Overview
RelativeLayout是Android中一种常用的布局，通过View之间的相互关系能够方便地确定子View的位置，通过源码的学习来看下Android是如何实现这一功能的。Android中的一个组件，工作原理主要在onMeasure(),onLayout(),onDraw()这3大阶段中。
onMeasure()
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
if (mDirtyHierarchy) {
mDirtyHierarchy = false;
sortChildren();
}
// ……
}
测量过程的第一步是对子View进行排序，在这一步骤会设置一个标志位mDirtyHierarchy进行优化，下面看下sortChildren的具体实现：
private void sortChildren() {
final int count = getChildCount();
if (mSortedVerticalChildren == null || mSortedVerticalChildren.length != count) {
mSortedVerticalChildren = new View[count];
}
if (mSortedHorizontalChildren == null || mSortedHorizontalChildren.length != count) {
mSortedHorizontalChildren = new View[count];
}
final DependencyGraph graph = mGraph;
graph.clear();
for (int i = 0; i < count; i++) {
graph.add(getChildAt(i));
}
graph.getSortedViews(mSortedVerticalChildren, RULES_VERTICAL);
graph.getSortedViews(mSortedHorizontalChildren, RULES_HORIZONTAL);
}
这个函数包含以下步骤：
- 初始化两个数组，分别用于记录水平方向和竖直方向的排序结果
- 初始化一个依赖图DependencyGraph，将子View都加入到依赖图中
- 调用DependencyGraph的getSortedViews方法，将两个方向的排序结果记录到对应数组中
下面首先看下DependencyGraph的结构：
private static class DependencyGraph {
/**
* List of all views in the graph.
*/
private ArrayList<Node> mNodes = new ArrayList<Node>();
/**
* List of nodes in the graph. Each node is identified by its
* view id (see View#getId()).
*/
private SparseArray<Node> mKeyNodes = new SparseArray<Node>();
/**
* Temporary data structure used to build the list of roots
* for this graph.
*/
private ArrayDeque<Node> mRoots = new ArrayDeque<Node>();
}
结果里每个成员代表的含义，其注释已经说清楚了；再具体看下Node的结构：
static class Node {
/**
* The view representing this node in the layout.
*/
View view;
/**
* The list of dependents for this node; a dependent is a node
* that needs this node to be processed first.
*/
final ArrayMap<Node, DependencyGraph> dependents =
new ArrayMap<Node, DependencyGraph>();
/**
* The list of dependencies for this node.
*/
final SparseArray<Node> dependencies = new SparseArray<Node>();
}
根据注释也能够明白每个成员的含义
接下来看下DependencyGraph是如何实现getSortedViews方法的：
void getSortedViews(View[] sorted, int... rules) {
// 找出所有的不依赖其他节点的节点，即根节点
final ArrayDeque<Node> roots = findRoots(rules);
int index = 0;
Node node;
// 循环从根节点的队尾中取出元素
while ((node = roots.pollLast()) != null) {
final View view = node.view;
final int key = view.getId();
// 将该节点加入排序结果
sorted[index++] = view;
// 取出该节点的所有伴随节点
final ArrayMap<Node, DependencyGraph> dependents = node.dependents;
final int count = dependents.size();
for (int i = 0; i < count; i++) {
// 遍历伴随节点
final Node dependent = dependents.keyAt(i);
// 取出伴随节点的所有依赖节点集合
final SparseArray<Node> dependencies = dependent.dependencies;
// 将当前节点从伴随节点的依赖节点集合中去掉
dependencies.remove(key);
// 如果该伴随节点去掉当前节点的依赖之后，没有再依赖其他节点，证明它是新的根节点，加入根节点集合
if (dependencies.size() == 0) {
roots.add(dependent);
}
}
}
if (index < sorted.length) {
throw new IllegalStateException("Circular dependencies cannot exist"
+ " in RelativeLayout");
}
}
这个排序的过程已经在注释中说明，实现过程还是挺明确的。
再看完排序过程，我们回到onMeasure()方法中
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
// ……
int myWidth = -1;
int myHeight = -1;
int width = 0;
int height = 0;
final int widthMode = MeasureSpec.getMode(widthMeasureSpec);
final int heightMode = MeasureSpec.getMode(heightMeasureSpec);
final int widthSize = MeasureSpec.getSize(widthMeasureSpec);
final int heightSize = MeasureSpec.getSize(heightMeasureSpec);
// Record our dimensions if they are known;
if (widthMode != MeasureSpec.UNSPECIFIED) {
myWidth = widthSize;
}
if (heightMode != MeasureSpec.UNSPECIFIED) {
myHeight = heightSize;
}
if (widthMode == MeasureSpec.EXACTLY) {
width = myWidth;
}
if (heightMode == MeasureSpec.EXACTLY) {
height = myHeight;
}
View ignore = null;
int gravity = mGravity & Gravity.RELATIVE_HORIZONTAL_GRAVITY_MASK;
final boolean horizontalGravity = gravity != Gravity.START && gravity != 0;
gravity = mGravity & Gravity.VERTICAL_GRAVITY_MASK;
final boolean verticalGravity = gravity != Gravity.TOP && gravity != 0;
int left = Integer.MAX_VALUE;
int top = Integer.MAX_VALUE;
int right = Integer.MIN_VALUE;
int bottom = Integer.MIN_VALUE;
boolean offsetHorizontalAxis = false;
boolean offsetVerticalAxis = false;
if ((horizontalGravity || verticalGravity) && mIgnoreGravity != View.NO_ID) {
ignore = findViewById(mIgnoreGravity);
}
final boolean isWrapContentWidth = widthMode != MeasureSpec.EXACTLY;
final boolean isWrapContentHeight = heightMode != MeasureSpec.EXACTLY;
// We need to know our size for doing the correct computation of children positioning in RTL
// mode but there is no practical way to get it instead of running the code below.
// So, instead of running the code twice, we just set the width to a "default display width"
// before the computation and then, as a last pass, we will update their real position with
// an offset equals to "DEFAULT_WIDTH - width".
final int layoutDirection = getLayoutDirection();
if (isLayoutRtl() && myWidth == -1) {
myWidth = DEFAULT_WIDTH;
}
// ……
}
中间这段是一些辅助变量的初始化，就不多说了，继续往下看
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
// ……
View[] views = mSortedHorizontalChildren;
int count = views.length;
for (int i = 0; i < count; i++) {
View child = views[i];
if (child.getVisibility() != GONE) {
LayoutParams params = (LayoutParams) child.getLayoutParams();
int[] rules = params.getRules(layoutDirection);
applyHorizontalSizeRules(params, myWidth, rules);
measureChildHorizontal(child, params, myWidth, myHeight);
if (positionChildHorizontal(child, params, myWidth, isWrapContentWidth)) {
offsetHorizontalAxis = true;
}
}
}
// ……
}
测量过程的第二步是遍历子View进行水平方向的依赖测量，这一步通过3个函数进行，逐个看下函数的实现。
介绍函数实现之前先介绍一下rule的概念，在RelativeLayout.LayoutParams进行解析时，会将RelativeLayout中的子View的依赖关系通过rules数组存储起来，数组中的下标代表某条规则，例如LEFT_OF，对应的元素代表这条规则依赖的View id：
switch (attr) {
case com.android.internal.R.styleable.RelativeLayout_Layout_layout_toLeftOf:
rules[LEFT_OF] = a.getResourceId(attr, 0);
break;
}
了解了rule的概念之后，看下遍历子View进行依赖测量时相关方法的实现：
private void applyHorizontalSizeRules(LayoutParams childParams, int myWidth, int[] rules) {
RelativeLayout.LayoutParams anchorParams;
// VALUE_NOT_SET indicates a "soft requirement" in that direction. For example:
// left=10, right=VALUE_NOT_SET means the view must start at 10, but can go as far as it
// wants to the right
// left=VALUE_NOT_SET, right=10 means the view must end at 10, but can go as far as it
// wants to the left
// left=10, right=20 means the left and right ends are both fixed
childParams.mLeft = VALUE_NOT_SET;
childParams.mRight = VALUE_NOT_SET;
anchorParams = getRelatedViewParams(rules, LEFT_OF);
if (anchorParams != null) {
childParams.mRight = anchorParams.mLeft - (anchorParams.leftMargin +
childParams.rightMargin);
} else if (childParams.alignWithParent && rules[LEFT_OF] != 0) {
if (myWidth >= 0) {
childParams.mRight = myWidth - mPaddingRight - childParams.rightMargin;
}
}
anchorParams = getRelatedViewParams(rules, RIGHT_OF);
if (anchorParams != null) {
childParams.mLeft = anchorParams.mRight + (anchorParams.rightMargin +
childParams.leftMargin);
} else if (childParams.alignWithParent && rules[RIGHT_OF] != 0) {
childParams.mLeft = mPaddingLeft + childParams.leftMargin;
}
anchorParams = getRelatedViewParams(rules, ALIGN_LEFT);
if (anchorParams != null) {
childParams.mLeft = anchorParams.mLeft + childParams.leftMargin;
} else if (childParams.alignWithParent && rules[ALIGN_LEFT] != 0) {
childParams.mLeft = mPaddingLeft + childParams.leftMargin;
}
anchorParams = getRelatedViewParams(rules, ALIGN_RIGHT);
if (anchorParams != null) {
childParams.mRight = anchorParams.mRight - childParams.rightMargin;
} else if (childParams.alignWithParent && rules[ALIGN_RIGHT] != 0) {
if (myWidth >= 0) {
childParams.mRight = myWidth - mPaddingRight - childParams.rightMargin;
}
}
if (0 != rules[ALIGN_PARENT_LEFT]) {
childParams.mLeft = mPaddingLeft + childParams.leftMargin;
}
if (0 != rules[ALIGN_PARENT_RIGHT]) {
if (myWidth >= 0) {
childParams.mRight = myWidth - mPaddingRight - childParams.rightMargin;
}
}
}
纵观整个函数，其实就是对水平方向的几条规则逐个进行处理，这几条规则分别是：LEFT_OF/RIGHT_OF/ALIGN_LEFT/ALIGN_RIGHT/ALIGN_PARENT_LEFT/ALIGN_PARENT_RIGHT；
对于每个属性的处理方法都是相似的：通过相应的规则从getRelativeViewParams中取到对应依赖的View，然后通过依赖的View的对应属性和padding来调整该View的对应属性
看看下getRelativeViewParams的实现：
private LayoutParams getRelatedViewParams(int[] rules, int relation) {
View v = getRelatedView(rules, relation);
if (v != null) {
ViewGroup.LayoutParams params = v.getLayoutParams();
if (params instanceof LayoutParams) {
return (LayoutParams) v.getLayoutParams();
}
}
return null;
}
真正的实现在getRelatedView中
private View getRelatedView(int[] rules, int relation) {
int id = rules[relation];
if (id != 0) {
DependencyGraph.Node node = mGraph.mKeyNodes.get(id);
if (node == null) return null;
View v = node.view;
// Find the first non-GONE view up the chain
while (v.getVisibility() == View.GONE) {
rules = ((LayoutParams) v.getLayoutParams()).getRules(v.getLayoutDirection());
node = mGraph.mKeyNodes.get((rules[relation]));
// ignore self dependency. for more info look in git commit: da3003
if (node == null || v == node.view) return null;
v = node.view;
}
return v;
}
return null;
}
关键点：
- 通过rules数组取到对应依赖的View id
- 通过View id从依赖图中取出对应的View
- 如果取到的View可见性为GONE，则继续通过对应的依赖规则取依赖的View所依赖的View(有点拗口...)
经过applyHorizontalSizeRules方法的处理之后，子View的params中的left/right值都已经确定好，然后会进入measureChildHorizontal方法
private void measureChildHorizontal(
View child, LayoutParams params, int myWidth, int myHeight) {
final int childWidthMeasureSpec = getChildMeasureSpec(params.mLeft, params.mRight,
params.width, params.leftMargin, params.rightMargin, mPaddingLeft, mPaddingRight,
myWidth);
final int childHeightMeasureSpec;
if (myHeight < 0 && !mAllowBrokenMeasureSpecs) {
if (params.height >= 0) {
childHeightMeasureSpec = MeasureSpec.makeMeasureSpec(
params.height, MeasureSpec.EXACTLY);
} else {
// Negative values in a mySize/myWidth/myWidth value in
// RelativeLayout measurement is code for, "we got an
// unspecified mode in the RelativeLayout's measure spec."
// Carry it forward.
childHeightMeasureSpec = MeasureSpec.makeMeasureSpec(0, MeasureSpec.UNSPECIFIED);
}
} else {
final int maxHeight;
if (mMeasureVerticalWithPaddingMargin) {
maxHeight = Math.max(0, myHeight - mPaddingTop - mPaddingBottom
- params.topMargin - params.bottomMargin);
} else {
maxHeight = Math.max(0, myHeight);
}
final int heightMode;
if (params.height == LayoutParams.MATCH_PARENT) {
heightMode = MeasureSpec.EXACTLY;
} else {
heightMode = MeasureSpec.AT_MOST;
}
childHeightMeasureSpec = MeasureSpec.makeMeasureSpec(maxHeight, heightMode);
}
child.measure(childWidthMeasureSpec, childHeightMeasureSpec);
}
在上一阶段我们确定了子View的left/right参数，在这一阶段，可以通过left/right得到子View的宽度，然后进行第一次子View的measure方法调用，进行第一次测量
接下来调用positionChildHorizontal方法：
private boolean positionChildHorizontal(View child, LayoutParams params, int myWidth,
boolean wrapContent) {
final int layoutDirection = getLayoutDirection();
int[] rules = params.getRules(layoutDirection);
if (params.mLeft == VALUE_NOT_SET && params.mRight != VALUE_NOT_SET) {
// Right is fixed, but left varies
params.mLeft = params.mRight - child.getMeasuredWidth();
} else if (params.mLeft != VALUE_NOT_SET && params.mRight == VALUE_NOT_SET) {
// Left is fixed, but right varies
params.mRight = params.mLeft + child.getMeasuredWidth();
} else if (params.mLeft == VALUE_NOT_SET && params.mRight == VALUE_NOT_SET) {
// Both left and right vary
if (rules[CENTER_IN_PARENT] != 0 || rules[CENTER_HORIZONTAL] != 0) {
if (!wrapContent) {
centerHorizontal(child, params, myWidth);
} else {
positionAtEdge(child, params, myWidth);
}
return true;
} else {
// This is the default case. For RTL we start from the right and for LTR we start
// from the left. This will give LEFT/TOP for LTR and RIGHT/TOP for RTL.
positionAtEdge(child, params, myWidth);
}
}
return rules[ALIGN_PARENT_END] != 0;
}
这个方法的代码不长，从实现中可以得知，该方法的作用在于，那些在前阶段没有设置的左右边界约束。
接下来是第三步，遍历子View进行垂直方向的依赖测量
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
// ……
for (int i = 0; i < count; i++) {
final View child = views[i];
if (child.getVisibility() != GONE) {
final LayoutParams params = (LayoutParams) child.getLayoutParams();
applyVerticalSizeRules(params, myHeight, child.getBaseline());
measureChild(child, params, myWidth, myHeight);
if (positionChildVertical(child, params, myHeight, isWrapContentHeight)) {
offsetVerticalAxis = true;
}
if (isWrapContentWidth) {
if (isLayoutRtl()) {
if (targetSdkVersion < Build.VERSION_CODES.KITKAT) {
width = Math.max(width, myWidth - params.mLeft);
} else {
width = Math.max(width, myWidth - params.mLeft + params.leftMargin);
}
} else {
if (targetSdkVersion < Build.VERSION_CODES.KITKAT) {
width = Math.max(width, params.mRight);
} else {
width = Math.max(width, params.mRight + params.rightMargin);
}
}
}
if (isWrapContentHeight) {
if (targetSdkVersion < Build.VERSION_CODES.KITKAT) {
height = Math.max(height, params.mBottom);
} else {
height = Math.max(height, params.mBottom + params.bottomMargin);
}
}
if (child != ignore || verticalGravity) {
left = Math.min(left, params.mLeft - params.leftMargin);
top = Math.min(top, params.mTop - params.topMargin);
}
if (child != ignore || horizontalGravity) {
right = Math.max(right, params.mRight + params.rightMargin);
bottom = Math.max(bottom, params.mBottom + params.bottomMargin);
}
}
}
// ……
}
垂直方向的约束测量过程和水平方向是一个套路：
- 遍历子View，根据垂直方向的规则进行top/bottom的约束
- 测量一次子View
- 补充边界约束
- 不同的是，在垂直方向的测量阶段，还会对RelativeLayout的最大宽度和最大高度进行计算
经过水平和垂直方向两轮对子View的分发测量，已经完成了子View的测量过程，接下来是根据是否wrapContent调整RelativeLayout自身的高度
if (isWrapContentWidth) {
// Width already has left padding in it since it was calculated by looking at
// the right of each child view
width += mPaddingRight;
if (mLayoutParams != null && mLayoutParams.width >= 0) {
width = Math.max(width, mLayoutParams.width);
}
width = Math.max(width, getSuggestedMinimumWidth());
width = resolveSize(width, widthMeasureSpec);
if (offsetHorizontalAxis) {
for (int i = 0; i < count; i++) {
final View child = views[i];
if (child.getVisibility() != GONE) {
final LayoutParams params = (LayoutParams) child.getLayoutParams();
final int[] rules = params.getRules(layoutDirection);
if (rules[CENTER_IN_PARENT] != 0 || rules[CENTER_HORIZONTAL] != 0) {
centerHorizontal(child, params, width);
} else if (rules[ALIGN_PARENT_RIGHT] != 0) {
final int childWidth = child.getMeasuredWidth();
params.mLeft = width - mPaddingRight - childWidth;
params.mRight = params.mLeft + childWidth;
}
}
}
}
}
if (isWrapContentHeight) {
// Height already has top padding in it since it was calculated by looking at
// the bottom of each child view
height += mPaddingBottom;
if (mLayoutParams != null && mLayoutParams.height >= 0) {
height = Math.max(height, mLayoutParams.height);
}
height = Math.max(height, getSuggestedMinimumHeight());
height = resolveSize(height, heightMeasureSpec);
if (offsetVerticalAxis) {
for (int i = 0; i < count; i++) {
final View child = views[i];
if (child.getVisibility() != GONE) {
final LayoutParams params = (LayoutParams) child.getLayoutParams();
final int[] rules = params.getRules(layoutDirection);
if (rules[CENTER_IN_PARENT] != 0 || rules[CENTER_VERTICAL] != 0) {
centerVertical(child, params, height);
} else if (rules[ALIGN_PARENT_BOTTOM] != 0) {
final int childHeight = child.getMeasuredHeight();
params.mTop = height - mPaddingBottom - childHeight;
params.mBottom = params.mTop + childHeight;
}
}
}
}
}
然后再根据Gravity属性调整位置
if (isWrapContentWidth) {
// Width already has left padding in it since it was calculated by looking at
// the right of each child view
width += mPaddingRight;
if (mLayoutParams != null && mLayoutParams.width >= 0) {
width = Math.max(width, mLayoutParams.width);
}
width = Math.max(width, getSuggestedMinimumWidth());
width = resolveSize(width, widthMeasureSpec);
if (offsetHorizontalAxis) {
for (int i = 0; i < count; i++) {
final View child = views[i];
if (child.getVisibility() != GONE) {
final LayoutParams params = (LayoutParams) child.getLayoutParams();
final int[] rules = params.getRules(layoutDirection);
if (rules[CENTER_IN_PARENT] != 0 || rules[CENTER_HORIZONTAL] != 0) {
centerHorizontal(child, params, width);
} else if (rules[ALIGN_PARENT_RIGHT] != 0) {
final int childWidth = child.getMeasuredWidth();
params.mLeft = width - mPaddingRight - childWidth;
params.mRight = params.mLeft + childWidth;
}
}
}
}
}
if (isWrapContentHeight) {
// Height already has top padding in it since it was calculated by looking at
// the bottom of each child view
height += mPaddingBottom;
if (mLayoutParams != null && mLayoutParams.height >= 0) {
height = Math.max(height, mLayoutParams.height);
}
height = Math.max(height, getSuggestedMinimumHeight());
height = resolveSize(height, heightMeasureSpec);
if (offsetVerticalAxis) {
for (int i = 0; i < count; i++) {
final View child = views[i];
if (child.getVisibility() != GONE) {
final LayoutParams params = (LayoutParams) child.getLayoutParams();
final int[] rules = params.getRules(layoutDirection);
if (rules[CENTER_IN_PARENT] != 0 || rules[CENTER_VERTICAL] != 0) {
centerVertical(child, params, height);
} else if (rules[ALIGN_PARENT_BOTTOM] != 0) {
final int childHeight = child.getMeasuredHeight();
params.mTop = height - mPaddingBottom - childHeight;
params.mBottom = params.mTop + childHeight;
}
}
}
}
}
最后就是设置自身的宽高，完成测量过程
if (isLayoutRtl()) {
final int offsetWidth = myWidth - width;
for (int i = 0; i < count; i++) {
final View child = views[i];
if (child.getVisibility() != GONE) {
final LayoutParams params = (LayoutParams) child.getLayoutParams();
params.mLeft -= offsetWidth;
params.mRight -= offsetWidth;
}
}
}
setMeasuredDimension(width, height);
上述就是onMeasure的过程。
onLayout
protected void onLayout(boolean changed, int l, int t, int r, int b) {
// The layout has actually already been performed and the positions
// cached. Apply the cached values to the children.
final int count = getChildCount();
for (int i = 0; i < count; i++) {
View child = getChildAt(i);
if (child.getVisibility() != GONE) {
RelativeLayout.LayoutParams st =
(RelativeLayout.LayoutParams) child.getLayoutParams();
child.layout(st.mLeft, st.mTop, st.mRight, st.mBottom);
}
}
}
onLayout的过程很简单，遍历子View，因为子View在分发测量过程中已经确定了上下左右的边界约束，所以直接调用对应的约束值进行layout。
onDraw
在RelativeLayout中，没有对onDraw进行重写。
总结
和大多数容器布局类似，RelativeLayout的核心在于测量阶段，需要遍历两次子View，原因是水平和垂直方向各需一次来确定相应方向的边界约束。
