前言
- React的diff算法是在render的beginWork阶段中进行处理
beginWork
是在向下深度遍历fiber树时会对途径的每个节点进行状态处理和进行diff对比-
首先diff的入口是在
reconcileChildFibers
中,然后会根据type来判断使用哪种diff函数进行处理function reconcileChildFibers( returnFiber: Fiber, currentFirstChild: Fiber | null, newChild: any, lanes: Lanes, ): Fiber | null { if (typeof newChild === 'object' && newChild !== null) { switch (newChild.$$typeof) { case REACT_ELEMENT_TYPE: return placeSingleChild( reconcileSingleElement( returnFiber, currentFirstChild, newChild, lanes, ), ); case REACT_PORTAL_TYPE: // ... case REACT_LAZY_TYPE: //... } if (isArray(newChild)) { return reconcileChildrenArray( returnFiber, currentFirstChild, newChild, lanes, ); } if (getIteratorFn(newChild)) { //... } } // ... }
-
我在本篇会针对两种较常用的diff函数进行分析
reconcileSingleElement
reconcileChildrenArray
reconcileSingleElement
-
reconcileSingleElement是针对新newChild是单节点,而oldChild单节点或者是多节点就无法确定了,所以在此diff算法中就会对旧节点进行遍历,然后删除不匹配的oldFiber
function reconcileSingleElement( returnFiber: Fiber, currentFirstChild: Fiber | null, element: ReactElement lanes: Lanes, ): Fiber { const key = element.key; let child = currentFirstChild; /** * 遍历旧节点,找到与newChild相同key的节点,不匹配的删除 * 针对匹配的oldFiber, 用newChild中新节点的props来生成新的fiber节点 */ while (child !== null) { if (child.key === key) { const elementType = element.type; /** * 通过useFiber创建一个新的Fiber * 如果element是一个Fragment,则以element.props.children建立Fiber * 将returnFiber赋给新的fiber的return字段,然后返回这个新的fiber */· if (elementType === REACT_FRAGMENT_TYPE) { if (child.tag === Fragment) { deleteRemainingChildren(returnFiber, child.sibling); const existing = useFiber(child, element.props.children); existing.return = returnFiber; if (__DEV__) { existing._debugSource = element._source; existing._debugOwner = element._owner; } return existing; } } else { if ( child.elementType === elementType || (__DEV__ ? isCompatibleFamilyForHotReloading(child, element) : false) || (typeof elementType === 'object' && elementType !== null && elementType.$$typeof === REACT_LAZY_TYPE && resolveLazy(elementType) === child.type) ) { deleteRemainingChildren(returnFiber, child.sibling); const existing = useFiber(child, element.props); existing.ref = coerceRef(returnFiber, child, element); existing.return = returnFiber; if (__DEV__) { existing._debugSource = element._source; existing._debugOwner = element._owner; } return existing; } } // Didn't match. deleteRemainingChildren(returnFiber, child); break; } else { // key不相同就删除 deleteChild(returnFiber, child); } child = child.sibling; } // 如果没有命中一个key,则通过createFiberFormElement或CreateFiberFormFragment创建一个新的fiber,然后返回 if (element.type === REACT_FRAGMENT_TYPE) { const created = createFiberFromFragment( element.props.children, returnFiber.mode, lanes, element.key, ); created.return = returnFiber; return created; } else { const created = createFiberFromElement(element, returnFiber.mode, lanes); created.ref = coerceRef(returnFiber, currentFirstChild, element); created.return = returnFiber; return created; } }
reconcileChildrenArray
- 针对
newChild
是多节点的情况就需要调用reconcileChildrenArray
进行diff操作 - 多节点会有四种可能性的变化:删除、新增、位移、更新
reconcileChildrenArray
针对这四种变化,首先会处理的是更新,当出现无法匹配的情况时,就会根据遍历的情况来判断是否处理删除或者新增,然后最后会根据情况处理位移- 因为fiber是单向链表,所以
reconcileChildrenArray
的遍历不是双端遍历 -
首先第一轮遍历,是处理节点更新
for (; oldFiber !== null && newIdx < newChildren.length; newIdx++) { // newChildren遍历完了,oldFiber没有遍历完,中断遍历 if (oldFiber.index > newIdx) { nextOldFiber = oldFiber; oldFiber = null; } else { // 记录oldFiber的下一个节点 nextOldFiber = oldFiber.sibling; } // 更新节点,如果节点没有匹配上,就会返回null const newFiber = updateSlot( returnFiber, oldFiber, newChildren[newIdx], lanes, ); // newFiber为null说明节点没有匹配上,中断遍历 if (newFiber === null) { // oldFiber为null说明oldFiber也遍历完了 if (oldFiber === null) { oldFiber = nextOldFiber; } break; } /** * shouldTrackSideEffects为true表示是更新过程 * mountChildFibers = ChildReconciler(false); * reconcileChildFibers = ChildReconciler(true); * ChildReconciler接收的就是shouldTrackSideEffects */ if (shouldTrackSideEffects) { if (oldFiber && newFiber.alternate === null) { // 新节点没有现有节点,需要删除 deleteChild(returnFiber, oldFiber); } } // 记录固定节点的位置 lastPlacedIndex = placeChild(newFiber, lastPlacedIndex, newIdx); // 将新节点拼接成以sibling为指针的单向链表 if (previousNewFiber === null) { resultingFirstChild = newFiber; } else { previousNewFiber.sibling = newFiber; } previousNewFiber = newFiber; oldFiber = nextOldFiber; }
-
遍历完匹配的节点后,就判断新节点是否遍历完,如果遍历完,那么剩余的oldFiber都是要删除的
if (newIdx === newChildren.length) { deleteRemainingChildren(returnFiber, oldFiber); if (getIsHydrating()) { const numberOfForks = newIdx; pushTreeFork(returnFiber, numberOfForks); } return resultingFirstChild; }
-
如果新旧点没有遍历完,就判断旧fiber链是否遍历完,如果遍历完那么剩余的新节点全部作为新fiber插入
if (oldFiber === null) { for (; newIdx < newChildren.length; newIdx++) { // 创建新fiber节点 const newFiber = createChild(returnFiber, newChildren[newIdx], lanes); if (newFiber === null) { continue; } // 记录固定节点 lastPlacedIndex = placeChild(newFiber, lastPlacedIndex, newIdx); // 将新fiber拼接成以sibling为指针的单向链表 if (previousNewFiber === null) { resultingFirstChild = newFiber; } else { previousNewFiber.sibling = newFiber; } previousNewFiber = newFiber; } if (getIsHydrating()) { const numberOfForks = newIdx; pushTreeFork(returnFiber, numberOfForks); } return resultingFirstChild; }
-
执行到这一步,说明新旧节点都没有遍历完,就说明存在有位移的未知序列
// 首先创建一个以oldFiber key为键,值为oldFiber的map const existingChildren = mapRemainingChildren(returnFiber, oldFiber); for (; newIdx < newChildren.length; newIdx++) { // 然后根据map中的oldFiber创建新fiber const newFiber = updateFromMap( existingChildren, returnFiber, newIdx, newChildren[newIdx], lanes, ); if (newFiber !== null) { if (shouldTrackSideEffects) { if (newFiber.alternate !== null) { // 如果newFiber.alternate不为null,说明是根据oldFiber创建的,那么就需要在map中删除oldFiber existingChildren.delete( newFiber.key === null ? newIdx : newFiber.key, ); } } // 根据lastPlacedIndex判断是否移动节点 lastPlacedIndex = placeChild(newFiber, lastPlacedIndex, newIdx); // 将新fiber拼接成以sibling为指针的单向链表 if (previousNewFiber === null) { resultingFirstChild = newFiber; } else { previousNewFiber.sibling = newFiber; } previousNewFiber = newFiber; } } if (shouldTrackSideEffects) { // 删除剩余的oldFiber existingChildren.forEach(child => deleteChild(returnFiber, child)); }
placeChild
-
移动节点的核心是在
placeChild
这个函数中,如果当前正在遍历的节点的oldIndex是在lastPlacedIndex
的右边,就说明它的位置没变化,因为旧节点中就处于右边,新节点中也处于右边。- 例如:old:A -> B -> C -> D,new:D -> A -> B -> C
- 遍历到D时,
lastPlacedIndex = D的oldIndex = 3
-
然后遍历到A时,A的
oldIndex
为0,小于3,说明A在旧序列中肯定不是D的右边,所以A肯定产生了位移function placeChild( newFiber: Fiber, lastPlacedIndex: number, newIndex: number, ): number { newFiber.index = newIndex; if (!shouldTrackSideEffects) { newFiber.flags |= Forked; return lastPlacedIndex; } const current = newFiber.alternate; if (current !== null) { const oldIndex = current.index; if (oldIndex < lastPlacedIndex) { // 小于lastPlacedIndex 产生了位移 newFiber.flags |= Placement | PlacementDEV; return lastPlacedIndex; } else { // 没有位移,返回当前的oldIndex return oldIndex; } } else { newFiber.flags |= Placement | PlacementDEV; return lastPlacedIndex; } }
总结
- 针对单节点的diff,会遍历oldFiber链,如果有匹配的fiber,就以匹配的生成新fiber,如果没有就新建一个fiber,然后删除不匹配的fiber
-
针对多节点diff
- 首先是从头向尾遍历,针对复用的fiber进行更新,如果无法复用就中断遍历
- 然后判断新旧节点的遍历情况,来判断是否新增或者删除
- 如果都没有遍历完,就创建一个map
Map<old key, old Fiber>
,然后遍历新节点,基于map来创建新fiber,然后根据lastPlacedIndex
来判断是否产生了位移,遍历完最后删除剩余的oldFiber
相关文章
暂无评论...