feat(getPageContentText): refactor content text extraction logic

- add special case handling for various block types
- maintain same functionality while improving structure

Close #2151

pages/search/[keyword]/index.js

@@ -3,6 +3,7 @@ import { getDataFromCache } from '@/lib/cache/cache_manager'
 import { siteConfig } from '@/lib/config'
 import { getGlobalData } from '@/lib/db/getSiteData'
 import { DynamicLayout } from '@/themes/theme'
+import { checkStrIsUuid } from '@/lib/utils'
 
 const Index = props => {
   const theme = siteConfig('THEME', BLOG.THEME, props.NOTION_CONFIG)
@@ -58,42 +59,6 @@ export function getStaticPaths() {
   }
 }
 
-/**
- * 将对象的指定字段拼接到字符串
- * @param sourceTextArray
- * @param targetObj
- * @param key
- * @returns {*}
- */
-function appendText(sourceTextArray, targetObj, key) {
-  if (!targetObj) {
-    return sourceTextArray
-  }
-  const textArray = targetObj[key]
-  const text = textArray ? getTextContent(textArray) : ''
-  if (text && text !== 'Untitled') {
-    return sourceTextArray.concat(text)
-  }
-  return sourceTextArray
-}
-
-/**
- * 递归获取层层嵌套的数组
- * @param {*} textArray
- * @returns
- */
-function getTextContent(textArray) {
-  if (typeof textArray === 'object' && isIterable(textArray)) {
-    let result = ''
-    for (const textObj of textArray) {
-      result = result + getTextContent(textObj)
-    }
-    return result
-  } else if (typeof textArray === 'string') {
-    return textArray
-  }
-}
-
 /**
  * 对象是否可以遍历
  * @param {*} obj
@@ -124,12 +89,12 @@ async function filterByMemCache(allPosts, keyword) {
         : ''
     const articleInfo = post.title + post.summary + tagContent + categoryContent
     let hit = articleInfo.toLowerCase().indexOf(keyword) > -1
-    const indexContent = getPageContentText(post, page)
+    const contentTextList = getPageContentText(post, page)
     // console.log('全文搜索缓存', cacheKey, page != null)
     post.results = []
     let hitCount = 0
-    for (const i of indexContent) {
-      const c = indexContent[i]
+    for (const i of contentTextList) {
+      const c = contentTextList[i]
       if (!c) {
         continue
       }
@@ -152,17 +117,120 @@ async function filterByMemCache(allPosts, keyword) {
 }
 
 export function getPageContentText(post, pageBlockMap) {
-  let indexContent = []
+  /**
+   * 将对象的指定字段拼接到字符串
+   * @param sourceTextArray
+   * @param targetObj
+   * @param key
+   * @returns string
+   */
+  function getText(targetObj) {
+    if (!targetObj) {
+      return ''
+    }
+    const textArray = targetObj['title'] || targetObj['caption']
+    return getTextArray(textArray)
+  }
+
+  function getTextArray(textArray) {
+    const text = textArray ? getTextContent(textArray) : ''
+    if (text && text !== 'Untitled') {
+      return text
+    }
+    return ''
+  }
+
+  const removeTypeFlag = ['a', 'p', '‣']
+
+  /**
+   * 递归获取层层嵌套的数组
+   * @param {*} textArray
+   * @returns string
+   */
+  function getTextContent(textArray) {
+    if (typeof textArray === 'object' && isIterable(textArray)) {
+      let result = ''
+      for (const textObj of textArray) {
+        if (textArray.length > 1 && removeTypeFlag.includes(textArray[0])) {
+          return result
+        }
+        result = result + getTextContent(textObj)
+      }
+      return result
+    } else if (typeof textArray === 'string') {
+      if (checkStrIsUuid(textArray) && pageBlockMap.block[textArray]) {
+        return getBlockContentText(textArray)
+      } else if (textArray === pageBlockMap.block[postId].value.space_id) {
+        return ''
+      }
+      return textArray
+    }
+  }
+
+  function getTransclusionReference(block) {
+    const result = []
+    const blockPointer = block.format.transclusion_reference_pointer
+    const blockPointerId = blockPointer.id
+    if (blockPointer) {
+      const blockContentList = pageBlockMap.block[blockPointerId].value.content
+      for (const blockContent of blockContentList) {
+        result.push(getBlockContentText(blockContent))
+      }
+    }
+    return result.join('')
+  }
+
+  function getBlockContentText(id) {
+    const block = pageBlockMap?.block[id].value
+    const blockType = block.type
+    switch (blockType) {
+      case 'transclusion_reference':
+        return getTransclusionReference(block)
+      case 'table':
+        return getTableText(block.content)
+      case 'page':
+        if (id !== postId) {
+          return getText(block.properties)
+        }
+        return ''
+      case 'breadcrumb':
+      case 'divider':
+        return ''
+      case 'quote':
+      default:
+        const properties = block?.properties
+        return getText(properties)
+    }
+  }
+
+  function getTableText(tableRowIds) {
+    const result = []
+    for (const blockRowId of tableRowIds) {
+      if (pageBlockMap.block[blockRowId]) {
+        const blockRow = pageBlockMap.block[blockRowId].value
+        const blockRowProperties = blockRow.properties
+        for (const blockRowPropertyValue of Object.values(blockRowProperties)) {
+          result.push(getTextArray(blockRowPropertyValue))
+        }
+      }
+    }
+    return result.join('')
+  }
+
+  const postId = post.id
+  let contentTextList = []
   // 防止搜到加密文章的内容
   if (pageBlockMap && pageBlockMap.block && !post.password) {
     const contentIds = Object.keys(pageBlockMap.block)
-    contentIds.forEach(id => {
-      const properties = pageBlockMap?.block[id]?.value?.properties
-      indexContent = appendText(indexContent, properties, 'title')
-      indexContent = appendText(indexContent, properties, 'caption')
-    })
+    for (const id of contentIds) {
+      const blockContentText = getBlockContentText(id)
+      if (blockContentText) {
+        contentTextList.push(blockContentText)
+      }
+    }
   }
-  return indexContent.join('')
+  console.log(contentTextList.join(''))
+  return contentTextList.join('')
 }
 
 export default Index