How to use diff algorithm

This time I will show you how to use the diff algorithm, what are the precautions when using the diff algorithm, the following is a practical case, let's take a look.

Virtual dom

The diff algorithm must first clarify the concept that the object of diff is the virtual dom, and updating the real dom is the result of the diff algorithm

Vnode base class

 constructor (
  。。。
 ) {
  this.tag = tag
  this.data = data
  this.children = children
  this.text = text
  this.elm = elm
  this.ns = undefined
  this.context = context
  this.fnContext = undefined
  this.fnOptions = undefined
  this.fnScopeId = undefined
  this.key = data && data.key
  this.componentOptions = componentOptions
  this.componentInstance = undefined
  this.parent = undefined
  this.raw = false
  this.isStatic = false
  this.isRootInsert = true
  this.isComment = false
  this.isCloned = false
  this.isOnce = false
  this.asyncFactory = asyncFactory
  this.asyncMeta = undefined
  this.isAsyncPlaceholder = false
 }

This part of the code is mainly to better understand the meaning of the attribute of the specific diff in the diff algorithm. Of course, it can also be better understood Understand the vnode instance

The overall process

The core function is the patch function

• isUndef judgment (whether it is undefined or null)

• // empty mount (likely as component), create new root elementcreateElm(vnode, insertedVnodeQueue) Here you can find that creating nodes is not inserted one by one, but put into a queue for unified batch processing

• Core function sameVnode

function sameVnode (a, b) {
 return (
  a.key === b.key && (
   (
    a.tag === b.tag &&
    a.isComment === b.isComment &&
    isDef(a.data) === isDef(b.data) &&
    sameInputType(a, b)
   ) || (
    isTrue(a.isAsyncPlaceholder) &&
    a.asyncFactory === b.asyncFactory &&
    isUndef(b.asyncFactory.error)
   )
  )
 )
}

Here is an outer comparison function that directly compares the keys and tags of the two nodes. ), compare data (note that data here refers to VNodeData), and directly compare types for input.

export interface VNodeData {
 key?: string | number;
 slot?: string;
 scopedSlots?: { [key: string]: ScopedSlot };
 ref?: string;
 tag?: string;
 staticClass?: string;
 class?: any;
 staticStyle?: { [key: string]: any };
 style?: object[] | object;
 props?: { [key: string]: any };
 attrs?: { [key: string]: any };
 domProps?: { [key: string]: any };
 hook?: { [key: string]: Function };
 on?: { [key: string]: Function | Function[] };
 nativeOn?: { [key: string]: Function | Function[] };
 transition?: object;
 show?: boolean;
 inlineTemplate?: {
  render: Function;
  staticRenderFns: Function[];
 };
 directives?: VNodeDirective[];
 keepAlive?: boolean;
}

This will confirm whether the two nodes have further comparison value, otherwise they will be replaced directly

The replacement process is mainly a createElm function and the other is to destroy the oldVNode

// destroy old node
    if (isDef(parentElm)) {
     removeVnodes(parentElm, [oldVnode], 0, 0)
    } else if (isDef(oldVnode.tag)) {
     invokeDestroyHook(oldVnode)
    }

Insert To simplify the process, it is to determine the type of the node and call

createComponent respectively (it will determine whether there are children and then call it recursively)

createComment

createTextNode

After creation After using the insert function

, you need to use the hydrate function to map the virtual dom and the real dom

function insert (parent, elm, ref) {
  if (isDef(parent)) {
   if (isDef(ref)) {
    if (ref.parentNode === parent) {
     nodeOps.insertBefore(parent, elm, ref)
    }
   } else {
    nodeOps.appendChild(parent, elm)
   }
  }
 }

Core function

 function patchVnode (oldVnode, vnode, insertedVnodeQueue, removeOnly) {
  if (oldVnode === vnode) {
   return
  }
  const elm = vnode.elm = oldVnode.elm
  if (isTrue(oldVnode.isAsyncPlaceholder)) {
   if (isDef(vnode.asyncFactory.resolved)) {
    hydrate(oldVnode.elm, vnode, insertedVnodeQueue)
   } else {
    vnode.isAsyncPlaceholder = true
   }
   return
  }
  if (isTrue(vnode.isStatic) &&
   isTrue(oldVnode.isStatic) &&
   vnode.key === oldVnode.key &&
   (isTrue(vnode.isCloned) || isTrue(vnode.isOnce))
  ) {
   vnode.componentInstance = oldVnode.componentInstance
   return
  }
  let i
  const data = vnode.data
  if (isDef(data) && isDef(i = data.hook) && isDef(i = i.prepatch)) {
   i(oldVnode, vnode)
  }
  const oldCh = oldVnode.children
  const ch = vnode.children
  if (isDef(data) && isPatchable(vnode)) {
   for (i = 0; i < cbs.update.length; ++i) cbs.update[i](oldVnode, vnode)
   if (isDef(i = data.hook) && isDef(i = i.update)) i(oldVnode, vnode)
  }
  if (isUndef(vnode.text)) {
   if (isDef(oldCh) && isDef(ch)) {
    if (oldCh !== ch) updateChildren(elm, oldCh, ch, insertedVnodeQueue, removeOnly)
   } else if (isDef(ch)) {
    if (isDef(oldVnode.text)) nodeOps.setTextContent(elm, &#39;&#39;)
    addVnodes(elm, null, ch, 0, ch.length - 1, insertedVnodeQueue)
   } else if (isDef(oldCh)) {
    removeVnodes(elm, oldCh, 0, oldCh.length - 1)
   } else if (isDef(oldVnode.text)) {
    nodeOps.setTextContent(elm, &#39;&#39;)
   }
  } else if (oldVnode.text !== vnode.text) {
   nodeOps.setTextContent(elm, vnode.text)
  }
  if (isDef(data)) {
   if (isDef(i = data.hook) && isDef(i = i.postpatch)) i(oldVnode, vnode)
  }
 }

const el = vnode.el = oldVnode.el This is a very important step. Let vnode.el refer to the current real dom. When el is modified, vnode.el will change synchronously.

1. Compare whether the two references are consistent

2. I don’t know what asyncFactory does after that, so I can’t understand this comparison

3. Static node comparison key, no re-rendering will be done after the same, directly copy componentInstance (once command takes effect here)

4. If vnode is a text node Or Comment node, but when vnode.text != oldVnode.text, you only need to update the text content of vnode.elm

5. Comparison of children

• If only oldVnode has child nodes, then delete these nodes

• If only vnode has child nodes, then delete them Create these child nodes. If oldVnode is a text node, set the text of vnode.elm to empty String

• If both are present, updateChildren will be updated. This will be detailed later.

• If neither oldVnode nor vnode has child nodes, but oldVnode is a text node or comment node, set the text of vnode.elm to an empty string

updateChildren

This part focuses on the entire algorithm

First four pointers, oldStart, oldEnd, newStart, newEnd, two arrays, oldVnode, Vnode .

function updateChildren (parentElm, oldCh, newCh, insertedVnodeQueue, removeOnly) {
  let oldStartIdx = 0
  let newStartIdx = 0
  let oldEndIdx = oldCh.length - 1
  let oldStartVnode = oldCh[0]
  let oldEndVnode = oldCh[oldEndIdx]
  let newEndIdx = newCh.length - 1
  let newStartVnode = newCh[0]
  let newEndVnode = newCh[newEndIdx]
  let oldKeyToIdx, idxInOld, vnodeToMove, refElm
  while (oldStartIdx <= oldEndIdx && newStartIdx <= newEndIdx) {
   if (isUndef(oldStartVnode)) {
    oldStartVnode = oldCh[++oldStartIdx] // Vnode has been moved left
   } else if (isUndef(oldEndVnode)) {
    oldEndVnode = oldCh[--oldEndIdx]
   } else if (sameVnode(oldStartVnode, newStartVnode)) {
    patchVnode(oldStartVnode, newStartVnode, insertedVnodeQueue)
    oldStartVnode = oldCh[++oldStartIdx]
    newStartVnode = newCh[++newStartIdx]
   } else if (sameVnode(oldEndVnode, newEndVnode)) {
    patchVnode(oldEndVnode, newEndVnode, insertedVnodeQueue)
    oldEndVnode = oldCh[--oldEndIdx]
    newEndVnode = newCh[--newEndIdx]
   } else if (sameVnode(oldStartVnode, newEndVnode)) { // Vnode moved right
    patchVnode(oldStartVnode, newEndVnode, insertedVnodeQueue)
    canMove && nodeOps.insertBefore(parentElm, oldStartVnode.elm, nodeOps.nextSibling(oldEndVnode.elm))
    oldStartVnode = oldCh[++oldStartIdx]
    newEndVnode = newCh[--newEndIdx]
   } else if (sameVnode(oldEndVnode, newStartVnode)) { // Vnode moved left
    patchVnode(oldEndVnode, newStartVnode, insertedVnodeQueue)
    canMove && nodeOps.insertBefore(parentElm, oldEndVnode.elm, oldStartVnode.elm)
    oldEndVnode = oldCh[--oldEndIdx]
    newStartVnode = newCh[++newStartIdx]
   } else {
    if (isUndef(oldKeyToIdx)) oldKeyToIdx = createKeyToOldIdx(oldCh, oldStartIdx, oldEndIdx)
    idxInOld = isDef(newStartVnode.key)
     ? oldKeyToIdx[newStartVnode.key]
     : findIdxInOld(newStartVnode, oldCh, oldStartIdx, oldEndIdx)
    if (isUndef(idxInOld)) { // New element
     createElm(newStartVnode, insertedVnodeQueue, parentElm, oldStartVnode.elm, false, newCh, newStartIdx)
    } else {
     vnodeToMove = oldCh[idxInOld]
     if (sameVnode(vnodeToMove, newStartVnode)) {
      patchVnode(vnodeToMove, newStartVnode, insertedVnodeQueue)
      oldCh[idxInOld] = undefined
      canMove && nodeOps.insertBefore(parentElm, vnodeToMove.elm, oldStartVnode.elm)
     } else {
      // same key but different element. treat as new element
      createElm(newStartVnode, insertedVnodeQueue, parentElm, oldStartVnode.elm, false, newCh, newStartIdx)
     }
    }
    newStartVnode = newCh[++newStartIdx]
   }
  }
  if (oldStartIdx > oldEndIdx) {
   refElm = isUndef(newCh[newEndIdx + 1]) ? null : newCh[newEndIdx + 1].elm
   addVnodes(parentElm, refElm, newCh, newStartIdx, newEndIdx, insertedVnodeQueue)
  } else if (newStartIdx > newEndIdx) {
   removeVnodes(parentElm, oldCh, oldStartIdx, oldEndIdx)
  }
 }

Several situations and processing of a loop comparison (the following - all refer to index -) The comparison is the node node of the comparison. The abbreviation is not rigorous and the comparison uses the sameVnode function, which is not true. Congruent

Conditions for the entire loop not to end oldStartIdx <= oldEndIdx && newStartIdx <= newEndIdx

1. ##oldStart === newStart, oldStart newStart

2. oldEnd === newEnd, oldEnd-- newEnd--

3. oldStart === newEnd, oldStart Insert to the end of the queue oldStart newEnd--

4. oldEnd === newStart, oldEnd is inserted into the beginning of the queue oldEnd-- newStart

5. This is the only way to handle all the remaining situations. kind of processing, after processing newStart

• newStart在old中发现一样的那么将这个移动到oldStart前

• 没有发现一样的那么创建一个放到oldStart之前

循环结束后并没有完成

还有一段判断才算完

if (oldStartIdx > oldEndIdx) {
   refElm = isUndef(newCh[newEndIdx + 1]) ? null : newCh[newEndIdx + 1].elm
   addVnodes(parentElm, refElm, newCh, newStartIdx, newEndIdx, insertedVnodeQueue)
  } else if (newStartIdx > newEndIdx) {
   removeVnodes(parentElm, oldCh, oldStartIdx, oldEndIdx)
  }

相信看了本文案例你已经掌握了方法，更多精彩请关注php中文网其它相关文章！
推荐阅读：
json对象如何实现数组键值大小写转换功能

全局搜索时结果高亮如何实现
The above is the detailed content of How to use diff algorithm. For more information, please follow other related articles on the PHP Chinese website!