); int hspace = mRight - mLeft - getCompoundPaddingLeft() - getCompoundPaddingRight();
int vspace = mBottom - mTop - getExtendedPaddingTop() - getExtendedPaddingBottom();
if (!mHorizontallyScrolling && right - left > hspace && right > x) {
// If cursor has been clamped, make sure we don't scroll.
right = Math.max(x, left + hspace);
} int hslack = (bottom - top) / 2;
int vslack = hslack; if (vslack > vspace / 4)
vslack = vspace / 4;
if (hslack > hspace / 4)
hslack = hspace / 4; int hs = mScrollX;
int vs = mScrollY; if (top - vs < vslack)
vs = top - vslack;
if (bottom - vs > vspace - vslack)
vs = bottom - (vspace - vslack);
if (ht - vs < vspace)
vs = ht - vspace;
if (0 - vs > 0)
vs = 0; if (grav != 0) {
if (x - hs < hslack) {
hs = x - hslack;
}
if (x - hs > hspace - hslack) {
hs = x - (hspace - hslack);
}
} if (grav < 0) {
if (left - hs > 0)
hs = left;
if (right - hs < hspace)
hs = right - hspace;
} else if (grav > 0) {
if (right - hs < hspace)
hs = right - hspace;
if (left - hs > 0)
hs = left;
} else /* grav == 0 */ {
if (right - left <= hspace) {
/*
* If the entire text fits, center it exactly.
*/
hs = left - (hspace - (right - left)) / 2;
} else if (x > right - hslack) {
/*
* If we are near the right edge, keep the right edge
* at the edge of the view.
*/
hs = right - hspace;
} else if (x < left + hslack) {
/*
* If we are near the left edge, keep the left edge
* at the edge of the view.
*/
hs = left;
} else if (left > hs) {
/*
* Is there whitespace visible at the left? Fix it if so.
*/
hs = left;
} else if (right < hs + hspace) {
/*
* Is there whitespace visible at the right? Fix it if so.
*/
hs = right - hspace;
} else {
/*
* Otherwise, float as needed.
*/
if (x - hs < hslack) {
hs = x - hslack;
}
if (x - hs > hspace - hslack) {
hs = x - (hspace - hslack);
}
}
} if (hs != mScrollX || vs != mScrollY) {
if (mScroller == null) {
scrollTo(hs, vs);
} else {
long duration = AnimationUtils.currentAnimationTimeMillis() - mLastScroll;
int dx = hs - mScrollX;
int dy = vs - mScrollY; if (duration > ANIMATED_SCROLL_GAP) {
mScroller.startScroll(mScrollX, mScrollY, dx, dy);
awakenScrollBars(mScroller.getDuration());
invalidate();
} else {
if (!mScroller.isFinished()) {
mScroller.abortAnimation();
} scrollBy(dx, dy);
} mLastScroll = AnimationUtils.currentAnimationTimeMillis();
} changed = true;
} if (isFocused()) {
// This offsets because getInterestingRect() is in terms of viewport coordinates, but
// requestRectangleOnScreen() is in terms of content coordinates. // The offsets here are to ensure the rectangle we are using is
// within our view bounds, in case the cursor is on the far left
// or right. If it isn't withing the bounds, then this request
// will be ignored.
if (mTempRect == null) mTempRect = new Rect();
mTempRect.set(x - 2, top, x + 2, bottom);
getInterestingRect(mTempRect, line);
///M: ALPS00605613 requestRectangleOnScreen() will return error result if setting the mTempRect to mScrollX, mScrollY
//mTempRect.offset(mScrollX, mScrollY); if (requestRectangleOnScreen(mTempRect)) {
changed = true;
}
} return changed;
}
有点没看明白bringPointIntoView方法的作用是什么，还得再研究。不过这不影响我们分析布局过程。
二、结合自己写的一个小Demo继续分析，代码如下：
MyLinear2.java
public class MyLinear2 extends ViewGroup {
private static final String TAG = "David_MyLinear2";
public MyLinear2(Context context) {
super(context);
}
public MyLinear2(Context context, AttributeSet attrs) {
super(context, attrs);
}
@Override
protected void onLayout(boolean changed, int l, int t, int r, int b) {
int count = getChildCount();
int cWidth = 0;
int cHeight = 0;
int top = 0;
MarginLayoutParams params = null;
Log.e(TAG + " onLayout", "l = " + l);
Log.e(TAG + " onLayout", "t = " + t);
Log.e(TAG + " onLayout", "r = " + r);
Log.e(TAG + " onLayout", "b = " + b);
for (int i = 0; i < count; i++) {
Log.e(TAG + " onLayout", "====================i = " + i);
View childView = getChildAt(i);
cWidth = childView.getMeasuredWidth();
cHeight = childView.getMeasuredHeight();
params = (MarginLayoutParams) childView.getLayoutParams();
Log.e(TAG + " onLayout", "params.height = " + params.height);
Log.e(TAG + " onLayout", "params.width = " + params.width);
int leftMargin = params.leftMargin;
Log.e(TAG + " onLayout", "leftMargin = " + leftMargin);
Log.e(TAG + " onLayout", "params.rightMargin = " + params.rightMargin);
/*Log.e(TAG + " onLayout", "cWidth = " + cWidth);
Log.e(TAG + " onLayout", "cHeight = " + cHeight);
Log.e(TAG + " onLayout", "getWidth() = " + childView.getWidth());
Log.e(TAG + " onLayout", "getHeight() = " + childView.getHeight());*/
int cl = 0, ct = 0, cr = 0, cb = 0;
cl = leftMargin;
cr = cl + cWidth;
ct = top;
cb = cHeight + ct;
/*Log.e(TAG + " onLayout", "cr = " + cr);
Log.e(TAG + " onLayout", "ct = " + ct);
Log.e(TAG + " onLayout", "cb = " + cb);
Log.e(TAG + " onLayout", "top = " + top);*/
childView.layout(cl, ct, cr, cb);
top += cHeight;
}
}
@Override
public ViewGroup.LayoutParams generateLayoutParams(AttributeSet attrs) {
Log.e(TAG, "generateLayoutParams attrs");
return new MarginLayoutParams(getContext(), attrs);
}
@Override
protected ViewGroup.LayoutParams generateDefaultLayoutParams() {
Log.e(TAG, "generateDefaultLayoutParams");
return new LayoutParams(LayoutParams.MATCH_PARENT, LayoutParams.MATCH_PARENT);
}
@Override
protected boolean checkLayoutParams(LayoutParams p) {
return super.checkLayoutParams(p);
}
@Override
protected ViewGroup.LayoutParams generateLayoutParams(ViewGroup.LayoutParams p) {
Log.e(TAG, "generateLayoutParams p");
return new MarginLayoutParams(p);
}
@Override
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
int measuredHeight = measureHeight(heightMeasureSpec);
int measuredWidth = measureWidth(widthMeasureSpec);
Log.e(TAG, "onMeasure measuredHeight = " + measuredHeight);
Log.e(TAG, "onMeasure measuredWidth = " + measuredWidth);
setMeasuredDimension(measuredWidth, measuredHeight);
measureChildren(widthMeasureSpec, heightMeasureSpec);
}
private int measureHeight(int measureSpec) {
int specMode = MeasureSpec.getMode(measureSpec);
int specSize = MeasureSpec.getSize(measureSpec);
int result = 500;
if (specMode == MeasureSpec.AT_MOST){
result = specSize;
} else if (specMode == MeasureSpec.EXACTLY){
result = specSize;
}
return result;
}
private int measureWidth(int measureSpec) {
int specMode = MeasureSpec.getMode(measureSpec);
int specSize = MeasureSpec.getSize(measureSpec);
int result = 100;
if (specMode == MeasureSpec.AT_MOST){
result = specSize;
} else if (specMode == MeasureSpec.EXACTLY){
result = specSize;
}
return result;
}
}
MyLinear2时模仿纵向布局的LinearLayout。
MyTextView.java
public class MyTextView extends View {
private static final String TAG = "David___MyTextView";
public MyTextView(Context context, AttributeSet attrs) {
super(context, attrs);
}
public MyTextView(Context context) {
super(context);
}
@Override
protected void onDraw(Canvas canvas) {
super.onDraw(canvas);
Paint paint = new Paint();
paint.setTextSize(22);
paint.setTextAlign(Align.CENTER);
Log.e(TAG, "onDraw getTop() = " + getTop());
Log.e(TAG, "onDraw getLeft() = " + getLeft());
canvas.drawText("nihao ----", getTop(), getLeft(), paint);
}
@Override
protected void onLayout(boolean changed, int left, int top, int right, int bottom) {
Log.e(TAG, "onLayout left = " + left);
Log.e(TAG, "onLayout top = " + top);
Log.e(TAG, "onLayout right = " + right);
Log.e(TAG, "onLayout bottom = " + bottom);
//super.onLayout(changed, left, top, right, bottom);
}
@Override
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
int measuredHeight = measureHeight(heightMeasureSpec);
int measuredWidth = measureWidth(widthMeasureSpec);
Log.e(TAG, "onMeasure measuredHeight = " + measuredHeight);
Log.e(TAG, "onMeasure measuredWidth = " + measuredWidth);
setMeasuredDimension(measuredWidth, measuredHeight);
}
private int measureHeight(int measureSpec) {
int specMode = MeasureSpec.getMode(measureSpec);
int specSize = MeasureSpec.getSize(measureSpec);
int result = 500;
if (specMode == MeasureSpec.AT_MOST){
result = specSize;
} else if (specMode == MeasureSpec.EXACTLY){
result = specSize;
}
return result;
}
private int measureWidth(int measureSpec) {
int specMode = MeasureSpec.getMode(measureSpec);
int specSize = MeasureSpec.getSize(measureSpec);
int result = 500;
if (specMode == MeasureSpec.AT_MOST){
result = specSize;
} else if (specMode == MeasureSpec.EXACTLY){
result = specSize;
}
return result;
}
}
这是一个简易的View,先不要关心onDraw()方法。布局文件如下：
<com.test.touch.MyLinear2 xmlns:android="http://schemas.android.com/apk/res/android"
android:layout_width="300dp"
android:layout_height="200dp"
android:paddingTop="22dp"
android:layout_marginLeft="10dp" > <com.test.touch.MyTextView
android:layout_width="match_parent"
android:layout_height="120dp"
android:background="#0f0"
android:layout_marginLeft="30dp"
android:paddingTop="22dp"
android:text="Hello World" /> <TextView
android:id="@+id/tv1"
android:layout_width="match_parent"
android:layout_height="110dp"
android:background="#f00"
android:gravity="center_vertical"
android:layout_marginLeft="10dp"
android:text="Hello World Text" />
23 <com.test.touch.MyTextView
android:layout_width="match_parent"
android:layout_height="60dp"
android:background="#00f"
android:text="Hello World88" /> </com.test.touch.MyLinear2>
OK,运行打印的日志如下：
这段日志其实是会被打印好几遍的，我只截取了完整的一次日志。从日志可以看出，首先执行的是测量过程，这个之前分析过了，请参考http://www.cnblogs.com/wlrhnh/p/4680636.html。然后才是布局过程，由于我们在MyLinear2的onLayout()中显式调用childView的layout()方法，代码如下：
@Override
protected void onLayout(boolean changed, int l, int t, int r, int b) {
int count = getChildCount();
int cWidth = 0;
int cHeight = 0;
int top = 0;
MarginLayoutParams params = null;
Log.e(TAG + " onLayout", "l = " + l);
Log.e(TAG + " onLayout", "t = " + t);
Log.e(TAG + " onLayout", "r = " + r);
Log.e(TAG + " onLayout", "b = " + b); for (int i = 0; i < count; i++) {
Log.e(TAG + " onLayout", "====================i = " + i);
View childView = getChildAt(i);
cWidth = childView.getMeasuredWidth();
cHeight = childView.getMeasuredHeight();
params = (MarginLayoutParams) childView.getLayoutParams(); Log.e(TAG + " onLayout", "params.height = " + params.height);
Log.e(TAG + " onLayout", "params.width = " + params.width);
int leftMargin = params.leftMargin;
Log.e(TAG + " onLayout", "leftMargin = " + leftMargin);
Log.e(TAG + " onLayout", "params.rightMargin = " + params.rightMargin);
int cl = 0, ct = 0, cr = 0, cb = 0;
cl = leftMargin;
cr = cl + cWidth;
ct = top;
cb = cHeight + ct;
childView.layout(cl, ct, cr, cb);
top += cHeight;
}
}
所以日志中显示:先调用了ViewGroup的onLayout,然后遍历每一个childView,取出它们已经计算好的坐标值，按照ViewGroup的既定布局策略，给childView布局，调用他们的layout()方法，进而调用了onLayout()。而很多继承自View的组件，比如ImageView等，由于没有实现onLayout()方法，那么其实只调用了View的layout()和View的空实现的onLayout()方法。当然了，我在MyTextView中重写onLayout()只是为了打印日志而已，并没有做什么实际操作。在日志的最后，打印了onDraw()方法，可见，先测量、后布局、最后才统一draw，并不是布局完一个就draw一个。而且我发现一个有意思的现象，那就是如果你的父控件拥有的space不足以显示所有的子View，那么不能显示出来的子View的onDraw方法是不会被调用的，这点好理解～
三、关于LayoutParams
注意上面的一行红色代码，每个childView都有LayoutParams,而且可以强转为MarginLayoutParams呢？注意看这几行代码：
@Override
public ViewGroup.LayoutParams generateLayoutParams(AttributeSet attrs) {
Log.e(TAG, "generateLayoutParams attrs");
return new MarginLayoutParams(getContext(), attrs);
} @Override
protected ViewGroup.LayoutParams generateDefaultLayoutParams() {
Log.e(TAG, "generateDefaultLayoutParams");
return new LayoutParams(LayoutParams.MATCH_PARENT, LayoutParams.MATCH_PARENT);
} @Override
protected boolean checkLayoutParams(LayoutParams p) {
return super.checkLayoutParams(p);
} @Override
protected ViewGroup.LayoutParams generateLayoutParams(ViewGroup.LayoutParams p) {
Log.e(TAG, "generateLayoutParams p");
return new MarginLayoutParams(p);
}
看一下日志：
看黄色部分日志，主要执行了generateLayoutParams(AttributeSet attrs)方法，而且是先于测量过程执行的，这个方法我们有自己的实现，具体的请参考http://www.cnblogs.com/wlrhnh/p/4683542.html
简单的总结一下布局过程：
Android View绘制系统首先取到Activity布局的根View，当然这一般是一个ViewGroup了。先测量，后布局。布局的时候调用ViewGroup子类比如MyLinear2的layout()方法，由于ViewGroup强制子类实现onLayout()方法，所以会调到MyLinear2的onLayout()方法，在这个方法中，需要遍历子View，按照布局策略，计算每一个子View的坐标，然后将它放在合适的位置上。至于继承自View的子类，则不需要实现onLayout，毕竟onLayout的作是布局，这是容器类该干的事情。而子View要做的就是根据父View在指定给自己的空间中draw。
通过以上三篇文章，想必应该对Android系统测量、布局View的流程有了一个大概的了解，这是一条主线。当然了，这里面涉及到的细节其实有很多，最好的办法是自己去看源码喽～
项目代码和View.java ViewGroup.java LinearLayout.java Button.java等源码请下载
后记：分析完三步骤之后觉得不过瘾，想起网上有人写过这样的布局，就顺手写了个，抄袭人家的创意了，但是代码是自己的写的～
public class MyLinear1 extends ViewGroup {
private static final String TAG = "David_MyLinear1";
public MyLinear1(Context context) {
super(context);
}
public MyLinear1(Context context, AttributeSet attrs) {
super(context, attrs);
}
@Override
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
int measuredHeightSize = MeasureSpec.getSize(heightMeasureSpec);
int measuredWidthSize = MeasureSpec.getSize(widthMeasureSpec);
int measuredHeightMode = MeasureSpec.getMode(heightMeasureSpec);
int measuredWidthMode = MeasureSpec.getMode(widthMeasureSpec);
measureChildren(widthMeasureSpec, heightMeasureSpec);
int lHeight = 0, rHeight = 0;
int tWidth = 0, bWidth = 0;
View childView = null;
MarginLayoutParams params = null;
for (int i = 0; i < getChildCount(); i++) {
childView = getChildAt(i);
params = (MarginLayoutParams) childView.getLayoutParams();
switch (i) {
case 0:
lHeight += childView.getMeasuredHeight();
tWidth += childView.getMeasuredWidth();
break;
case 1:
rHeight += childView.getMeasuredHeight();
tWidth += childView.getMeasuredWidth();
break;
case 2:
lHeight += childView.getMeasuredHeight();
bWidth += childView.getMeasuredWidth();
break;
case 3:
rHeight += childView.getMeasuredHeight();
bWidth += childView.getMeasuredWidth();
break;
default:
break;
}
}
int realHeightSize = Math.max(lHeight, rHeight);
int realWidthSize = Math.max(bWidth, tWidth);
int h = measuredHeightMode == MeasureSpec.EXACTLY ? measuredHeightSize : realHeightSize;
int w = measuredWidthMode == MeasureSpec.EXACTLY ? measuredWidthSize : realWidthSize;
setMeasuredDimension(w, h);
}
@Override
protected void onLayout(boolean changed, int l, int t, int r, int b) {
int realLeft = 0;
int realTop = 0;
int realRight = 0;
int realBottom = 0;
int count = getChildCount();
View childView = null;
MarginLayoutParams params = null;
for (int i = 0; i < count; i++) {
Log.e(TAG, "------------------------- i = " + i);
childView = getChildAt(i);
params = (MarginLayoutParams) childView.getLayoutParams();
Log.d(TAG, "childView.getMeasuredWidth() = " + childView.getMeasuredWidth());
Log.d(TAG, "childView.getMeasuredHeight() = " + childView.getMeasuredHeight());
switch (i) {
case 0:
realLeft = l;
realTop = t;
realRight = realLeft + childView.getMeasuredWidth();
realBottom = realTop + childView.getMeasuredHeight();
break;
case 1:
realLeft = r - childView.getMeasuredWidth();
realTop = t;
realRight = realLeft + childView.getMeasuredWidth();
realBottom = realTop + childView.getMeasuredHeight();
break;
case 2:
realLeft = l;
realTop = b - childView.getMeasuredHeight();
realRight = realLeft + childView.getMeasuredWidth();
realBottom = realTop + childView.getMeasuredHeight();
break;
case 3:
realLeft = r - childView.getMeasuredWidth();
realTop = b - childView.getMeasuredHeight();
realRight = realLeft + childView.getMeasuredWidth();
realBottom = realTop + childView.getMeasuredHeight();
break;
default:
break;
}
childView.layout(realLeft, realTop, realRight, realBottom);
}
}
@Override
public ViewGroup.LayoutParams generateLayoutParams(AttributeSet attrs) {
return new MarginLayoutParams(getContext(), attrs);
}
}
代码随手写的，有些问题考虑的还不够周全，各位自己完善吧～而且没有做注释，但是相信看完这三篇文章，不需要看注释了吧～