dify

dify/web/app/components/datasets/documents/create-from-pipeline/actions/index.tsx

@@ -0,0 +1,90 @@
+import React, { useMemo } from 'react'
+import Button from '@/app/components/base/button'
+import { useTranslation } from 'react-i18next'
+import { useParams } from 'next/navigation'
+import { RiArrowRightLine } from '@remixicon/react'
+import Link from 'next/link'
+import Checkbox from '@/app/components/base/checkbox'
+
+type ActionsProps = {
+  disabled?: boolean
+  handleNextStep: () => void
+  showSelect?: boolean
+  totalOptions?: number
+  selectedOptions?: number
+  onSelectAll?: () => void
+  tip?: string
+}
+
+const Actions = ({
+  disabled,
+  handleNextStep,
+  showSelect = false,
+  totalOptions,
+  selectedOptions,
+  onSelectAll,
+  tip = '',
+}: ActionsProps) => {
+  const { t } = useTranslation()
+  const { datasetId } = useParams()
+
+  const indeterminate = useMemo(() => {
+    if (!showSelect) return false
+    if (selectedOptions === undefined || totalOptions === undefined) return false
+    return selectedOptions > 0 && selectedOptions < totalOptions
+  }, [showSelect, selectedOptions, totalOptions])
+
+  const checked = useMemo(() => {
+    if (!showSelect) return false
+    if (selectedOptions === undefined || totalOptions === undefined) return false
+    return selectedOptions > 0 && selectedOptions === totalOptions
+  }, [showSelect, selectedOptions, totalOptions])
+
+  return (
+    <div className='flex items-center gap-x-2 overflow-hidden'>
+      {showSelect && (
+        <>
+          <div className='flex shrink-0 items-center gap-x-2 py-[3px] pl-4 pr-2'>
+            <Checkbox
+              onCheck={onSelectAll}
+              indeterminate={indeterminate}
+              checked={checked}
+            />
+            <span className='system-sm-medium text-text-accent'>
+              {t('common.operation.selectAll')}
+            </span>
+          </div>
+          {tip && (
+            <div title={tip} className='system-xs-regular max-w-full truncate text-text-tertiary'>
+              {tip}
+            </div>
+          )}
+        </>
+      )}
+      <div className='flex grow items-center justify-end gap-x-2'>
+        <Link
+          href={`/datasets/${datasetId}/documents`}
+          replace
+        >
+          <Button
+            variant='ghost'
+            className='px-3 py-2'
+          >
+            {t('common.operation.cancel')}
+          </Button>
+        </Link>
+        <Button
+          disabled={disabled}
+          variant='primary'
+          onClick={handleNextStep}
+          className='gap-x-0.5'
+        >
+          <span className='px-0.5'>{t('datasetCreation.stepOne.button')}</span>
+          <RiArrowRightLine className='size-4' />
+        </Button>
+      </div>
+    </div>
+  )
+}
+
+export default React.memo(Actions)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source-options/datasource-icon.tsx

@@ -0,0 +1,40 @@
+import type { FC } from 'react'
+import { memo } from 'react'
+import cn from '@/utils/classnames'
+
+type DatasourceIconProps = {
+  size?: string
+  className?: string
+  iconUrl: string
+}
+
+const ICON_CONTAINER_CLASSNAME_SIZE_MAP: Record<string, string> = {
+  xs: 'w-4 h-4 rounded-[5px] shadow-xs',
+  sm: 'w-5 h-5 rounded-md shadow-xs',
+  md: 'w-6 h-6 rounded-lg shadow-md',
+}
+
+const DatasourceIcon: FC<DatasourceIconProps> = ({
+  size = 'sm',
+  className,
+  iconUrl,
+}) => {
+  return (
+    <div className={
+      cn(
+        'flex items-center justify-center shadow-none',
+        ICON_CONTAINER_CLASSNAME_SIZE_MAP[size],
+        className,
+      )}
+    >
+      <div
+        className='h-full w-full shrink-0 rounded-md bg-cover bg-center'
+        style={{
+          backgroundImage: `url(${iconUrl})`,
+        }}
+      />
+    </div>
+  )
+}
+
+export default memo(DatasourceIcon)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source-options/hooks.tsx

@@ -0,0 +1,23 @@
+import { useMemo } from 'react'
+import type { DataSourceNodeType } from '@/app/components/workflow/nodes/data-source/types'
+import { basePath } from '@/utils/var'
+import { useDataSourceList } from '@/service/use-pipeline'
+import { transformDataSourceToTool } from '@/app/components/workflow/block-selector/utils'
+
+export const useDatasourceIcon = (data: DataSourceNodeType) => {
+  const { data: dataSourceListData, isSuccess } = useDataSourceList(true)
+
+  const datasourceIcon = useMemo(() => {
+    if (!isSuccess) return
+    const dataSourceList = [...(dataSourceListData || [])]
+    dataSourceList.forEach((item) => {
+      const icon = item.declaration.identity.icon
+      if (typeof icon == 'string' && !icon.includes(basePath))
+        item.declaration.identity.icon = `${basePath}${icon}`
+    })
+    const formattedDataSourceList = dataSourceList.map(item => transformDataSourceToTool(item))
+    return formattedDataSourceList?.find(toolWithProvider => toolWithProvider.plugin_id === data.plugin_id)?.icon
+  }, [data.plugin_id, dataSourceListData, isSuccess])
+
+  return datasourceIcon
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source-options/index.tsx

@@ -0,0 +1,52 @@
+import { useCallback, useEffect } from 'react'
+import { useDatasourceOptions } from '../hooks'
+import OptionCard from './option-card'
+import type { Datasource } from '@/app/components/rag-pipeline/components/panel/test-run/types'
+import type { Node } from '@/app/components/workflow/types'
+import type { DataSourceNodeType } from '@/app/components/workflow/nodes/data-source/types'
+
+type DataSourceOptionsProps = {
+  pipelineNodes: Node<DataSourceNodeType>[]
+  datasourceNodeId: string
+  onSelect: (option: Datasource) => void
+}
+
+const DataSourceOptions = ({
+  pipelineNodes,
+  datasourceNodeId,
+  onSelect,
+}: DataSourceOptionsProps) => {
+  const options = useDatasourceOptions(pipelineNodes)
+
+  const handelSelect = useCallback((value: string) => {
+    const selectedOption = options.find(option => option.value === value)
+    if (!selectedOption)
+      return
+    const datasource = {
+      nodeId: selectedOption.value,
+      nodeData: selectedOption.data,
+    }
+    onSelect(datasource)
+  }, [onSelect, options])
+
+  useEffect(() => {
+    if (options.length > 0 && !datasourceNodeId)
+      handelSelect(options[0].value)
+  }, [])
+
+  return (
+    <div className='grid w-full grid-cols-4 gap-1'>
+      {options.map(option => (
+        <OptionCard
+          key={option.value}
+          label={option.label}
+          selected={datasourceNodeId === option.value}
+          nodeData={option.data}
+          onClick={handelSelect.bind(null, option.value)}
+        />
+      ))}
+    </div>
+  )
+}
+
+export default DataSourceOptions

dify/web/app/components/datasets/documents/create-from-pipeline/data-source-options/option-card.tsx

@@ -0,0 +1,45 @@
+import React from 'react'
+import cn from '@/utils/classnames'
+import type { DataSourceNodeType } from '@/app/components/workflow/nodes/data-source/types'
+import DatasourceIcon from './datasource-icon'
+import { useDatasourceIcon } from './hooks'
+
+type OptionCardProps = {
+  label: string
+  selected: boolean
+  nodeData: DataSourceNodeType
+  onClick?: () => void
+}
+
+const OptionCard = ({
+  label,
+  selected,
+  nodeData,
+  onClick,
+}: OptionCardProps) => {
+  const iconUrl = useDatasourceIcon(nodeData) as string
+
+  return (
+    <div
+      className={cn(
+        'flex cursor-pointer items-center gap-2 rounded-xl border border-components-option-card-option-border bg-components-option-card-option-bg p-3 shadow-shadow-shadow-3',
+        selected
+          ? 'border-components-option-card-option-selected-border bg-components-option-card-option-selected-bg shadow-xs ring-[0.5px] ring-inset ring-components-option-card-option-selected-border'
+          : 'hover:bg-components-option-card-bg-hover hover:border-components-option-card-option-border-hover hover:shadow-xs',
+      )}
+      onClick={onClick}
+    >
+      <div className='flex size-8 shrink-0 items-center justify-center rounded-lg border-[0.5px] border-components-panel-border bg-background-default-dodge p-1.5'>
+        <DatasourceIcon iconUrl={iconUrl} />
+      </div>
+      <div
+        className={cn('system-sm-medium line-clamp-2 grow text-text-secondary', selected && 'text-text-primary')}
+        title={label}
+      >
+        {label}
+      </div>
+    </div>
+  )
+}
+
+export default React.memo(OptionCard)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/base/credential-selector/index.tsx

@@ -0,0 +1,69 @@
+import React, { useCallback, useEffect, useMemo } from 'react'
+import {
+  PortalToFollowElem,
+  PortalToFollowElemContent,
+  PortalToFollowElemTrigger,
+} from '@/app/components/base/portal-to-follow-elem'
+import type { DataSourceCredential } from '@/types/pipeline'
+import { useBoolean } from 'ahooks'
+import Trigger from './trigger'
+import List from './list'
+
+export type CredentialSelectorProps = {
+  pluginName: string
+  currentCredentialId: string
+  onCredentialChange: (credentialId: string) => void
+  credentials: Array<DataSourceCredential>
+}
+
+const CredentialSelector = ({
+  pluginName,
+  currentCredentialId,
+  onCredentialChange,
+  credentials,
+}: CredentialSelectorProps) => {
+  const [open, { toggle }] = useBoolean(false)
+
+  const currentCredential = useMemo(() => {
+    return credentials.find(cred => cred.id === currentCredentialId)
+  }, [credentials, currentCredentialId])
+
+  useEffect(() => {
+    if (!currentCredential && credentials.length)
+      onCredentialChange(credentials[0].id)
+  }, [currentCredential, credentials])
+
+  const handleCredentialChange = useCallback((credentialId: string) => {
+    onCredentialChange(credentialId)
+    toggle()
+  }, [onCredentialChange, toggle])
+
+  return (
+    <PortalToFollowElem
+      open={open}
+      onOpenChange={toggle}
+      placement='bottom-start'
+      offset={{
+        mainAxis: 4,
+      }}
+    >
+      <PortalToFollowElemTrigger onClick={toggle} className='grow overflow-hidden'>
+        <Trigger
+          currentCredential={currentCredential}
+          pluginName={pluginName}
+          isOpen={open}
+        />
+      </PortalToFollowElemTrigger>
+      <PortalToFollowElemContent className='z-10'>
+        <List
+          currentCredentialId={currentCredentialId}
+          credentials={credentials}
+          pluginName={pluginName}
+          onCredentialChange={handleCredentialChange}
+        />
+      </PortalToFollowElemContent>
+    </PortalToFollowElem>
+  )
+}
+
+export default React.memo(CredentialSelector)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/base/credential-selector/item.tsx

@@ -0,0 +1,52 @@
+import { CredentialIcon } from '@/app/components/datasets/common/credential-icon'
+import type { DataSourceCredential } from '@/types/pipeline'
+import { RiCheckLine } from '@remixicon/react'
+import React, { useCallback } from 'react'
+import { useTranslation } from 'react-i18next'
+
+type ItemProps = {
+  credential: DataSourceCredential
+  pluginName: string
+  isSelected: boolean
+  onCredentialChange: (credentialId: string) => void
+}
+
+const Item = ({
+  credential,
+  pluginName,
+  isSelected,
+  onCredentialChange,
+}: ItemProps) => {
+  const { t } = useTranslation()
+  const { avatar_url, name } = credential
+
+  const handleCredentialChange = useCallback(() => {
+    onCredentialChange(credential.id)
+  }, [credential.id, onCredentialChange])
+
+  return (
+    <div
+      className='flex cursor-pointer items-center gap-x-2 rounded-lg p-2 hover:bg-state-base-hover'
+      onClick={handleCredentialChange}
+    >
+      <CredentialIcon
+        avatar_url={avatar_url}
+        name={name}
+        size={20}
+      />
+      <span className='system-sm-medium grow truncate text-text-secondary'>
+        {t('datasetPipeline.credentialSelector.name', {
+          credentialName: name,
+          pluginName,
+        })}
+      </span>
+      {
+        isSelected && (
+          <RiCheckLine className='size-4 shrink-0 text-text-accent' />
+        )
+      }
+    </div>
+  )
+}
+
+export default React.memo(Item)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/base/credential-selector/list.tsx

@@ -0,0 +1,38 @@
+import type { DataSourceCredential } from '@/types/pipeline'
+import React from 'react'
+import Item from './item'
+
+type ListProps = {
+  currentCredentialId: string
+  credentials: Array<DataSourceCredential>
+  pluginName: string
+  onCredentialChange: (credentialId: string) => void
+}
+
+const List = ({
+  currentCredentialId,
+  credentials,
+  pluginName,
+  onCredentialChange,
+}: ListProps) => {
+  return (
+    <div className='flex w-[280px] flex-col gap-y-1 rounded-xl border-[0.5px] border-components-panel-border bg-components-panel-bg-blur p-1 shadow-lg shadow-shadow-shadow-5 backdrop-blur-[5px]'>
+      {
+        credentials.map((credential) => {
+          const isSelected = credential.id === currentCredentialId
+          return (
+            <Item
+              key={credential.id}
+              credential={credential}
+              pluginName={pluginName}
+              isSelected={isSelected}
+              onCredentialChange={onCredentialChange}
+            />
+          )
+        })
+      }
+    </div>
+  )
+}
+
+export default React.memo(List)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/base/credential-selector/trigger.tsx

@@ -0,0 +1,51 @@
+import React from 'react'
+import type { DataSourceCredential } from '@/types/pipeline'
+import { useTranslation } from 'react-i18next'
+import { RiArrowDownSLine } from '@remixicon/react'
+import cn from '@/utils/classnames'
+import { CredentialIcon } from '@/app/components/datasets/common/credential-icon'
+
+type TriggerProps = {
+  currentCredential: DataSourceCredential | undefined
+  pluginName: string
+  isOpen: boolean
+}
+
+const Trigger = ({
+  currentCredential,
+  pluginName,
+  isOpen,
+}: TriggerProps) => {
+  const { t } = useTranslation()
+
+  const {
+    avatar_url,
+    name = '',
+  } = currentCredential || {}
+
+  return (
+    <div
+      className={cn(
+        'flex cursor-pointer items-center gap-x-2 rounded-md p-1 pr-2',
+        isOpen ? 'bg-state-base-hover' : 'hover:bg-state-base-hover',
+      )}
+    >
+      <CredentialIcon
+        avatar_url={avatar_url}
+        name={name}
+        size={20}
+      />
+      <div className='flex grow items-center gap-x-1 overflow-hidden'>
+        <span className='system-md-semibold grow truncate text-text-secondary'>
+          {t('datasetPipeline.credentialSelector.name', {
+            credentialName: name,
+            pluginName,
+          })}
+        </span>
+        <RiArrowDownSLine className='size-4 shrink-0 text-text-secondary' />
+      </div>
+    </div>
+  )
+}
+
+export default React.memo(Trigger)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/base/header.tsx

@@ -0,0 +1,60 @@
+import React from 'react'
+import Divider from '@/app/components/base/divider'
+import Button from '@/app/components/base/button'
+import { RiBookOpenLine, RiEqualizer2Line } from '@remixicon/react'
+import type { CredentialSelectorProps } from './credential-selector'
+import CredentialSelector from './credential-selector'
+import Tooltip from '@/app/components/base/tooltip'
+import { useTranslation } from 'react-i18next'
+
+type HeaderProps = {
+  docTitle: string
+  docLink: string
+  onClickConfiguration?: () => void
+} & CredentialSelectorProps
+
+const Header = ({
+  docTitle,
+  docLink,
+  onClickConfiguration,
+  ...rest
+}: HeaderProps) => {
+  const { t } = useTranslation()
+
+  return (
+    <div className='flex items-center justify-between gap-x-2'>
+      <div className='flex items-center gap-x-1 overflow-hidden'>
+        <CredentialSelector
+          {...rest}
+        />
+        <Divider type='vertical' className='mx-1 h-3.5 shrink-0' />
+        <Tooltip
+          popupContent={t('datasetPipeline.configurationTip', { pluginName: rest.pluginName })}
+          position='top'
+        >
+          <Button
+            variant='ghost'
+            size='small'
+            className='size-6 shrink-0 px-1'
+          >
+            <RiEqualizer2Line
+              className='h-4 w-4'
+              onClick={onClickConfiguration}
+            />
+          </Button>
+        </Tooltip>
+      </div>
+      <a
+        className='system-xs-medium flex shrink-0 items-center gap-x-1 text-text-accent'
+        href={docLink}
+        target='_blank'
+        rel='noopener noreferrer'
+      >
+        <RiBookOpenLine className='size-3.5 shrink-0' />
+        <span title={docTitle}>{docTitle}</span>
+      </a>
+    </div>
+  )
+}
+
+export default React.memo(Header)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/local-file/index.tsx

@@ -0,0 +1,369 @@
+'use client'
+import React, { useCallback, useEffect, useMemo, useRef, useState } from 'react'
+import { useTranslation } from 'react-i18next'
+import { useContext } from 'use-context-selector'
+import { RiDeleteBinLine, RiErrorWarningFill, RiUploadCloud2Line } from '@remixicon/react'
+import DocumentFileIcon from '@/app/components/datasets/common/document-file-icon'
+import cn from '@/utils/classnames'
+import type { CustomFile as File, FileItem } from '@/models/datasets'
+import { ToastContext } from '@/app/components/base/toast'
+import { upload } from '@/service/base'
+import { getFileUploadErrorMessage } from '@/app/components/base/file-uploader/utils'
+import I18n from '@/context/i18n'
+import { LanguagesSupported } from '@/i18n-config/language'
+import { IS_CE_EDITION } from '@/config'
+import { Theme } from '@/types/app'
+import useTheme from '@/hooks/use-theme'
+import { useFileUploadConfig } from '@/service/use-common'
+import { useDataSourceStore, useDataSourceStoreWithSelector } from '../store'
+import { produce } from 'immer'
+import dynamic from 'next/dynamic'
+
+const SimplePieChart = dynamic(() => import('@/app/components/base/simple-pie-chart'), { ssr: false })
+
+const FILES_NUMBER_LIMIT = 20
+
+export type LocalFileProps = {
+  allowedExtensions: string[]
+  notSupportBatchUpload?: boolean
+}
+
+const LocalFile = ({
+  allowedExtensions,
+  notSupportBatchUpload,
+}: LocalFileProps) => {
+  const { t } = useTranslation()
+  const { notify } = useContext(ToastContext)
+  const { locale } = useContext(I18n)
+  const localFileList = useDataSourceStoreWithSelector(state => state.localFileList)
+  const dataSourceStore = useDataSourceStore()
+  const [dragging, setDragging] = useState(false)
+
+  const dropRef = useRef<HTMLDivElement>(null)
+  const dragRef = useRef<HTMLDivElement>(null)
+  const fileUploader = useRef<HTMLInputElement>(null)
+  const fileListRef = useRef<FileItem[]>([])
+
+  const hideUpload = notSupportBatchUpload && localFileList.length > 0
+
+  const { data: fileUploadConfigResponse } = useFileUploadConfig()
+  const supportTypesShowNames = useMemo(() => {
+    const extensionMap: { [key: string]: string } = {
+      md: 'markdown',
+      pptx: 'pptx',
+      htm: 'html',
+      xlsx: 'xlsx',
+      docx: 'docx',
+    }
+
+    return allowedExtensions
+      .map(item => extensionMap[item] || item) // map to standardized extension
+      .map(item => item.toLowerCase()) // convert to lower case
+      .filter((item, index, self) => self.indexOf(item) === index) // remove duplicates
+      .map(item => item.toUpperCase()) // convert to upper case
+      .join(locale !== LanguagesSupported[1] ? ', ' : '、 ')
+  }, [locale, allowedExtensions])
+  const ACCEPTS = allowedExtensions.map((ext: string) => `.${ext}`)
+  const fileUploadConfig = useMemo(() => fileUploadConfigResponse ?? {
+    file_size_limit: 15,
+    batch_count_limit: 5,
+  }, [fileUploadConfigResponse])
+
+  const updateFile = useCallback((fileItem: FileItem, progress: number, list: FileItem[]) => {
+    const { setLocalFileList } = dataSourceStore.getState()
+    const newList = produce(list, (draft) => {
+      const targetIndex = draft.findIndex(file => file.fileID === fileItem.fileID)
+      draft[targetIndex] = {
+        ...draft[targetIndex],
+        progress,
+      }
+    })
+    setLocalFileList(newList)
+  }, [dataSourceStore])
+
+  const updateFileList = useCallback((preparedFiles: FileItem[]) => {
+    const { setLocalFileList } = dataSourceStore.getState()
+    setLocalFileList(preparedFiles)
+  }, [dataSourceStore])
+
+  const handlePreview = useCallback((file: File) => {
+    const { setCurrentLocalFile } = dataSourceStore.getState()
+    if (file.id)
+      setCurrentLocalFile(file)
+  }, [dataSourceStore])
+
+  // utils
+  const getFileType = (currentFile: File) => {
+    if (!currentFile)
+      return ''
+
+    const arr = currentFile.name.split('.')
+    return arr[arr.length - 1]
+  }
+
+  const getFileSize = (size: number) => {
+    if (size / 1024 < 10)
+      return `${(size / 1024).toFixed(2)}KB`
+
+    return `${(size / 1024 / 1024).toFixed(2)}MB`
+  }
+
+  const isValid = useCallback((file: File) => {
+    const { size } = file
+    const ext = `.${getFileType(file)}`
+    const isValidType = ACCEPTS.includes(ext.toLowerCase())
+    if (!isValidType)
+      notify({ type: 'error', message: t('datasetCreation.stepOne.uploader.validation.typeError') })
+
+    const isValidSize = size <= fileUploadConfig.file_size_limit * 1024 * 1024
+    if (!isValidSize)
+      notify({ type: 'error', message: t('datasetCreation.stepOne.uploader.validation.size', { size: fileUploadConfig.file_size_limit }) })
+
+    return isValidType && isValidSize
+  }, [fileUploadConfig, notify, t, ACCEPTS])
+
+  type UploadResult = Awaited<ReturnType<typeof upload>>
+
+  const fileUpload = useCallback(async (fileItem: FileItem): Promise<FileItem> => {
+    const formData = new FormData()
+    formData.append('file', fileItem.file)
+    const onProgress = (e: ProgressEvent) => {
+      if (e.lengthComputable) {
+        const percent = Math.floor(e.loaded / e.total * 100)
+        updateFile(fileItem, percent, fileListRef.current)
+      }
+    }
+
+    return upload({
+      xhr: new XMLHttpRequest(),
+      data: formData,
+      onprogress: onProgress,
+    }, false, undefined, '?source=datasets')
+      .then((res: UploadResult) => {
+        const updatedFile = Object.assign({}, fileItem.file, {
+          id: res.id,
+          ...(res as Partial<File>),
+        }) as File
+        const completeFile: FileItem = {
+          fileID: fileItem.fileID,
+          file: updatedFile,
+          progress: -1,
+        }
+        const index = fileListRef.current.findIndex(item => item.fileID === fileItem.fileID)
+        fileListRef.current[index] = completeFile
+        updateFile(completeFile, 100, fileListRef.current)
+        return Promise.resolve({ ...completeFile })
+      })
+      .catch((e) => {
+        const errorMessage = getFileUploadErrorMessage(e, t('datasetCreation.stepOne.uploader.failed'), t)
+        notify({ type: 'error', message: errorMessage })
+        updateFile(fileItem, -2, fileListRef.current)
+        return Promise.resolve({ ...fileItem })
+      })
+      .finally()
+  }, [fileListRef, notify, updateFile, t])
+
+  const uploadBatchFiles = useCallback((bFiles: FileItem[]) => {
+    bFiles.forEach(bf => (bf.progress = 0))
+    return Promise.all(bFiles.map(fileUpload))
+  }, [fileUpload])
+
+  const uploadMultipleFiles = useCallback(async (files: FileItem[]) => {
+    const batchCountLimit = fileUploadConfig.batch_count_limit
+    const length = files.length
+    let start = 0
+    let end = 0
+
+    while (start < length) {
+      if (start + batchCountLimit > length)
+        end = length
+      else
+        end = start + batchCountLimit
+      const bFiles = files.slice(start, end)
+      await uploadBatchFiles(bFiles)
+      start = end
+    }
+  }, [fileUploadConfig, uploadBatchFiles])
+
+  const initialUpload = useCallback((files: File[]) => {
+    if (!files.length)
+      return false
+
+    if (files.length + localFileList.length > FILES_NUMBER_LIMIT && !IS_CE_EDITION) {
+      notify({ type: 'error', message: t('datasetCreation.stepOne.uploader.validation.filesNumber', { filesNumber: FILES_NUMBER_LIMIT }) })
+      return false
+    }
+
+    const preparedFiles = files.map((file, index) => ({
+      fileID: `file${index}-${Date.now()}`,
+      file,
+      progress: -1,
+    }))
+    const newFiles = [...fileListRef.current, ...preparedFiles]
+    updateFileList(newFiles)
+    fileListRef.current = newFiles
+    uploadMultipleFiles(preparedFiles)
+  }, [updateFileList, uploadMultipleFiles, notify, t, localFileList])
+
+  const handleDragEnter = (e: DragEvent) => {
+    e.preventDefault()
+    e.stopPropagation()
+    if (e.target !== dragRef.current)
+      setDragging(true)
+  }
+  const handleDragOver = (e: DragEvent) => {
+    e.preventDefault()
+    e.stopPropagation()
+  }
+  const handleDragLeave = (e: DragEvent) => {
+    e.preventDefault()
+    e.stopPropagation()
+    if (e.target === dragRef.current)
+      setDragging(false)
+  }
+
+  const handleDrop = useCallback((e: DragEvent) => {
+    e.preventDefault()
+    e.stopPropagation()
+    setDragging(false)
+    if (!e.dataTransfer)
+      return
+
+    let files = [...e.dataTransfer.files] as File[]
+    if (notSupportBatchUpload)
+      files = files.slice(0, 1)
+
+    const validFiles = files.filter(isValid)
+    initialUpload(validFiles)
+  }, [initialUpload, isValid, notSupportBatchUpload])
+
+  const selectHandle = useCallback(() => {
+    if (fileUploader.current)
+      fileUploader.current.click()
+  }, [])
+
+  const removeFile = (fileID: string) => {
+    if (fileUploader.current)
+      fileUploader.current.value = ''
+
+    fileListRef.current = fileListRef.current.filter(item => item.fileID !== fileID)
+    updateFileList([...fileListRef.current])
+  }
+  const fileChangeHandle = useCallback((e: React.ChangeEvent<HTMLInputElement>) => {
+    const files = [...(e.target.files ?? [])] as File[]
+    initialUpload(files.filter(isValid))
+  }, [isValid, initialUpload])
+
+  const { theme } = useTheme()
+  const chartColor = useMemo(() => theme === Theme.dark ? '#5289ff' : '#296dff', [theme])
+
+  useEffect(() => {
+    const dropElement = dropRef.current
+    dropElement?.addEventListener('dragenter', handleDragEnter)
+    dropElement?.addEventListener('dragover', handleDragOver)
+    dropElement?.addEventListener('dragleave', handleDragLeave)
+    dropElement?.addEventListener('drop', handleDrop)
+    return () => {
+      dropElement?.removeEventListener('dragenter', handleDragEnter)
+      dropElement?.removeEventListener('dragover', handleDragOver)
+      dropElement?.removeEventListener('dragleave', handleDragLeave)
+      dropElement?.removeEventListener('drop', handleDrop)
+    }
+  }, [handleDrop])
+
+  return (
+    <div className='flex flex-col'>
+      {!hideUpload && (
+        <input
+          ref={fileUploader}
+          id='fileUploader'
+          className='hidden'
+          type='file'
+          multiple={!notSupportBatchUpload}
+          accept={ACCEPTS.join(',')}
+          onChange={fileChangeHandle}
+        />
+      )}
+      {!hideUpload && (
+        <div
+          ref={dropRef}
+          className={cn(
+            'relative box-border flex min-h-20 flex-col items-center justify-center gap-1 rounded-xl border border-dashed border-components-dropzone-border bg-components-dropzone-bg px-4 py-3 text-xs leading-4 text-text-tertiary',
+            dragging && 'border-components-dropzone-border-accent bg-components-dropzone-bg-accent',
+          )}>
+          <div className='flex min-h-5 items-center justify-center text-sm leading-4 text-text-secondary'>
+            <RiUploadCloud2Line className='mr-2 size-5' />
+
+            <span>
+              {notSupportBatchUpload ? t('datasetCreation.stepOne.uploader.buttonSingleFile') : t('datasetCreation.stepOne.uploader.button')}
+              {allowedExtensions.length > 0 && (
+                <label className='ml-1 cursor-pointer text-text-accent' onClick={selectHandle}>{t('datasetCreation.stepOne.uploader.browse')}</label>
+              )}
+            </span>
+          </div>
+          <div>{t('datasetCreation.stepOne.uploader.tip', {
+            size: fileUploadConfig.file_size_limit,
+            supportTypes: supportTypesShowNames,
+            batchCount: notSupportBatchUpload ? 1 : fileUploadConfig.batch_count_limit,
+          })}</div>
+          {dragging && <div ref={dragRef} className='absolute left-0 top-0 h-full w-full' />}
+        </div>
+      )}
+      {localFileList.length > 0 && (
+        <div className='mt-1 flex flex-col gap-y-1'>
+          {localFileList.map((fileItem, index) => {
+            const isUploading = fileItem.progress >= 0 && fileItem.progress < 100
+            const isError = fileItem.progress === -2
+            return (
+              <div
+                key={`${fileItem.fileID}-${index}`}
+                onClick={handlePreview.bind(null, fileItem.file)}
+                className={cn(
+                  'flex h-12 items-center rounded-lg border border-components-panel-border bg-components-panel-on-panel-item-bg shadow-xs shadow-shadow-shadow-4',
+                  isError && 'border-state-destructive-border bg-state-destructive-hover',
+                )}
+              >
+                <div className='flex w-12 shrink-0 items-center justify-center'>
+                  <DocumentFileIcon
+                    size='lg'
+                    className='shrink-0'
+                    name={fileItem.file.name}
+                    extension={getFileType(fileItem.file)}
+                  />
+                </div>
+                <div className='flex shrink grow flex-col gap-0.5'>
+                  <div className='flex w-full'>
+                    <div className='w-0 grow truncate text-xs text-text-secondary'>{fileItem.file.name}</div>
+                  </div>
+                  <div className='w-full truncate text-2xs leading-3 text-text-tertiary'>
+                    <span className='uppercase'>{getFileType(fileItem.file)}</span>
+                    <span className='px-1 text-text-quaternary'>·</span>
+                    <span>{getFileSize(fileItem.file.size)}</span>
+                  </div>
+                </div>
+                <div className='flex w-16 shrink-0 items-center justify-end gap-1 pr-3'>
+                  {isUploading && (
+                    <SimplePieChart percentage={fileItem.progress} stroke={chartColor} fill={chartColor} animationDuration={0} />
+                  )}
+                  {
+                    isError && (
+                      <RiErrorWarningFill className='size-4 text-text-destructive' />
+                    )
+                  }
+                  <span className='flex h-6 w-6 cursor-pointer items-center justify-center' onClick={(e) => {
+                    e.stopPropagation()
+                    removeFile(fileItem.fileID)
+                  }}>
+                    <RiDeleteBinLine className='size-4 text-text-tertiary' />
+                  </span>
+                </div>
+              </div>
+            )
+          })}
+        </div>
+      )}
+    </div>
+  )
+}
+
+export default LocalFile

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-documents/index.tsx

@@ -0,0 +1,174 @@
+import { useCallback, useEffect, useMemo } from 'react'
+import SearchInput from '@/app/components/base/notion-page-selector/search-input'
+import PageSelector from './page-selector'
+import type { DataSourceNotionPageMap, DataSourceNotionWorkspace } from '@/models/common'
+import Header from '../base/header'
+import { useDatasetDetailContextWithSelector } from '@/context/dataset-detail'
+import { DatasourceType } from '@/models/pipeline'
+import { ssePost } from '@/service/base'
+import Toast from '@/app/components/base/toast'
+import type { DataSourceNodeCompletedResponse, DataSourceNodeErrorResponse } from '@/types/pipeline'
+import type { DataSourceNodeType } from '@/app/components/workflow/nodes/data-source/types'
+import { useDataSourceStore, useDataSourceStoreWithSelector } from '../store'
+import { useShallow } from 'zustand/react/shallow'
+import { useModalContextSelector } from '@/context/modal-context'
+import Title from './title'
+import { useGetDataSourceAuth } from '@/service/use-datasource'
+import Loading from '@/app/components/base/loading'
+import { useDocLink } from '@/context/i18n'
+import { ACCOUNT_SETTING_TAB } from '@/app/components/header/account-setting/constants'
+
+type OnlineDocumentsProps = {
+  isInPipeline?: boolean
+  nodeId: string
+  nodeData: DataSourceNodeType
+  onCredentialChange: (credentialId: string) => void
+}
+
+const OnlineDocuments = ({
+  nodeId,
+  nodeData,
+  isInPipeline = false,
+  onCredentialChange,
+}: OnlineDocumentsProps) => {
+  const docLink = useDocLink()
+  const pipelineId = useDatasetDetailContextWithSelector(s => s.dataset?.pipeline_id)
+  const setShowAccountSettingModal = useModalContextSelector(s => s.setShowAccountSettingModal)
+  const {
+    documentsData,
+    searchValue,
+    selectedPagesId,
+    currentCredentialId,
+  } = useDataSourceStoreWithSelector(useShallow(state => ({
+    documentsData: state.documentsData,
+    searchValue: state.searchValue,
+    selectedPagesId: state.selectedPagesId,
+    currentCredentialId: state.currentCredentialId,
+  })))
+
+  const { data: dataSourceAuth } = useGetDataSourceAuth({
+    pluginId: nodeData.plugin_id,
+    provider: nodeData.provider_name,
+  })
+
+  const dataSourceStore = useDataSourceStore()
+
+  const PagesMapAndSelectedPagesId: DataSourceNotionPageMap = useMemo(() => {
+    const pagesMap = (documentsData || []).reduce((prev: DataSourceNotionPageMap, next: DataSourceNotionWorkspace) => {
+      next.pages.forEach((page) => {
+        prev[page.page_id] = {
+          ...page,
+          workspace_id: next.workspace_id,
+        }
+      })
+
+      return prev
+    }, {})
+    return pagesMap
+  }, [documentsData])
+
+  const datasourceNodeRunURL = !isInPipeline
+    ? `/rag/pipelines/${pipelineId}/workflows/published/datasource/nodes/${nodeId}/run`
+    : `/rag/pipelines/${pipelineId}/workflows/draft/datasource/nodes/${nodeId}/run`
+
+  const getOnlineDocuments = useCallback(async () => {
+    const { currentCredentialId } = dataSourceStore.getState()
+    ssePost(
+      datasourceNodeRunURL,
+      {
+        body: {
+          inputs: {},
+          credential_id: currentCredentialId,
+          datasource_type: DatasourceType.onlineDocument,
+        },
+      },
+      {
+        onDataSourceNodeCompleted: (documentsData: DataSourceNodeCompletedResponse) => {
+          const { setDocumentsData } = dataSourceStore.getState()
+          setDocumentsData(documentsData.data as DataSourceNotionWorkspace[])
+        },
+        onDataSourceNodeError: (error: DataSourceNodeErrorResponse) => {
+          Toast.notify({
+            type: 'error',
+            message: error.error,
+          })
+        },
+      },
+    )
+  }, [dataSourceStore, datasourceNodeRunURL])
+
+  useEffect(() => {
+    if (!currentCredentialId) return
+    getOnlineDocuments()
+  }, [currentCredentialId])
+
+  const handleSearchValueChange = useCallback((value: string) => {
+    const { setSearchValue } = dataSourceStore.getState()
+    setSearchValue(value)
+  }, [dataSourceStore])
+
+  const handleSelectPages = useCallback((newSelectedPagesId: Set<string>) => {
+    const { setSelectedPagesId, setOnlineDocuments } = dataSourceStore.getState()
+    const selectedPages = Array.from(newSelectedPagesId).map(pageId => PagesMapAndSelectedPagesId[pageId])
+    setSelectedPagesId(new Set(Array.from(newSelectedPagesId)))
+    setOnlineDocuments(selectedPages)
+  }, [dataSourceStore, PagesMapAndSelectedPagesId])
+
+  const handlePreviewPage = useCallback((previewPageId: string) => {
+    const { setCurrentDocument } = dataSourceStore.getState()
+    setCurrentDocument(PagesMapAndSelectedPagesId[previewPageId])
+  }, [PagesMapAndSelectedPagesId, dataSourceStore])
+
+  const handleSetting = useCallback(() => {
+    setShowAccountSettingModal({
+      payload: ACCOUNT_SETTING_TAB.DATA_SOURCE,
+    })
+  }, [setShowAccountSettingModal])
+
+  return (
+    <div className='flex flex-col gap-y-2'>
+      <Header
+        docTitle='Docs'
+        docLink={docLink('/guides/knowledge-base/knowledge-pipeline/authorize-data-source')}
+        onClickConfiguration={handleSetting}
+        pluginName={nodeData.datasource_label}
+        currentCredentialId={currentCredentialId}
+        onCredentialChange={onCredentialChange}
+        credentials={dataSourceAuth?.result || []}
+      />
+      <div className='rounded-xl border border-components-panel-border bg-background-default-subtle'>
+        <div className='flex items-center gap-x-2 rounded-t-xl border-b border-b-divider-regular bg-components-panel-bg p-1 pl-3'>
+          <div className='flex grow items-center'>
+            <Title name={nodeData.datasource_label} />
+          </div>
+          <SearchInput
+            value={searchValue}
+            onChange={handleSearchValueChange}
+          />
+        </div>
+        <div className='overflow-hidden rounded-b-xl'>
+          {documentsData?.length ? (
+            <PageSelector
+              checkedIds={selectedPagesId}
+              disabledValue={new Set()}
+              searchValue={searchValue}
+              list={documentsData[0].pages || []}
+              pagesMap={PagesMapAndSelectedPagesId}
+              onSelect={handleSelectPages}
+              canPreview={!isInPipeline}
+              onPreview={handlePreviewPage}
+              isMultipleChoice={!isInPipeline}
+              currentCredentialId={currentCredentialId}
+            />
+          ) : (
+            <div className='flex h-[296px] items-center justify-center'>
+              <Loading type='app' />
+            </div>
+          )}
+        </div>
+      </div>
+    </div>
+  )
+}
+
+export default OnlineDocuments

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-documents/page-selector/index.tsx

@@ -0,0 +1,190 @@
+import { useCallback, useEffect, useMemo, useState } from 'react'
+import { useTranslation } from 'react-i18next'
+import { FixedSizeList as List } from 'react-window'
+import type { DataSourceNotionPage, DataSourceNotionPageMap } from '@/models/common'
+import Item from './item'
+import { recursivePushInParentDescendants } from './utils'
+
+type PageSelectorProps = {
+  checkedIds: Set<string>
+  disabledValue: Set<string>
+  searchValue: string
+  pagesMap: DataSourceNotionPageMap
+  list: DataSourceNotionPage[]
+  onSelect: (selectedPagesId: Set<string>) => void
+  canPreview?: boolean
+  onPreview?: (selectedPageId: string) => void
+  isMultipleChoice?: boolean
+  currentCredentialId: string
+}
+
+export type NotionPageTreeItem = {
+  children: Set<string>
+  descendants: Set<string>
+  depth: number
+  ancestors: string[]
+} & DataSourceNotionPage
+
+export type NotionPageTreeMap = Record<string, NotionPageTreeItem>
+
+type NotionPageItem = {
+  expand: boolean
+  depth: number
+} & DataSourceNotionPage
+
+const PageSelector = ({
+  checkedIds,
+  disabledValue,
+  searchValue,
+  pagesMap,
+  list,
+  onSelect,
+  canPreview = true,
+  onPreview,
+  isMultipleChoice = true,
+  currentCredentialId,
+}: PageSelectorProps) => {
+  const { t } = useTranslation()
+  const [dataList, setDataList] = useState<NotionPageItem[]>([])
+  const [currentPreviewPageId, setCurrentPreviewPageId] = useState('')
+
+  useEffect(() => {
+    setDataList(list.filter(item => item.parent_id === 'root' || !pagesMap[item.parent_id]).map((item) => {
+      return {
+        ...item,
+        expand: false,
+        depth: 0,
+      }
+    }))
+  }, [currentCredentialId])
+
+  const searchDataList = list.filter((item) => {
+    return item.page_name.includes(searchValue)
+  }).map((item) => {
+    return {
+      ...item,
+      expand: false,
+      depth: 0,
+    }
+  })
+  const currentDataList = searchValue ? searchDataList : dataList
+
+  const listMapWithChildrenAndDescendants = useMemo(() => {
+    return list.reduce((prev: NotionPageTreeMap, next: DataSourceNotionPage) => {
+      const pageId = next.page_id
+      if (!prev[pageId])
+        prev[pageId] = { ...next, children: new Set(), descendants: new Set(), depth: 0, ancestors: [] }
+
+      recursivePushInParentDescendants(pagesMap, prev, prev[pageId], prev[pageId])
+      return prev
+    }, {})
+  }, [list, pagesMap])
+
+  const handleToggle = useCallback((index: number) => {
+    const current = dataList[index]
+    const pageId = current.page_id
+    const currentWithChildrenAndDescendants = listMapWithChildrenAndDescendants[pageId]
+    const descendantsIds = Array.from(currentWithChildrenAndDescendants.descendants)
+    const childrenIds = Array.from(currentWithChildrenAndDescendants.children)
+    let newDataList = []
+
+    if (current.expand) {
+      current.expand = false
+
+      newDataList = dataList.filter(item => !descendantsIds.includes(item.page_id))
+    }
+    else {
+      current.expand = true
+
+      newDataList = [
+        ...dataList.slice(0, index + 1),
+        ...childrenIds.map(item => ({
+          ...pagesMap[item],
+          expand: false,
+          depth: listMapWithChildrenAndDescendants[item].depth,
+        })),
+        ...dataList.slice(index + 1),
+      ]
+    }
+    setDataList(newDataList)
+  }, [dataList, listMapWithChildrenAndDescendants, pagesMap])
+
+  const handleCheck = useCallback((index: number) => {
+    const copyValue = new Set(checkedIds)
+    const current = currentDataList[index]
+    const pageId = current.page_id
+    const currentWithChildrenAndDescendants = listMapWithChildrenAndDescendants[pageId]
+
+    if (copyValue.has(pageId)) {
+      if (!searchValue && isMultipleChoice) {
+        for (const item of currentWithChildrenAndDescendants.descendants)
+          copyValue.delete(item)
+      }
+
+      copyValue.delete(pageId)
+    }
+    else {
+      if (!searchValue && isMultipleChoice) {
+        for (const item of currentWithChildrenAndDescendants.descendants)
+          copyValue.add(item)
+      }
+      // Single choice mode, clear previous selection
+      if (!isMultipleChoice && copyValue.size > 0) {
+        copyValue.clear()
+        copyValue.add(pageId)
+      }
+      else {
+        copyValue.add(pageId)
+      }
+    }
+
+    onSelect(new Set(copyValue))
+  }, [currentDataList, isMultipleChoice, listMapWithChildrenAndDescendants, onSelect, searchValue, checkedIds])
+
+  const handlePreview = useCallback((index: number) => {
+    const current = currentDataList[index]
+    const pageId = current.page_id
+
+    setCurrentPreviewPageId(pageId)
+
+    if (onPreview)
+      onPreview(pageId)
+  }, [currentDataList, onPreview])
+
+  if (!currentDataList.length) {
+    return (
+      <div className='flex h-[296px] items-center justify-center text-[13px] text-text-tertiary'>
+        {t('common.dataSource.notion.selector.noSearchResult')}
+      </div>
+    )
+  }
+
+  return (
+    <List
+      className='py-2'
+      height={296}
+      itemCount={currentDataList.length}
+      itemSize={28}
+      width='100%'
+      itemKey={(index, data) => data.dataList[index].page_id}
+      itemData={{
+        dataList: currentDataList,
+        handleToggle,
+        checkedIds,
+        disabledCheckedIds: disabledValue,
+        handleCheck,
+        canPreview,
+        handlePreview,
+        listMapWithChildrenAndDescendants,
+        searchValue,
+        previewPageId: currentPreviewPageId,
+        pagesMap,
+        isMultipleChoice,
+      }}
+    >
+      {Item}
+    </List>
+  )
+}
+
+export default PageSelector

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-documents/page-selector/item.tsx

@@ -0,0 +1,149 @@
+import React from 'react'
+import { useTranslation } from 'react-i18next'
+import { areEqual } from 'react-window'
+import type { ListChildComponentProps } from 'react-window'
+import { RiArrowDownSLine, RiArrowRightSLine } from '@remixicon/react'
+import Checkbox from '@/app/components/base/checkbox'
+import NotionIcon from '@/app/components/base/notion-icon'
+import cn from '@/utils/classnames'
+import type { DataSourceNotionPage, DataSourceNotionPageMap } from '@/models/common'
+import Radio from '@/app/components/base/radio/ui'
+
+type NotionPageTreeItem = {
+  children: Set<string>
+  descendants: Set<string>
+  depth: number
+  ancestors: string[]
+} & DataSourceNotionPage
+
+type NotionPageTreeMap = Record<string, NotionPageTreeItem>
+
+type NotionPageItem = {
+  expand: boolean
+  depth: number
+} & DataSourceNotionPage
+
+const Item = ({ index, style, data }: ListChildComponentProps<{
+  dataList: NotionPageItem[]
+  handleToggle: (index: number) => void
+  checkedIds: Set<string>
+  disabledCheckedIds: Set<string>
+  handleCheck: (index: number) => void
+  canPreview?: boolean
+  handlePreview: (index: number) => void
+  listMapWithChildrenAndDescendants: NotionPageTreeMap
+  searchValue: string
+  previewPageId: string
+  pagesMap: DataSourceNotionPageMap
+  isMultipleChoice?: boolean
+}>) => {
+  const { t } = useTranslation()
+  const {
+    dataList,
+    handleToggle,
+    checkedIds,
+    disabledCheckedIds,
+    handleCheck,
+    canPreview,
+    handlePreview,
+    listMapWithChildrenAndDescendants,
+    searchValue,
+    previewPageId,
+    pagesMap,
+    isMultipleChoice,
+  } = data
+  const current = dataList[index]
+  const currentWithChildrenAndDescendants = listMapWithChildrenAndDescendants[current.page_id]
+  const hasChild = currentWithChildrenAndDescendants.descendants.size > 0
+  const ancestors = currentWithChildrenAndDescendants.ancestors
+  const breadCrumbs = ancestors.length ? [...ancestors, current.page_name] : [current.page_name]
+  const disabled = disabledCheckedIds.has(current.page_id)
+
+  const renderArrow = () => {
+    if (hasChild) {
+      return (
+        <div
+          className='mr-1 flex h-5 w-5 shrink-0 items-center justify-center rounded-md hover:bg-components-button-ghost-bg-hover'
+          style={{ marginLeft: current.depth * 8 }}
+          onClick={() => handleToggle(index)}
+        >
+          {
+            current.expand
+              ? <RiArrowDownSLine className='h-4 w-4 text-text-tertiary' />
+              : <RiArrowRightSLine className='h-4 w-4 text-text-tertiary' />
+          }
+        </div>
+      )
+    }
+    if (current.parent_id === 'root' || !pagesMap[current.parent_id]) {
+      return (
+        <div></div>
+      )
+    }
+    return (
+      <div className='mr-1 h-5 w-5 shrink-0' style={{ marginLeft: current.depth * 8 }} />
+    )
+  }
+
+  return (
+    <div
+      className={cn('group flex cursor-pointer items-center rounded-md pl-2 pr-[2px] hover:bg-state-base-hover',
+        previewPageId === current.page_id && 'bg-state-base-hover')}
+      style={{ ...style, top: style.top as number + 8, left: 8, right: 8, width: 'calc(100% - 16px)' }}
+    >
+      {isMultipleChoice ? (
+        <Checkbox
+          className='mr-2 shrink-0'
+          checked={checkedIds.has(current.page_id)}
+          disabled={disabled}
+          onCheck={() => {
+            handleCheck(index)
+          }}
+        />) : (
+        <Radio
+          className='mr-2 shrink-0'
+          isChecked={checkedIds.has(current.page_id)}
+          disabled={disabled}
+          onCheck={() => {
+            handleCheck(index)
+          }}
+        />
+      )}
+      {!searchValue && renderArrow()}
+      <NotionIcon
+        className='mr-1 shrink-0'
+        type='page'
+        src={current.page_icon}
+      />
+      <div
+        className='grow truncate text-[13px] font-medium leading-4 text-text-secondary'
+        title={current.page_name}
+      >
+        {current.page_name}
+      </div>
+      {
+        canPreview && (
+          <div
+            className='ml-1 hidden h-6 shrink-0 cursor-pointer items-center rounded-md border-[0.5px] border-components-button-secondary-border bg-components-button-secondary-bg px-2 text-xs
+            font-medium leading-4 text-components-button-secondary-text shadow-xs shadow-shadow-shadow-3 backdrop-blur-[10px]
+            hover:border-components-button-secondary-border-hover hover:bg-components-button-secondary-bg-hover group-hover:flex'
+            onClick={() => handlePreview(index)}>
+            {t('common.dataSource.notion.selector.preview')}
+          </div>
+        )
+      }
+      {
+        searchValue && (
+          <div
+            className='ml-1 max-w-[120px] shrink-0 truncate text-xs text-text-quaternary'
+            title={breadCrumbs.join(' / ')}
+          >
+            {breadCrumbs.join(' / ')}
+          </div>
+        )
+      }
+    </div>
+  )
+}
+
+export default React.memo(Item, areEqual)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-documents/page-selector/utils.ts

@@ -0,0 +1,39 @@
+import type { DataSourceNotionPageMap } from '@/models/common'
+import type { NotionPageTreeItem, NotionPageTreeMap } from './index'
+
+export const recursivePushInParentDescendants = (
+  pagesMap: DataSourceNotionPageMap,
+  listTreeMap: NotionPageTreeMap,
+  current: NotionPageTreeItem,
+  leafItem: NotionPageTreeItem,
+) => {
+  const parentId = current.parent_id
+  const pageId = current.page_id
+
+  if (!parentId || !pageId)
+    return
+
+  if (parentId !== 'root' && pagesMap[parentId]) {
+    if (!listTreeMap[parentId]) {
+      const children = new Set([pageId])
+      const descendants = new Set([pageId, leafItem.page_id])
+      listTreeMap[parentId] = {
+        ...pagesMap[parentId],
+        children,
+        descendants,
+        depth: 0,
+        ancestors: [],
+      }
+    }
+    else {
+      listTreeMap[parentId].children.add(pageId)
+      listTreeMap[parentId].descendants.add(pageId)
+      listTreeMap[parentId].descendants.add(leafItem.page_id)
+    }
+    leafItem.depth++
+    leafItem.ancestors.unshift(listTreeMap[parentId].page_name)
+
+    if (listTreeMap[parentId].parent_id !== 'root')
+      recursivePushInParentDescendants(pagesMap, listTreeMap, listTreeMap[parentId], leafItem)
+  }
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-documents/title.tsx

@@ -0,0 +1,20 @@
+import React from 'react'
+import { useTranslation } from 'react-i18next'
+
+type TitleProps = {
+  name: string
+}
+
+const Title = ({
+  name,
+}: TitleProps) => {
+  const { t } = useTranslation()
+
+  return (
+    <div className='system-sm-medium px-[5px] py-1 text-text-secondary'>
+      {t('datasetPipeline.onlineDocument.pageSelectorTitle', { name })}
+    </div>
+  )
+}
+
+export default React.memo(Title)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/connect/index.tsx

@@ -0,0 +1,50 @@
+import { useTranslation } from 'react-i18next'
+import Button from '@/app/components/base/button'
+import { Icon3Dots } from '@/app/components/base/icons/src/vender/line/others'
+import BlockIcon from '@/app/components/workflow/block-icon'
+import { useToolIcon } from '@/app/components/workflow/hooks'
+import type { DataSourceNodeType } from '@/app/components/workflow/nodes/data-source/types'
+import { BlockEnum } from '@/app/components/workflow/types'
+
+type ConnectProps = {
+  nodeData: DataSourceNodeType
+  onSetting: () => void
+}
+
+const Connect = ({
+  nodeData,
+  onSetting,
+}: ConnectProps) => {
+  const { t } = useTranslation()
+  const toolIcon = useToolIcon(nodeData)
+
+  return (
+    <div className='flex flex-col items-start gap-y-2 rounded-xl bg-workflow-process-bg p-6'>
+      <div className='flex size-12 items-center justify-center rounded-[10px] border-[0.5px] border-components-card-border bg-components-card-bg p-1 shadow-lg shadow-shadow-shadow-5'>
+        <BlockIcon
+          type={BlockEnum.DataSource}
+          toolIcon={toolIcon}
+          size='md'
+        />
+      </div>
+      <div className='flex flex-col gap-y-1'>
+        <div className='flex flex-col gap-y-1 pb-3 pt-1'>
+          <div className='system-md-semibold text-text-secondary'>
+            <span className='relative'>
+              {t('datasetPipeline.onlineDrive.notConnected', { name: nodeData.title })}
+              <Icon3Dots className='absolute -right-2.5 -top-1.5 size-4 text-text-secondary' />
+            </span>
+          </div>
+          <div className='system-sm-regular text-text-tertiary'>
+            {t('datasetPipeline.onlineDrive.notConnectedTip', { name: nodeData.title })}
+          </div>
+        </div>
+        <Button className='w-fit' variant='primary' onClick={onSetting}>
+          {t('datasetCreation.stepOne.connect')}
+        </Button>
+      </div>
+    </div>
+  )
+}
+
+export default Connect

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/header/breadcrumbs/bucket.tsx

@@ -0,0 +1,62 @@
+import React, { useCallback } from 'react'
+import { BucketsGray } from '@/app/components/base/icons/src/public/knowledge/online-drive'
+import Tooltip from '@/app/components/base/tooltip'
+import { useTranslation } from 'react-i18next'
+import cn from '@/utils/classnames'
+
+type BucketProps = {
+  bucketName: string
+  isActive?: boolean
+  disabled?: boolean
+  showSeparator?: boolean
+  handleBackToBucketList: () => void
+  handleClickBucketName: () => void
+}
+
+const Bucket = ({
+  bucketName,
+  handleBackToBucketList,
+  handleClickBucketName,
+  disabled = false,
+  isActive = false,
+  showSeparator = true,
+}: BucketProps) => {
+  const { t } = useTranslation()
+  const handleClickItem = useCallback(() => {
+    if (!disabled)
+      handleClickBucketName()
+  }, [disabled, handleClickBucketName])
+
+  return (
+    <>
+      <Tooltip
+        popupContent={t('datasetPipeline.onlineDrive.breadcrumbs.allBuckets')}
+      >
+        <button
+          type='button'
+          className='flex size-6 shrink-0 cursor-pointer items-center justify-center rounded-md hover:bg-state-base-hover'
+          onClick={handleBackToBucketList}
+        >
+          <BucketsGray />
+        </button>
+      </Tooltip>
+      <span className='system-xs-regular text-divider-deep'>/</span>
+      <button
+        type='button'
+        className={cn(
+          'max-w-full shrink truncate rounded-md px-[5px] py-1',
+          isActive ? 'system-sm-medium text-text-secondary' : 'system-sm-regular text-text-tertiary',
+          !disabled && 'hover:bg-state-base-hover',
+        )}
+        disabled={disabled}
+        onClick={handleClickItem}
+        title={bucketName}
+      >
+        {bucketName}
+      </button>
+      {showSeparator && <span className='system-xs-regular shrink-0 text-divider-deep'>/</span>}
+    </>
+  )
+}
+
+export default React.memo(Bucket)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/header/breadcrumbs/drive.tsx

@@ -0,0 +1,35 @@
+import React from 'react'
+import cn from '@/utils/classnames'
+import { useTranslation } from 'react-i18next'
+
+type DriveProps = {
+  breadcrumbs: string[]
+  handleBackToRoot: () => void
+}
+
+const Drive = ({
+  breadcrumbs,
+  handleBackToRoot,
+}: DriveProps) => {
+  const { t } = useTranslation()
+
+  return (
+    <>
+      <button
+        type='button'
+        className={cn(
+          'max-w-full shrink truncate rounded-md px-[5px] py-1',
+          breadcrumbs.length > 0 && 'system-sm-regular text-text-tertiary hover:bg-state-base-hover',
+          breadcrumbs.length === 0 && 'system-sm-medium text-text-secondary',
+        )}
+        onClick={handleBackToRoot}
+        disabled={breadcrumbs.length === 0}
+      >
+        {t('datasetPipeline.onlineDrive.breadcrumbs.allFiles')}
+      </button>
+      {breadcrumbs.length > 0 && <span className='system-xs-regular text-divider-deep'>/</span>}
+    </>
+  )
+}
+
+export default React.memo(Drive)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/header/breadcrumbs/dropdown/index.tsx

@@ -0,0 +1,66 @@
+import React, { useCallback, useState } from 'react'
+import {
+  PortalToFollowElem,
+  PortalToFollowElemContent,
+  PortalToFollowElemTrigger,
+} from '@/app/components/base/portal-to-follow-elem'
+import { RiMoreFill } from '@remixicon/react'
+import cn from '@/utils/classnames'
+import Menu from './menu'
+
+type DropdownProps = {
+  startIndex: number
+  breadcrumbs: string[]
+  onBreadcrumbClick: (index: number) => void
+}
+
+const Dropdown = ({
+  startIndex,
+  breadcrumbs,
+  onBreadcrumbClick,
+}: DropdownProps) => {
+  const [open, setOpen] = useState(false)
+
+  const handleTrigger = useCallback(() => {
+    setOpen(prev => !prev)
+  }, [])
+
+  const handleBreadCrumbClick = useCallback((index: number) => {
+    onBreadcrumbClick(index)
+    setOpen(false)
+  }, [onBreadcrumbClick])
+
+  return (
+    <PortalToFollowElem
+      open={open}
+      onOpenChange={setOpen}
+      placement='bottom-start'
+      offset={{
+        mainAxis: 4,
+        crossAxis: -13,
+      }}
+    >
+      <PortalToFollowElemTrigger onClick={handleTrigger}>
+        <button
+          type='button'
+          className={cn(
+            'flex size-6 items-center justify-center rounded-md',
+            open ? 'bg-state-base-hover' : 'hover:bg-state-base-hover',
+          )}
+        >
+          <RiMoreFill className='size-4 text-text-tertiary' />
+        </button>
+      </PortalToFollowElemTrigger>
+      <PortalToFollowElemContent className='z-[11]'>
+        <Menu
+          breadcrumbs={breadcrumbs}
+          startIndex={startIndex}
+          onBreadcrumbClick={handleBreadCrumbClick}
+        />
+      </PortalToFollowElemContent>
+      <span className='system-xs-regular text-divider-deep'>/</span>
+    </PortalToFollowElem>
+  )
+}
+
+export default React.memo(Dropdown)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/header/breadcrumbs/dropdown/item.tsx

@@ -0,0 +1,28 @@
+import React, { useCallback } from 'react'
+
+type ItemProps = {
+  name: string
+  index: number
+  onBreadcrumbClick: (index: number) => void
+}
+
+const Item = ({
+  name,
+  index,
+  onBreadcrumbClick,
+}: ItemProps) => {
+  const handleClick = useCallback(() => {
+    onBreadcrumbClick(index)
+  }, [index, onBreadcrumbClick])
+
+  return (
+    <div
+      className='system-md-regular rounded-lg px-3 py-1.5 text-text-secondary hover:bg-state-base-hover'
+      onClick={handleClick}
+    >
+      {name}
+    </div>
+  )
+}
+
+export default React.memo(Item)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/header/breadcrumbs/dropdown/menu.tsx

@@ -0,0 +1,31 @@
+import React from 'react'
+import Item from './item'
+
+type MenuProps = {
+  breadcrumbs: string[]
+  startIndex: number
+  onBreadcrumbClick: (index: number) => void
+}
+
+const Menu = ({
+  breadcrumbs,
+  startIndex,
+  onBreadcrumbClick,
+}: MenuProps) => {
+  return (
+    <div className='flex w-[136px] flex-col rounded-xl border-[0.5px] border-components-panel-border bg-components-panel-bg-blur p-1 shadow-lg shadow-shadow-shadow-5 backdrop-blur-[5px]'>
+      {breadcrumbs.map((breadcrumb, index) => {
+        return (
+          <Item
+            key={`${breadcrumb}-${index}`}
+            name={breadcrumb}
+            index={startIndex + index}
+            onBreadcrumbClick={onBreadcrumbClick}
+          />
+        )
+      })}
+    </div>
+  )
+}
+
+export default React.memo(Menu)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/header/breadcrumbs/index.tsx

@@ -0,0 +1,166 @@
+import React, { useCallback, useMemo } from 'react'
+import { useTranslation } from 'react-i18next'
+import { useDataSourceStore, useDataSourceStoreWithSelector } from '../../../../store'
+import Bucket from './bucket'
+import BreadcrumbItem from './item'
+import Dropdown from './dropdown'
+import Drive from './drive'
+
+type BreadcrumbsProps = {
+  breadcrumbs: string[]
+  keywords: string
+  bucket: string
+  searchResultsLength: number
+  isInPipeline: boolean
+}
+
+const Breadcrumbs = ({
+  breadcrumbs,
+  keywords,
+  bucket,
+  searchResultsLength,
+  isInPipeline,
+}: BreadcrumbsProps) => {
+  const { t } = useTranslation()
+  const dataSourceStore = useDataSourceStore()
+  const hasBucket = useDataSourceStoreWithSelector(s => s.hasBucket)
+  const showSearchResult = !!keywords && searchResultsLength > 0
+  const showBucketListTitle = breadcrumbs.length === 0 && hasBucket && bucket === ''
+
+  const displayBreadcrumbNum = useMemo(() => {
+    const num = isInPipeline ? 2 : 3
+    return bucket ? num - 1 : num
+  }, [isInPipeline, bucket])
+
+  const breadcrumbsConfig = useMemo(() => {
+    const prefixToDisplay = breadcrumbs.slice(0, displayBreadcrumbNum - 1)
+    const collapsedBreadcrumbs = breadcrumbs.slice(displayBreadcrumbNum - 1, breadcrumbs.length - 1)
+    return {
+      original: breadcrumbs,
+      needCollapsed: breadcrumbs.length > displayBreadcrumbNum,
+      prefixBreadcrumbs: prefixToDisplay,
+      collapsedBreadcrumbs,
+      lastBreadcrumb: breadcrumbs[breadcrumbs.length - 1],
+    }
+  }, [displayBreadcrumbNum, breadcrumbs])
+
+  const handleBackToBucketList = useCallback(() => {
+    const { setOnlineDriveFileList, setSelectedFileIds, setBreadcrumbs, setPrefix, setBucket } = dataSourceStore.getState()
+    setOnlineDriveFileList([])
+    setSelectedFileIds([])
+    setBucket('')
+    setBreadcrumbs([])
+    setPrefix([])
+  }, [dataSourceStore])
+
+  const handleClickBucketName = useCallback(() => {
+    const { setOnlineDriveFileList, setSelectedFileIds, setBreadcrumbs, setPrefix } = dataSourceStore.getState()
+    setOnlineDriveFileList([])
+    setSelectedFileIds([])
+    setBreadcrumbs([])
+    setPrefix([])
+  }, [dataSourceStore])
+
+  const handleBackToRoot = useCallback(() => {
+    const { setOnlineDriveFileList, setSelectedFileIds, setBreadcrumbs, setPrefix } = dataSourceStore.getState()
+    setOnlineDriveFileList([])
+    setSelectedFileIds([])
+    setBreadcrumbs([])
+    setPrefix([])
+  }, [dataSourceStore])
+
+  const handleClickBreadcrumb = useCallback((index: number) => {
+    const { breadcrumbs, prefix, setOnlineDriveFileList, setSelectedFileIds, setBreadcrumbs, setPrefix } = dataSourceStore.getState()
+    const newBreadcrumbs = breadcrumbs.slice(0, index + 1)
+    const newPrefix = prefix.slice(0, index + 1)
+    setOnlineDriveFileList([])
+    setSelectedFileIds([])
+    setBreadcrumbs(newBreadcrumbs)
+    setPrefix(newPrefix)
+  }, [dataSourceStore])
+
+  return (
+    <div className='flex grow items-center overflow-hidden'>
+      {showSearchResult && (
+        <div className='system-sm-medium text-test-secondary px-[5px]'>
+          {t('datasetPipeline.onlineDrive.breadcrumbs.searchResult', {
+            searchResultsLength,
+            folderName: breadcrumbs.length > 0 ? breadcrumbs[breadcrumbs.length - 1] : bucket,
+          })}
+        </div>
+      )}
+      {!showSearchResult && showBucketListTitle && (
+        <div className='system-sm-medium text-test-secondary px-[5px]'>
+          {t('datasetPipeline.onlineDrive.breadcrumbs.allBuckets')}
+        </div>
+      )}
+      {!showSearchResult && !showBucketListTitle && (
+        <div className='flex w-full items-center gap-x-0.5 overflow-hidden'>
+          {hasBucket && bucket && (
+            <Bucket
+              bucketName={bucket}
+              handleBackToBucketList={handleBackToBucketList}
+              handleClickBucketName={handleClickBucketName}
+              isActive={breadcrumbs.length === 0}
+              disabled={breadcrumbs.length === 0}
+              showSeparator={breadcrumbs.length > 0}
+            />
+          )}
+          {!hasBucket && (
+            <Drive
+              breadcrumbs={breadcrumbs}
+              handleBackToRoot={handleBackToRoot}
+            />
+          )}
+          {!breadcrumbsConfig.needCollapsed && (
+            <>
+              {breadcrumbsConfig.original.map((breadcrumb, index) => {
+                const isLast = index === breadcrumbsConfig.original.length - 1
+                return (
+                  <BreadcrumbItem
+                    key={`${breadcrumb}-${index}`}
+                    index={index}
+                    handleClick={handleClickBreadcrumb}
+                    name={breadcrumb}
+                    isActive={isLast}
+                    showSeparator={!isLast}
+                    disabled={isLast}
+                  />
+                )
+              })}
+            </>
+          )}
+          {breadcrumbsConfig.needCollapsed && (
+            <>
+              {breadcrumbsConfig.prefixBreadcrumbs.map((breadcrumb, index) => {
+                return (
+                  <BreadcrumbItem
+                    key={`${breadcrumb}-${index}`}
+                    index={index}
+                    handleClick={handleClickBreadcrumb}
+                    name={breadcrumb}
+                  />
+                )
+              })}
+              <Dropdown
+                startIndex={breadcrumbsConfig.prefixBreadcrumbs.length}
+                breadcrumbs={breadcrumbsConfig.collapsedBreadcrumbs}
+                onBreadcrumbClick={handleClickBreadcrumb}
+              />
+              <BreadcrumbItem
+                index={breadcrumbs.length - 1}
+                handleClick={handleClickBreadcrumb}
+                name={breadcrumbsConfig.lastBreadcrumb}
+                isActive={true}
+                disabled={true}
+                showSeparator={false}
+              />
+            </>
+          )}
+        </div>
+      )}
+    </div>
+  )
+}
+
+export default React.memo(Breadcrumbs)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/header/breadcrumbs/item.tsx

@@ -0,0 +1,48 @@
+import React, { useCallback } from 'react'
+import cn from '@/utils/classnames'
+
+type BreadcrumbItemProps = {
+  name: string
+  index: number
+  handleClick: (index: number) => void
+  disabled?: boolean
+  isActive?: boolean
+  showSeparator?: boolean
+}
+
+const BreadcrumbItem = ({
+  name,
+  index,
+  handleClick,
+  disabled = false,
+  isActive = false,
+  showSeparator = true,
+}: BreadcrumbItemProps) => {
+  const handleClickItem = useCallback(() => {
+    if (!disabled)
+      handleClick(index)
+  }, [disabled, handleClick, index])
+
+  return (
+    <>
+      <button
+        type='button'
+        className={cn(
+          'max-w-full shrink truncate rounded-md px-[5px] py-1',
+          isActive ? 'system-sm-medium text-text-secondary' : 'system-sm-regular text-text-tertiary',
+          !disabled && 'hover:bg-state-base-hover',
+        )}
+        disabled={disabled}
+        onClick={handleClickItem}
+        title={name}
+      >
+        {name}
+      </button>
+      {showSeparator && <span className='system-xs-regular shrink-0 text-divider-deep'>/</span>}
+    </>
+  )
+}
+
+BreadcrumbItem.displayName = 'BreadcrumbItem'
+
+export default React.memo(BreadcrumbItem)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/header/index.tsx

@@ -0,0 +1,51 @@
+import React from 'react'
+import Breadcrumbs from './breadcrumbs'
+import Input from '@/app/components/base/input'
+import { useTranslation } from 'react-i18next'
+
+type HeaderProps = {
+  breadcrumbs: string[]
+  inputValue: string
+  keywords: string
+  bucket: string
+  searchResultsLength: number
+  handleInputChange: React.ChangeEventHandler<HTMLInputElement>
+  handleResetKeywords: () => void
+  isInPipeline: boolean
+}
+
+const Header = ({
+  breadcrumbs,
+  inputValue,
+  keywords,
+  bucket,
+  isInPipeline,
+  searchResultsLength,
+  handleInputChange,
+  handleResetKeywords,
+}: HeaderProps) => {
+  const { t } = useTranslation()
+
+  return (
+    <div className='flex items-center gap-x-2 bg-components-panel-bg p-1 pl-3'>
+      <Breadcrumbs
+        breadcrumbs={breadcrumbs}
+        keywords={keywords}
+        bucket={bucket}
+        searchResultsLength={searchResultsLength}
+        isInPipeline={isInPipeline}
+      />
+      <Input
+        value={inputValue}
+        onChange={handleInputChange}
+        onClear={handleResetKeywords}
+        placeholder={t('datasetPipeline.onlineDrive.breadcrumbs.searchPlaceholder')}
+        showLeftIcon
+        showClearIcon
+        wrapperClassName='w-[200px] h-8 shrink-0'
+      />
+    </div>
+  )
+}
+
+export default React.memo(Header)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/index.tsx

@@ -0,0 +1,82 @@
+import type { OnlineDriveFile } from '@/models/pipeline'
+import Header from './header'
+import List from './list'
+import { useState } from 'react'
+import { useDebounceFn } from 'ahooks'
+
+type FileListProps = {
+  fileList: OnlineDriveFile[]
+  selectedFileIds: string[]
+  breadcrumbs: string[]
+  keywords: string
+  bucket: string
+  isInPipeline: boolean
+  resetKeywords: () => void
+  updateKeywords: (keywords: string) => void
+  searchResultsLength: number
+  handleSelectFile: (file: OnlineDriveFile) => void
+  handleOpenFolder: (file: OnlineDriveFile) => void
+  isLoading: boolean
+}
+
+const FileList = ({
+  fileList,
+  selectedFileIds,
+  breadcrumbs,
+  keywords,
+  bucket,
+  resetKeywords,
+  updateKeywords,
+  searchResultsLength,
+  handleSelectFile,
+  handleOpenFolder,
+  isInPipeline,
+  isLoading,
+}: FileListProps) => {
+  const [inputValue, setInputValue] = useState(keywords)
+
+  const { run: updateKeywordsWithDebounce } = useDebounceFn(
+    (keywords: string) => {
+      updateKeywords(keywords)
+    },
+    { wait: 500 },
+  )
+
+  const handleInputChange = (e: React.ChangeEvent<HTMLInputElement>) => {
+    const keywords = e.target.value
+    setInputValue(keywords)
+    updateKeywordsWithDebounce(keywords)
+  }
+
+  const handleResetKeywords = () => {
+    setInputValue('')
+    resetKeywords()
+  }
+
+  return (
+    <div className='flex h-[400px] flex-col overflow-hidden rounded-xl border border-components-panel-border bg-components-panel-bg shadow-xs shadow-shadow-shadow-3'>
+      <Header
+        breadcrumbs={breadcrumbs}
+        inputValue={inputValue}
+        keywords={keywords}
+        bucket={bucket}
+        isInPipeline={isInPipeline}
+        handleInputChange={handleInputChange}
+        searchResultsLength={searchResultsLength}
+        handleResetKeywords={handleResetKeywords}
+      />
+      <List
+        fileList={fileList}
+        selectedFileIds={selectedFileIds}
+        keywords={keywords}
+        handleResetKeywords={handleResetKeywords}
+        handleOpenFolder={handleOpenFolder}
+        handleSelectFile={handleSelectFile}
+        isInPipeline={isInPipeline}
+        isLoading={isLoading}
+      />
+    </div>
+  )
+}
+
+export default FileList

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/list/empty-folder.tsx

@@ -0,0 +1,14 @@
+import React from 'react'
+import { useTranslation } from 'react-i18next'
+
+const EmptyFolder = () => {
+  const { t } = useTranslation()
+
+  return (
+    <div className='flex size-full items-center justify-center rounded-[10px] bg-background-section px-1 py-1.5'>
+      <span className='system-xs-regular text-text-tertiary'>{t('datasetPipeline.onlineDrive.emptyFolder')}</span>
+    </div>
+  )
+}
+
+export default React.memo(EmptyFolder)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/list/empty-search-result.tsx

@@ -0,0 +1,35 @@
+import Button from '@/app/components/base/button'
+import { SearchMenu } from '@/app/components/base/icons/src/vender/knowledge'
+import React from 'react'
+import { useTranslation } from 'react-i18next'
+
+type EmptySearchResultProps = {
+  onResetKeywords: () => void
+}
+
+const EmptySearchResult = ({
+  onResetKeywords,
+}: EmptySearchResultProps & {
+  className?: string
+}) => {
+  const { t } = useTranslation()
+
+  return (
+    <div className='flex size-full flex-col items-center justify-center gap-y-2 rounded-[10px] bg-background-section p-6'>
+      <SearchMenu className='size-8 text-text-tertiary' />
+      <div className='system-sm-regular text-text-secondary'>
+        {t('datasetPipeline.onlineDrive.emptySearchResult')}
+      </div>
+      <Button
+        variant='secondary-accent'
+        size='small'
+        onClick={onResetKeywords}
+        className='px-1.5'
+      >
+        <span className='px-[3px]'>{t('datasetPipeline.onlineDrive.resetKeywords')}</span>
+      </Button>
+    </div>
+  )
+}
+
+export default React.memo(EmptySearchResult)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/list/file-icon.tsx

@@ -0,0 +1,49 @@
+import React, { useMemo } from 'react'
+import { OnlineDriveFileType } from '@/models/pipeline'
+import { BucketsBlue, Folder } from '@/app/components/base/icons/src/public/knowledge/online-drive'
+import FileTypeIcon from '@/app/components/base/file-uploader/file-type-icon'
+import { getFileType } from './utils'
+import cn from '@/utils/classnames'
+
+type FileIconProps = {
+  type: OnlineDriveFileType
+  fileName: string
+  size?: 'sm' | 'md' | 'lg' | 'xl'
+  className?: string
+}
+
+const FileIcon = ({
+  type,
+  fileName,
+  size = 'md',
+  className,
+}: FileIconProps) => {
+  const fileType = useMemo(() => {
+    if (type === OnlineDriveFileType.bucket || type === OnlineDriveFileType.folder)
+      return 'custom'
+
+    return getFileType(fileName)
+  }, [type, fileName])
+
+  if (type === OnlineDriveFileType.bucket) {
+    return (
+      <BucketsBlue className={cn('size-[18px]', className)} />
+    )
+  }
+
+  if (type === OnlineDriveFileType.folder) {
+    return (
+      <Folder className={cn('size-[18px]', className)} />
+    )
+  }
+
+  return (
+    <FileTypeIcon
+      size={size}
+      type={fileType}
+      className={cn('size-[18px]', className)}
+    />
+  )
+}
+
+export default React.memo(FileIcon)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/list/index.tsx

@@ -0,0 +1,102 @@
+import React, { useEffect, useRef } from 'react'
+import type { OnlineDriveFile } from '@/models/pipeline'
+import Item from './item'
+import EmptyFolder from './empty-folder'
+import EmptySearchResult from './empty-search-result'
+import Loading from '@/app/components/base/loading'
+import { RiLoader2Line } from '@remixicon/react'
+import { useDataSourceStore } from '../../../store'
+
+type FileListProps = {
+  fileList: OnlineDriveFile[]
+  selectedFileIds: string[]
+  keywords: string
+  isInPipeline: boolean
+  isLoading: boolean
+  handleResetKeywords: () => void
+  handleSelectFile: (file: OnlineDriveFile) => void
+  handleOpenFolder: (file: OnlineDriveFile) => void
+}
+
+const List = ({
+  fileList,
+  selectedFileIds,
+  keywords,
+  handleResetKeywords,
+  handleSelectFile,
+  handleOpenFolder,
+  isInPipeline,
+  isLoading,
+}: FileListProps) => {
+  const anchorRef = useRef<HTMLDivElement>(null)
+  const observerRef = useRef<IntersectionObserver>(null)
+  const dataSourceStore = useDataSourceStore()
+
+  useEffect(() => {
+    if (anchorRef.current) {
+      observerRef.current = new IntersectionObserver((entries) => {
+        const { setNextPageParameters, currentNextPageParametersRef, isTruncated } = dataSourceStore.getState()
+        if (entries[0].isIntersecting && isTruncated.current && !isLoading)
+          setNextPageParameters(currentNextPageParametersRef.current)
+      }, {
+        rootMargin: '100px',
+      })
+      observerRef.current.observe(anchorRef.current)
+    }
+    return () => observerRef.current?.disconnect()
+  }, [anchorRef, isLoading, dataSourceStore])
+
+  const isAllLoading = isLoading && fileList.length === 0 && keywords.length === 0
+  const isPartialLoading = isLoading && fileList.length > 0
+  const isEmptyFolder = !isLoading && fileList.length === 0 && keywords.length === 0
+  const isSearchResultEmpty = !isLoading && fileList.length === 0 && keywords.length > 0
+
+  return (
+    <div className='grow overflow-hidden p-1 pt-0'>
+      {
+        isAllLoading && (
+          <Loading type='app' />
+        )
+      }
+      {
+        isEmptyFolder && (
+          <EmptyFolder />
+        )
+      }
+      {
+        isSearchResultEmpty && (
+          <EmptySearchResult onResetKeywords={handleResetKeywords} />
+        )
+      }
+      {fileList.length > 0 && (
+        <div className='flex h-full flex-col gap-y-px overflow-y-auto rounded-[10px] bg-background-section px-1 py-1.5'>
+          {
+            fileList.map((file) => {
+              const isSelected = selectedFileIds.includes(file.id)
+              return (
+                <Item
+                  key={file.id}
+                  file={file}
+                  isSelected={isSelected}
+                  onSelect={handleSelectFile}
+                  onOpen={handleOpenFolder}
+                  isMultipleChoice={!isInPipeline}
+                />
+              )
+            })
+          }
+          {
+            isPartialLoading && (
+              <div className='flex items-center justify-center py-2'>
+                <RiLoader2Line className='animation-spin size-4 text-text-tertiary' />
+              </div>
+            )
+          }
+          <div ref={anchorRef} className='h-0' />
+        </div>
+      )}
+    </div>
+  )
+}
+
+export default React.memo(List)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/list/item.tsx

@@ -0,0 +1,103 @@
+import Checkbox from '@/app/components/base/checkbox'
+import Radio from '@/app/components/base/radio/ui'
+import type { OnlineDriveFile } from '@/models/pipeline'
+import React, { useCallback } from 'react'
+import FileIcon from './file-icon'
+import { formatFileSize } from '@/utils/format'
+import Tooltip from '@/app/components/base/tooltip'
+import { useTranslation } from 'react-i18next'
+import cn from '@/utils/classnames'
+import type { Placement } from '@floating-ui/react'
+
+type ItemProps = {
+  file: OnlineDriveFile
+  isSelected: boolean
+  disabled?: boolean
+  isMultipleChoice?: boolean
+  onSelect: (file: OnlineDriveFile) => void
+  onOpen: (file: OnlineDriveFile) => void
+}
+
+const Item = ({
+  file,
+  isSelected,
+  disabled = false,
+  isMultipleChoice = true,
+  onSelect,
+  onOpen,
+}: ItemProps) => {
+  const { t } = useTranslation()
+  const { id, name, type, size } = file
+
+  const isBucket = type === 'bucket'
+  const isFolder = type === 'folder'
+
+  const Wrapper = disabled ? Tooltip : React.Fragment
+  const wrapperProps = disabled ? {
+    popupContent: t('datasetPipeline.onlineDrive.notSupportedFileType'),
+    position: 'top-end' as Placement,
+    offset: { mainAxis: 4, crossAxis: -104 },
+  } : {}
+
+  const handleSelect = useCallback((e: React.MouseEvent<HTMLDivElement>) => {
+    e.stopPropagation()
+    onSelect(file)
+  }, [file, onSelect])
+
+  const handleClickItem = useCallback((e: React.MouseEvent<HTMLDivElement>) => {
+    e.stopPropagation()
+    if (disabled) return
+    if (isBucket || isFolder) {
+      onOpen(file)
+      return
+    }
+    onSelect(file)
+  }, [disabled, file, isBucket, isFolder, onOpen, onSelect])
+
+  return (
+    <div
+      className='flex cursor-pointer items-center gap-2 rounded-md px-2 py-[3px] hover:bg-state-base-hover'
+      onClick={handleClickItem}
+    >
+      {!isBucket && isMultipleChoice && (
+        <Checkbox
+          className='shrink-0'
+          disabled={disabled}
+          id={id}
+          checked={isSelected}
+          onCheck={handleSelect}
+        />
+      )}
+      {!isBucket && !isMultipleChoice && (
+        <Radio
+          className='shrink-0'
+          disabled={disabled}
+          isChecked={isSelected}
+          onCheck={handleSelect}
+        />
+      )}
+      <Wrapper
+        {...wrapperProps}
+      >
+        <div
+          className={cn(
+            'flex grow items-center gap-x-1 overflow-hidden py-0.5',
+            disabled && 'opacity-30',
+          )}>
+          <FileIcon type={type} fileName={name} className='shrink-0 transform-gpu' />
+          <span
+            className='system-sm-medium grow truncate text-text-secondary'
+            title={name}
+          >
+            {name}
+          </span>
+          {!isFolder && typeof size === 'number' && (
+            <span className='system-xs-regular shrink-0 text-text-tertiary'>{formatFileSize(size)}</span>
+          )}
+        </div>
+      </Wrapper>
+    </div>
+  )
+}
+
+export default React.memo(Item)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/file-list/list/utils.ts

@@ -0,0 +1,51 @@
+import { FileAppearanceTypeEnum } from '@/app/components/base/file-uploader/types'
+import { FILE_EXTS } from '@/app/components/base/prompt-editor/constants'
+
+export const getFileExtension = (fileName: string): string => {
+  if (!fileName)
+    return ''
+  const parts = fileName.split('.')
+  if (parts.length <= 1 || (parts[0] === '' && parts.length === 2))
+    return ''
+
+  return parts[parts.length - 1].toLowerCase()
+}
+
+export const getFileType = (fileName: string) => {
+  const extension = getFileExtension(fileName)
+
+  if (extension === 'gif')
+    return FileAppearanceTypeEnum.gif
+
+  if (FILE_EXTS.image.includes(extension.toUpperCase()))
+    return FileAppearanceTypeEnum.image
+
+  if (FILE_EXTS.video.includes(extension.toUpperCase()))
+    return FileAppearanceTypeEnum.video
+
+  if (FILE_EXTS.audio.includes(extension.toUpperCase()))
+    return FileAppearanceTypeEnum.audio
+
+  if (extension === 'html' || extension === 'htm' || extension === 'xml' || extension === 'json')
+    return FileAppearanceTypeEnum.code
+
+  if (extension === 'pdf')
+    return FileAppearanceTypeEnum.pdf
+
+  if (extension === 'md' || extension === 'markdown' || extension === 'mdx')
+    return FileAppearanceTypeEnum.markdown
+
+  if (extension === 'xlsx' || extension === 'xls' || extension === 'csv')
+    return FileAppearanceTypeEnum.excel
+
+  if (extension === 'docx' || extension === 'doc')
+    return FileAppearanceTypeEnum.word
+
+  if (extension === 'pptx' || extension === 'ppt')
+    return FileAppearanceTypeEnum.ppt
+
+  if (FILE_EXTS.document.includes(extension.toUpperCase()))
+    return FileAppearanceTypeEnum.document
+
+  return FileAppearanceTypeEnum.custom
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/header.tsx

@@ -0,0 +1,48 @@
+import React from 'react'
+import Button from '@/app/components/base/button'
+import Divider from '@/app/components/base/divider'
+import { RiBookOpenLine, RiEqualizer2Line } from '@remixicon/react'
+
+type HeaderProps = {
+  onClickConfiguration?: () => void
+  docTitle: string
+  docLink: string
+}
+
+const Header = ({
+  onClickConfiguration,
+  docTitle,
+  docLink,
+}: HeaderProps) => {
+  return (
+    <div className='flex items-center gap-x-2'>
+      <div className='flex shrink-0 grow items-center gap-x-1'>
+        <div className='w-20 bg-black'>
+          {/* placeholder */}
+        </div>
+        <Divider type='vertical' className='mx-1 h-3.5' />
+        <Button
+          variant='ghost'
+          size='small'
+          className='px-1'
+        >
+          <RiEqualizer2Line
+            className='size-4'
+            onClick={onClickConfiguration}
+          />
+        </Button>
+      </div>
+      <a
+        className='system-xs-medium flex items-center gap-x-1 overflow-hidden text-text-accent'
+        href={docLink}
+        target='_blank'
+        rel='noopener noreferrer'
+      >
+        <RiBookOpenLine className='size-3.5 shrink-0' />
+        <span className='grow truncate' title={docTitle}>{docTitle}</span>
+      </a>
+    </div>
+  )
+}
+
+export default React.memo(Header)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/index.tsx

@@ -0,0 +1,217 @@
+import type { DataSourceNodeType } from '@/app/components/workflow/nodes/data-source/types'
+import Header from '../base/header'
+import { useCallback, useEffect, useMemo, useRef, useState } from 'react'
+import FileList from './file-list'
+import type { OnlineDriveFile } from '@/models/pipeline'
+import { DatasourceType, OnlineDriveFileType } from '@/models/pipeline'
+import { useDatasetDetailContextWithSelector } from '@/context/dataset-detail'
+import { ssePost } from '@/service/base'
+import type { DataSourceNodeCompletedResponse, DataSourceNodeErrorResponse } from '@/types/pipeline'
+import Toast from '@/app/components/base/toast'
+import { useDataSourceStore, useDataSourceStoreWithSelector } from '../store'
+import { convertOnlineDriveData } from './utils'
+import { produce } from 'immer'
+import { useShallow } from 'zustand/react/shallow'
+import { useModalContextSelector } from '@/context/modal-context'
+import { useGetDataSourceAuth } from '@/service/use-datasource'
+import { useDocLink } from '@/context/i18n'
+import { ACCOUNT_SETTING_TAB } from '@/app/components/header/account-setting/constants'
+
+type OnlineDriveProps = {
+  nodeId: string
+  nodeData: DataSourceNodeType
+  isInPipeline?: boolean
+  onCredentialChange: (credentialId: string) => void
+}
+
+const OnlineDrive = ({
+  nodeId,
+  nodeData,
+  isInPipeline = false,
+  onCredentialChange,
+}: OnlineDriveProps) => {
+  const docLink = useDocLink()
+  const [isInitialMount, setIsInitialMount] = useState(true)
+  const pipelineId = useDatasetDetailContextWithSelector(s => s.dataset?.pipeline_id)
+  const setShowAccountSettingModal = useModalContextSelector(s => s.setShowAccountSettingModal)
+  const {
+    nextPageParameters,
+    breadcrumbs,
+    prefix,
+    keywords,
+    bucket,
+    selectedFileIds,
+    onlineDriveFileList,
+    currentCredentialId,
+  } = useDataSourceStoreWithSelector(useShallow(state => ({
+    nextPageParameters: state.nextPageParameters,
+    breadcrumbs: state.breadcrumbs,
+    prefix: state.prefix,
+    keywords: state.keywords,
+    bucket: state.bucket,
+    selectedFileIds: state.selectedFileIds,
+    onlineDriveFileList: state.onlineDriveFileList,
+    currentCredentialId: state.currentCredentialId,
+  })))
+  const dataSourceStore = useDataSourceStore()
+  const [isLoading, setIsLoading] = useState(false)
+  const isLoadingRef = useRef(false)
+
+  const { data: dataSourceAuth } = useGetDataSourceAuth({
+    pluginId: nodeData.plugin_id,
+    provider: nodeData.provider_name,
+  })
+
+  const datasourceNodeRunURL = !isInPipeline
+    ? `/rag/pipelines/${pipelineId}/workflows/published/datasource/nodes/${nodeId}/run`
+    : `/rag/pipelines/${pipelineId}/workflows/draft/datasource/nodes/${nodeId}/run`
+
+  const getOnlineDriveFiles = useCallback(async () => {
+    if (isLoadingRef.current) return
+    const { nextPageParameters, prefix, bucket, onlineDriveFileList, currentCredentialId } = dataSourceStore.getState()
+    setIsLoading(true)
+    isLoadingRef.current = true
+    ssePost(
+      datasourceNodeRunURL,
+      {
+        body: {
+          inputs: {
+            prefix: prefix[prefix.length - 1],
+            bucket,
+            next_page_parameters: nextPageParameters,
+            max_keys: 30,
+          },
+          datasource_type: DatasourceType.onlineDrive,
+          credential_id: currentCredentialId,
+        },
+      },
+      {
+        onDataSourceNodeCompleted: (documentsData: DataSourceNodeCompletedResponse) => {
+          const { setOnlineDriveFileList, isTruncated, currentNextPageParametersRef, setHasBucket } = dataSourceStore.getState()
+          const {
+            fileList: newFileList,
+            isTruncated: newIsTruncated,
+            nextPageParameters: newNextPageParameters,
+            hasBucket: newHasBucket,
+          } = convertOnlineDriveData(documentsData.data, breadcrumbs, bucket)
+          setOnlineDriveFileList([...onlineDriveFileList, ...newFileList])
+          isTruncated.current = newIsTruncated
+          currentNextPageParametersRef.current = newNextPageParameters
+          setHasBucket(newHasBucket)
+          setIsLoading(false)
+          isLoadingRef.current = false
+        },
+        onDataSourceNodeError: (error: DataSourceNodeErrorResponse) => {
+          Toast.notify({
+            type: 'error',
+            message: error.error,
+          })
+          setIsLoading(false)
+          isLoadingRef.current = false
+        },
+      },
+    )
+  }, [datasourceNodeRunURL, dataSourceStore])
+
+  useEffect(() => {
+    if (!currentCredentialId) return
+    if (isInitialMount) {
+      // Only fetch files on initial mount if fileList is empty
+      if (onlineDriveFileList.length === 0)
+        getOnlineDriveFiles()
+      setIsInitialMount(false)
+    }
+    else {
+      getOnlineDriveFiles()
+    }
+  }, [nextPageParameters, prefix, bucket, currentCredentialId])
+
+  const filteredOnlineDriveFileList = useMemo(() => {
+    if (keywords)
+      return onlineDriveFileList.filter(file => file.name.toLowerCase().includes(keywords.toLowerCase()))
+    return onlineDriveFileList
+  }, [onlineDriveFileList, keywords])
+
+  const updateKeywords = useCallback((keywords: string) => {
+    const { setKeywords } = dataSourceStore.getState()
+    setKeywords(keywords)
+  }, [dataSourceStore])
+
+  const resetKeywords = useCallback(() => {
+    const { setKeywords } = dataSourceStore.getState()
+
+    setKeywords('')
+  }, [dataSourceStore])
+
+  const handleSelectFile = useCallback((file: OnlineDriveFile) => {
+    const { selectedFileIds, setSelectedFileIds } = dataSourceStore.getState()
+    if (file.type === OnlineDriveFileType.bucket) return
+    const newSelectedFileList = produce(selectedFileIds, (draft) => {
+      if (draft.includes(file.id)) {
+        const index = draft.indexOf(file.id)
+        draft.splice(index, 1)
+      }
+      else {
+        if (isInPipeline && draft.length >= 1) return
+        draft.push(file.id)
+      }
+    })
+    setSelectedFileIds(newSelectedFileList)
+  }, [dataSourceStore, isInPipeline])
+
+  const handleOpenFolder = useCallback((file: OnlineDriveFile) => {
+    const { breadcrumbs, prefix, setBreadcrumbs, setPrefix, setBucket, setOnlineDriveFileList, setSelectedFileIds } = dataSourceStore.getState()
+    if (file.type === OnlineDriveFileType.file) return
+    setOnlineDriveFileList([])
+    if (file.type === OnlineDriveFileType.bucket) {
+      setBucket(file.name)
+    }
+    else {
+      setSelectedFileIds([])
+      const newBreadcrumbs = produce(breadcrumbs, (draft) => {
+        draft.push(file.name)
+      })
+      const newPrefix = produce(prefix, (draft) => {
+        draft.push(file.id)
+      })
+      setBreadcrumbs(newBreadcrumbs)
+      setPrefix(newPrefix)
+    }
+  }, [dataSourceStore, getOnlineDriveFiles])
+
+  const handleSetting = useCallback(() => {
+    setShowAccountSettingModal({
+      payload: ACCOUNT_SETTING_TAB.DATA_SOURCE,
+    })
+  }, [setShowAccountSettingModal])
+
+  return (
+    <div className='flex flex-col gap-y-2'>
+      <Header
+        docTitle='Docs'
+        docLink={docLink('/guides/knowledge-base/knowledge-pipeline/authorize-data-source')}
+        onClickConfiguration={handleSetting}
+        pluginName={nodeData.datasource_label}
+        currentCredentialId={currentCredentialId}
+        onCredentialChange={onCredentialChange}
+        credentials={dataSourceAuth?.result || []}
+      />
+      <FileList
+        fileList={filteredOnlineDriveFileList}
+        selectedFileIds={selectedFileIds}
+        breadcrumbs={breadcrumbs}
+        keywords={keywords}
+        bucket={bucket}
+        resetKeywords={resetKeywords}
+        updateKeywords={updateKeywords}
+        searchResultsLength={filteredOnlineDriveFileList.length}
+        handleSelectFile={handleSelectFile}
+        handleOpenFolder={handleOpenFolder}
+        isInPipeline={isInPipeline}
+        isLoading={isLoading}
+      />
+    </div>
+  )
+}
+
+export default OnlineDrive

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/online-drive/utils.ts

@@ -0,0 +1,54 @@
+import { type OnlineDriveFile, OnlineDriveFileType } from '@/models/pipeline'
+import type { OnlineDriveData } from '@/types/pipeline'
+
+export const isFile = (type: 'file' | 'folder'): boolean => {
+  return type === 'file'
+}
+
+export const isBucketListInitiation = (data: OnlineDriveData[], prefix: string[], bucket: string): boolean => {
+  if (bucket || prefix.length > 0) return false
+  const hasBucket = data.every(item => !!item.bucket)
+  return hasBucket && (data.length > 1 || (data.length === 1 && !!data[0].bucket && data[0].files.length === 0))
+}
+
+export const convertOnlineDriveData = (data: OnlineDriveData[], prefix: string[], bucket: string): {
+  fileList: OnlineDriveFile[],
+  isTruncated: boolean,
+  nextPageParameters: Record<string, any>
+  hasBucket: boolean
+} => {
+  const fileList: OnlineDriveFile[] = []
+  let isTruncated = false
+  let nextPageParameters: Record<string, any> = {}
+  let hasBucket = false
+
+  if (data.length === 0)
+    return { fileList, isTruncated, nextPageParameters, hasBucket }
+
+  if (isBucketListInitiation(data, prefix, bucket)) {
+    data.forEach((item) => {
+      fileList.push({
+        id: item.bucket,
+        name: item.bucket,
+        type: OnlineDriveFileType.bucket,
+      })
+    })
+    hasBucket = true
+  }
+  else {
+    data[0].files.forEach((file) => {
+      const { id, name, size, type } = file
+      const isFileType = isFile(type)
+      fileList.push({
+        id,
+        name,
+        size: isFileType ? size : undefined,
+        type: isFileType ? OnlineDriveFileType.file : OnlineDriveFileType.folder,
+      })
+    })
+    isTruncated = data[0].is_truncated ?? false
+    nextPageParameters = data[0].next_page_parameters ?? {}
+    hasBucket = !!data[0].bucket
+  }
+  return { fileList, isTruncated, nextPageParameters, hasBucket }
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/store/index.ts

@@ -0,0 +1,45 @@
+import { useContext } from 'react'
+import { createStore, useStore } from 'zustand'
+import { DataSourceContext } from './provider'
+import type { CommonShape } from './slices/common'
+import { createCommonSlice } from './slices/common'
+import type { LocalFileSliceShape } from './slices/local-file'
+import { createLocalFileSlice } from './slices/local-file'
+import type { OnlineDocumentSliceShape } from './slices/online-document'
+import { createOnlineDocumentSlice } from './slices/online-document'
+import type { WebsiteCrawlSliceShape } from './slices/website-crawl'
+import { createWebsiteCrawlSlice } from './slices/website-crawl'
+import type { OnlineDriveSliceShape } from './slices/online-drive'
+import { createOnlineDriveSlice } from './slices/online-drive'
+
+export type DataSourceShape = CommonShape
+  & LocalFileSliceShape
+  & OnlineDocumentSliceShape
+  & WebsiteCrawlSliceShape
+  & OnlineDriveSliceShape
+
+export const createDataSourceStore = () => {
+  return createStore<DataSourceShape>((...args) => ({
+    ...createCommonSlice(...args),
+    ...createLocalFileSlice(...args),
+    ...createOnlineDocumentSlice(...args),
+    ...createWebsiteCrawlSlice(...args),
+    ...createOnlineDriveSlice(...args),
+  }))
+}
+
+export const useDataSourceStoreWithSelector = <T>(selector: (state: DataSourceShape) => T): T => {
+  const store = useContext(DataSourceContext)
+  if (!store)
+    throw new Error('Missing DataSourceContext.Provider in the tree')
+
+  return useStore(store, selector)
+}
+
+export const useDataSourceStore = () => {
+  const store = useContext(DataSourceContext)
+  if (!store)
+    throw new Error('Missing DataSourceContext.Provider in the tree')
+
+  return store
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/store/provider.tsx

@@ -0,0 +1,29 @@
+import { createContext, useRef } from 'react'
+import { createDataSourceStore } from './'
+
+type DataSourceStoreApi = ReturnType<typeof createDataSourceStore>
+
+type DataSourceContextType = DataSourceStoreApi | null
+
+export const DataSourceContext = createContext<DataSourceContextType>(null)
+
+type DataSourceProviderProps = {
+  children: React.ReactNode
+}
+
+const DataSourceProvider = ({
+  children,
+}: DataSourceProviderProps) => {
+  const storeRef = useRef<DataSourceStoreApi>(null)
+
+  if (!storeRef.current)
+    storeRef.current = createDataSourceStore()
+
+  return (
+    <DataSourceContext.Provider value={storeRef.current!}>
+      {children}
+    </DataSourceContext.Provider>
+  )
+}
+
+export default DataSourceProvider

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/store/slices/common.ts

@@ -0,0 +1,19 @@
+import type { StateCreator } from 'zustand'
+
+export type CommonShape = {
+  currentNodeIdRef: React.RefObject<string>
+  currentCredentialId: string
+  setCurrentCredentialId: (credentialId: string) => void
+  currentCredentialIdRef: React.RefObject<string>
+}
+
+export const createCommonSlice: StateCreator<CommonShape> = (set) => {
+  return ({
+    currentNodeIdRef: { current: '' },
+    currentCredentialId: '',
+    setCurrentCredentialId: (credentialId: string) => {
+      set({ currentCredentialId: credentialId })
+    },
+    currentCredentialIdRef: { current: '' },
+  })
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/store/slices/local-file.ts

@@ -0,0 +1,28 @@
+import type { StateCreator } from 'zustand'
+import type { DocumentItem, CustomFile as File, FileItem } from '@/models/datasets'
+
+export type LocalFileSliceShape = {
+  localFileList: FileItem[]
+  setLocalFileList: (fileList: FileItem[]) => void
+  currentLocalFile: File | undefined
+  setCurrentLocalFile: (file: File | undefined) => void
+  previewLocalFileRef: React.RefObject<DocumentItem | undefined>
+}
+
+export const createLocalFileSlice: StateCreator<LocalFileSliceShape> = (set, get) => {
+  return ({
+    localFileList: [],
+    setLocalFileList: (fileList: FileItem[]) => {
+      set(() => ({
+        localFileList: fileList,
+      }))
+      const { previewLocalFileRef } = get()
+      previewLocalFileRef.current = fileList[0]?.file as DocumentItem
+    },
+    currentLocalFile: undefined,
+    setCurrentLocalFile: (file: File | undefined) => set(() => ({
+      currentLocalFile: file,
+    })),
+    previewLocalFileRef: { current: undefined },
+  })
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/store/slices/online-document.ts

@@ -0,0 +1,46 @@
+import type { StateCreator } from 'zustand'
+import type { DataSourceNotionWorkspace, NotionPage } from '@/models/common'
+
+export type OnlineDocumentSliceShape = {
+  documentsData: DataSourceNotionWorkspace[]
+  setDocumentsData: (documentData: DataSourceNotionWorkspace[]) => void
+  searchValue: string
+  setSearchValue: (searchValue: string) => void
+  onlineDocuments: NotionPage[]
+  setOnlineDocuments: (documents: NotionPage[]) => void
+  currentDocument: NotionPage | undefined
+  setCurrentDocument: (document: NotionPage | undefined) => void
+  selectedPagesId: Set<string>
+  setSelectedPagesId: (selectedPagesId: Set<string>) => void
+  previewOnlineDocumentRef: React.RefObject<NotionPage | undefined>
+}
+
+export const createOnlineDocumentSlice: StateCreator<OnlineDocumentSliceShape> = (set, get) => {
+  return ({
+    documentsData: [],
+    setDocumentsData: (documentsData: DataSourceNotionWorkspace[]) => set(() => ({
+      documentsData,
+    })),
+    searchValue: '',
+    setSearchValue: (searchValue: string) => set(() => ({
+      searchValue,
+    })),
+    onlineDocuments: [],
+    setOnlineDocuments: (documents: NotionPage[]) => {
+      set(() => ({
+        onlineDocuments: documents,
+      }))
+      const { previewOnlineDocumentRef } = get()
+      previewOnlineDocumentRef.current = documents[0]
+    },
+    currentDocument: undefined,
+    setCurrentDocument: (document: NotionPage | undefined) => set(() => ({
+      currentDocument: document,
+    })),
+    selectedPagesId: new Set(),
+    setSelectedPagesId: (selectedPagesId: Set<string>) => set(() => ({
+      selectedPagesId,
+    })),
+    previewOnlineDocumentRef: { current: undefined },
+  })
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/store/slices/online-drive.ts

@@ -0,0 +1,69 @@
+import type { StateCreator } from 'zustand'
+import type { OnlineDriveFile } from '@/models/pipeline'
+
+export type OnlineDriveSliceShape = {
+  breadcrumbs: string[]
+  setBreadcrumbs: (breadcrumbs: string[]) => void
+  prefix: string[]
+  setPrefix: (prefix: string[]) => void
+  keywords: string
+  setKeywords: (keywords: string) => void
+  selectedFileIds: string[]
+  setSelectedFileIds: (selectedFileIds: string[]) => void
+  onlineDriveFileList: OnlineDriveFile[]
+  setOnlineDriveFileList: (onlineDriveFileList: OnlineDriveFile[]) => void
+  bucket: string
+  setBucket: (bucket: string) => void
+  nextPageParameters: Record<string, any>
+  currentNextPageParametersRef: React.RefObject<Record<string, any>>
+  setNextPageParameters: (nextPageParameters: Record<string, any>) => void
+  isTruncated: React.RefObject<boolean>
+  previewOnlineDriveFileRef: React.RefObject<OnlineDriveFile | undefined>
+  hasBucket: boolean
+  setHasBucket: (hasBucket: boolean) => void
+}
+
+export const createOnlineDriveSlice: StateCreator<OnlineDriveSliceShape> = (set, get) => {
+  return ({
+    breadcrumbs: [],
+    setBreadcrumbs: (breadcrumbs: string[]) => set(() => ({
+      breadcrumbs,
+    })),
+    prefix: [],
+    setPrefix: (prefix: string[]) => set(() => ({
+      prefix,
+    })),
+    keywords: '',
+    setKeywords: (keywords: string) => set(() => ({
+      keywords,
+    })),
+    selectedFileIds: [],
+    setSelectedFileIds: (selectedFileIds: string[]) => {
+      set(() => ({
+        selectedFileIds,
+      }))
+      const id = selectedFileIds[0]
+      const { onlineDriveFileList, previewOnlineDriveFileRef } = get()
+      previewOnlineDriveFileRef.current = onlineDriveFileList.find(file => file.id === id)
+    },
+    onlineDriveFileList: [],
+    setOnlineDriveFileList: (onlineDriveFileList: OnlineDriveFile[]) => set(() => ({
+      onlineDriveFileList,
+    })),
+    bucket: '',
+    setBucket: (bucket: string) => set(() => ({
+      bucket,
+    })),
+    nextPageParameters: {},
+    currentNextPageParametersRef: { current: {} },
+    setNextPageParameters: (nextPageParameters: Record<string, any>) => set(() => ({
+      nextPageParameters,
+    })),
+    isTruncated: { current: false },
+    previewOnlineDriveFileRef: { current: undefined },
+    hasBucket: false,
+    setHasBucket: (hasBucket: boolean) => set(() => ({
+      hasBucket,
+    })),
+  })
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/store/slices/website-crawl.ts

@@ -0,0 +1,47 @@
+import type { StateCreator } from 'zustand'
+import type { CrawlResult, CrawlResultItem } from '@/models/datasets'
+import { CrawlStep } from '@/models/datasets'
+
+export type WebsiteCrawlSliceShape = {
+  websitePages: CrawlResultItem[]
+  setWebsitePages: (pages: CrawlResultItem[]) => void
+  currentWebsite: CrawlResultItem | undefined
+  setCurrentWebsite: (website: CrawlResultItem | undefined) => void
+  crawlResult: CrawlResult | undefined
+  setCrawlResult: (result: CrawlResult | undefined) => void
+  step: CrawlStep
+  setStep: (step: CrawlStep) => void
+  previewIndex: number
+  setPreviewIndex: (index: number) => void
+  previewWebsitePageRef: React.RefObject<CrawlResultItem | undefined>
+}
+
+export const createWebsiteCrawlSlice: StateCreator<WebsiteCrawlSliceShape> = (set, get) => {
+  return ({
+    websitePages: [],
+    setWebsitePages: (pages: CrawlResultItem[]) => {
+      set(() => ({
+        websitePages: pages,
+      }))
+      const { previewWebsitePageRef } = get()
+      previewWebsitePageRef.current = pages[0]
+    },
+    currentWebsite: undefined,
+    setCurrentWebsite: (website: CrawlResultItem | undefined) => set(() => ({
+      currentWebsite: website,
+    })),
+    crawlResult: undefined,
+    setCrawlResult: (result: CrawlResult | undefined) => set(() => ({
+      crawlResult: result,
+    })),
+    step: CrawlStep.init,
+    setStep: (step: CrawlStep) => set(() => ({
+      step,
+    })),
+    previewIndex: -1,
+    setPreviewIndex: (index: number) => set(() => ({
+      previewIndex: index,
+    })),
+    previewWebsitePageRef: { current: undefined },
+  })
+}

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/website-crawl/base/checkbox-with-label.tsx

@@ -0,0 +1,39 @@
+'use client'
+import React from 'react'
+import cn from '@/utils/classnames'
+import Checkbox from '@/app/components/base/checkbox'
+import Tooltip from '@/app/components/base/tooltip'
+
+type CheckboxWithLabelProps = {
+  className?: string
+  isChecked: boolean
+  onChange: (isChecked: boolean) => void
+  label: string
+  labelClassName?: string
+  tooltip?: string
+}
+
+const CheckboxWithLabel = ({
+  className = '',
+  isChecked,
+  onChange,
+  label,
+  labelClassName,
+  tooltip,
+}: CheckboxWithLabelProps) => {
+  return (
+    <label className={cn('flex items-center space-x-2', className)}>
+      <Checkbox checked={isChecked} onCheck={() => onChange(!isChecked)} />
+      <div className={cn('system-sm-medium text-text-secondary', labelClassName)}>{label}</div>
+      {tooltip && (
+        <Tooltip
+          popupContent={
+            <div className='w-[200px]'>{tooltip}</div>
+          }
+          triggerClassName='ml-0.5 w-4 h-4'
+        />
+      )}
+    </label>
+  )
+}
+export default React.memo(CheckboxWithLabel)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/website-crawl/base/crawled-result-item.tsx

@@ -0,0 +1,80 @@
+'use client'
+import React, { useCallback } from 'react'
+import cn from '@/utils/classnames'
+import type { CrawlResultItem as CrawlResultItemType } from '@/models/datasets'
+import Checkbox from '@/app/components/base/checkbox'
+import Button from '@/app/components/base/button'
+import { useTranslation } from 'react-i18next'
+import Radio from '@/app/components/base/radio/ui'
+
+type CrawledResultItemProps = {
+  payload: CrawlResultItemType
+  isChecked: boolean
+  onCheckChange: (checked: boolean) => void
+  isPreview: boolean
+  showPreview: boolean
+  onPreview: () => void
+  isMultipleChoice?: boolean
+}
+
+const CrawledResultItem = ({
+  payload,
+  isChecked,
+  onCheckChange,
+  isPreview,
+  onPreview,
+  showPreview,
+  isMultipleChoice = true,
+}: CrawledResultItemProps) => {
+  const { t } = useTranslation()
+
+  const handleCheckChange = useCallback(() => {
+    onCheckChange(!isChecked)
+  }, [isChecked, onCheckChange])
+
+  return (
+    <div className={cn(
+      'relative flex cursor-pointer gap-x-2 rounded-lg p-2',
+      isPreview ? 'bg-state-base-active' : 'group hover:bg-state-base-hover',
+    )}>
+      {
+        isMultipleChoice ? (
+          <Checkbox
+            className='shrink-0'
+            checked={isChecked}
+            onCheck={handleCheckChange}
+          />
+        ) : (
+          <Radio
+            isChecked={isChecked}
+            onCheck={handleCheckChange}
+          />
+        )
+      }
+      <div className='flex min-w-0 grow flex-col gap-y-0.5'>
+        <div
+          className='system-sm-medium truncate text-text-secondary'
+          title={payload.title}
+        >
+          {payload.title}
+        </div>
+        <div
+          className='system-xs-regular truncate text-text-tertiary'
+          title={payload.source_url}
+        >
+          {payload.source_url}
+        </div>
+      </div>
+      {showPreview && (
+        <Button
+          size='small'
+          onClick={onPreview}
+          className='system-xs-medium-uppercase right-2 top-2 hidden px-1.5 group-hover:absolute group-hover:block'
+        >
+          {t('datasetCreation.stepOne.website.preview')}
+        </Button>
+      )}
+    </div>
+  )
+}
+export default React.memo(CrawledResultItem)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/website-crawl/base/crawled-result.tsx

@@ -0,0 +1,98 @@
+'use client'
+import React, { useCallback } from 'react'
+import { useTranslation } from 'react-i18next'
+import cn from '@/utils/classnames'
+import type { CrawlResultItem } from '@/models/datasets'
+import CheckboxWithLabel from './checkbox-with-label'
+import CrawledResultItem from './crawled-result-item'
+
+const I18N_PREFIX = 'datasetCreation.stepOne.website'
+
+type CrawledResultProps = {
+  className?: string
+  previewIndex?: number
+  list: CrawlResultItem[]
+  checkedList: CrawlResultItem[]
+  onSelectedChange: (selected: CrawlResultItem[]) => void
+  onPreview?: (payload: CrawlResultItem, index: number) => void
+  showPreview?: boolean
+  usedTime: number
+  isMultipleChoice?: boolean
+}
+
+const CrawledResult = ({
+  className = '',
+  previewIndex,
+  list,
+  checkedList,
+  onSelectedChange,
+  usedTime,
+  onPreview,
+  showPreview = false,
+  isMultipleChoice = true,
+}: CrawledResultProps) => {
+  const { t } = useTranslation()
+
+  const isCheckAll = checkedList.length === list.length
+
+  const handleCheckedAll = useCallback(() => {
+    if (!isCheckAll)
+      onSelectedChange(list)
+
+    else
+      onSelectedChange([])
+  }, [isCheckAll, list, onSelectedChange])
+
+  const handleItemCheckChange = useCallback((item: CrawlResultItem) => {
+    return (checked: boolean) => {
+      if (checked) {
+        if (isMultipleChoice)
+          onSelectedChange([...checkedList, item])
+        else
+          onSelectedChange([item])
+      }
+      else { onSelectedChange(checkedList.filter(checkedItem => checkedItem.source_url !== item.source_url)) }
+    }
+  }, [checkedList, onSelectedChange, isMultipleChoice])
+
+  const handlePreview = useCallback((index: number) => {
+    if (!onPreview) return
+    onPreview(list[index], index)
+  }, [list, onPreview])
+
+  return (
+    <div className={cn('flex flex-col gap-y-2', className)}>
+      <div className='system-sm-medium pt-2 text-text-primary'>
+        {t(`${I18N_PREFIX}.scrapTimeInfo`, {
+          total: list.length,
+          time: usedTime.toFixed(1),
+        })}
+      </div>
+      <div className='overflow-hidden rounded-xl border border-components-panel-border bg-components-panel-bg'>
+        {isMultipleChoice && (
+          <div className='flex items-center px-4 py-2'>
+            <CheckboxWithLabel
+              isChecked={isCheckAll}
+              onChange={handleCheckedAll} label={isCheckAll ? t(`${I18N_PREFIX}.resetAll`) : t(`${I18N_PREFIX}.selectAll`)}
+            />
+          </div>
+        )}
+        <div className='flex flex-col gap-y-px border-t border-divider-subtle bg-background-default-subtle p-2'>
+          {list.map((item, index) => (
+            <CrawledResultItem
+              key={item.source_url}
+              payload={item}
+              isChecked={checkedList.some(checkedItem => checkedItem.source_url === item.source_url)}
+              onCheckChange={handleItemCheckChange(item)}
+              isPreview={index === previewIndex}
+              onPreview={handlePreview.bind(null, index)}
+              showPreview={showPreview}
+              isMultipleChoice={isMultipleChoice}
+            />
+          ))}
+        </div>
+      </div>
+    </div>
+  )
+}
+export default React.memo(CrawledResult)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/website-crawl/base/crawling.tsx

@@ -0,0 +1,89 @@
+'use client'
+import React from 'react'
+import { useTranslation } from 'react-i18next'
+import cn from '@/utils/classnames'
+
+type CrawlingProps = {
+  className?: string
+  crawledNum: number
+  totalNum: number
+}
+
+type BlockProps = {
+  className?: string
+}
+
+type ItemProps = {
+  firstLineWidth: string
+  secondLineWidth: string
+}
+
+const Block = React.memo(({
+  className,
+}: BlockProps) => {
+  return <div className={cn('bg-text-quaternary opacity-20', className)} />
+})
+
+const Item = React.memo(({
+  firstLineWidth,
+  secondLineWidth,
+}: ItemProps) => {
+  return (
+    <div className='flex gap-x-2 px-2 py-[5px]'>
+      <div className='py-0.5'>
+        <Block className='size-4 rounded-[4px]' />
+      </div>
+      <div className='flex grow flex-col'>
+        <div className='flex h-5 w-full items-center'>
+          <Block className={cn('h-2.5 rounded-sm', firstLineWidth)} />
+        </div>
+        <div className='flex h-[18px] w-full items-center'>
+          <Block className={cn('h-1.5 rounded-sm', secondLineWidth)} />
+        </div>
+      </div>
+    </div>
+  )
+})
+
+const Crawling = ({
+  className = '',
+  crawledNum,
+  totalNum,
+}: CrawlingProps) => {
+  const { t } = useTranslation()
+
+  const itemsConfig = [{
+    firstLineWidth: 'w-[35%]',
+    secondLineWidth: 'w-[50%]',
+  }, {
+    firstLineWidth: 'w-[40%]',
+    secondLineWidth: 'w-[45%]',
+  }, {
+    firstLineWidth: 'w-[30%]',
+    secondLineWidth: 'w-[36%]',
+  }]
+
+  return (
+    <div className={cn('mt-2 flex flex-col gap-y-2 pt-2', className)}>
+      <div className='system-sm-medium text-text-primary'>
+        {t('datasetCreation.stepOne.website.totalPageScraped')} {crawledNum}/{totalNum}
+      </div>
+      <div className='overflow-hidden rounded-xl border border-components-panel-border bg-components-panel-bg'>
+        <div className='flex items-center gap-x-2 px-4 py-2'>
+          <Block className='size-4 rounded-[4px]' />
+          <Block className='h-2.5 w-14 rounded-sm' />
+        </div>
+        <div className='flex flex-col gap-px border-t border-divider-subtle bg-background-default-subtle p-2'>
+          {itemsConfig.map((item, index) => (
+            <Item
+              key={index}
+              firstLineWidth={item.firstLineWidth}
+              secondLineWidth={item.secondLineWidth}
+            />
+          ))}
+        </div>
+      </div>
+    </div>
+  )
+}
+export default React.memo(Crawling)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/website-crawl/base/error-message.tsx

@@ -0,0 +1,34 @@
+import React from 'react'
+import cn from '@/utils/classnames'
+import { RiErrorWarningFill } from '@remixicon/react'
+
+type ErrorMessageProps = {
+  className?: string
+  title: string
+  errorMsg?: string
+}
+
+const ErrorMessage = ({
+  className,
+  title,
+  errorMsg,
+}: ErrorMessageProps) => {
+  return (
+    // eslint-disable-next-line tailwindcss/migration-from-tailwind-2
+    <div className={cn(
+      'flex gap-x-0.5 rounded-xl border-[0.5px] border-components-panel-border bg-opacity-40 bg-toast-error-bg p-2 shadow-xs shadow-shadow-shadow-3',
+      className,
+    )}>
+      <div className='flex size-6 items-center justify-center'>
+        <RiErrorWarningFill className='h-4 w-4 text-text-destructive' />
+      </div>
+      <div className='flex flex-col gap-y-0.5 py-1'>
+        <div className='system-xs-medium text-text-primary'>{title}</div>
+        {errorMsg && (
+          <div className='system-xs-regular text-text-secondary'>{errorMsg}</div>
+        )}
+      </div>
+    </div>
+  )
+}
+export default React.memo(ErrorMessage)

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/website-crawl/base/options/index.tsx

@@ -0,0 +1,123 @@
+import Button from '@/app/components/base/button'
+import { useAppForm } from '@/app/components/base/form'
+import BaseField from '@/app/components/base/form/form-scenarios/base/field'
+import { ArrowDownRoundFill } from '@/app/components/base/icons/src/vender/solid/general'
+import cn from '@/utils/classnames'
+import { RiPlayLargeLine } from '@remixicon/react'
+import { useBoolean } from 'ahooks'
+import { useEffect, useMemo } from 'react'
+import { useTranslation } from 'react-i18next'
+import Toast from '@/app/components/base/toast'
+import type { RAGPipelineVariables } from '@/models/pipeline'
+import { useConfigurations, useInitialData } from '@/app/components/rag-pipeline/hooks/use-input-fields'
+import { generateZodSchema } from '@/app/components/base/form/form-scenarios/base/utils'
+import { CrawlStep } from '@/models/datasets'
+
+const I18N_PREFIX = 'datasetCreation.stepOne.website'
+
+type OptionsProps = {
+  variables: RAGPipelineVariables
+  step: CrawlStep
+  runDisabled?: boolean
+  onSubmit: (data: Record<string, any>) => void
+}
+
+const Options = ({
+  variables,
+  step,
+  runDisabled,
+  onSubmit,
+}: OptionsProps) => {
+  const { t } = useTranslation()
+  const initialData = useInitialData(variables)
+  const configurations = useConfigurations(variables)
+  const schema = useMemo(() => {
+    return generateZodSchema(configurations)
+  }, [configurations])
+
+  const form = useAppForm({
+    defaultValues: initialData,
+    validators: {
+      onSubmit: ({ value }) => {
+        const result = schema.safeParse(value)
+        if (!result.success) {
+          const issues = result.error.issues
+          const firstIssue = issues[0]
+          const errorMessage = `"${firstIssue.path.join('.')}" ${firstIssue.message}`
+          Toast.notify({
+            type: 'error',
+            message: errorMessage,
+          })
+          return errorMessage
+        }
+        return undefined
+      },
+    },
+    onSubmit: ({ value }) => {
+      onSubmit(value)
+    },
+  })
+
+  const [fold, {
+    toggle: foldToggle,
+    setTrue: foldHide,
+    setFalse: foldShow,
+  }] = useBoolean(false)
+
+  useEffect(() => {
+    // When the step change
+    if (step !== CrawlStep.init)
+      foldHide()
+    else
+      foldShow()
+  }, [step])
+
+  const isRunning = useMemo(() => step === CrawlStep.running, [step])
+
+  return (
+    <form
+      className='w-full'
+      onSubmit={(e) => {
+        e.preventDefault()
+        e.stopPropagation()
+        form.handleSubmit()
+      }}
+    >
+      <div className='flex items-center gap-x-1 px-4 py-2'>
+        <div
+          className='flex grow cursor-pointer select-none items-center gap-x-0.5'
+          onClick={foldToggle}
+        >
+          <span className='system-sm-semibold-uppercase text-text-secondary'>
+            {t(`${I18N_PREFIX}.options`)}
+          </span>
+          <ArrowDownRoundFill className={cn('h-4 w-4 shrink-0 text-text-quaternary', fold && '-rotate-90')} />
+        </div>
+        <Button
+          variant='primary'
+          onClick={form.handleSubmit}
+          disabled={runDisabled || isRunning}
+          loading={isRunning}
+          className='shrink-0 gap-x-0.5'
+          spinnerClassName='!ml-0'
+        >
+          <RiPlayLargeLine className='size-4' />
+          <span className='px-0.5'>{!isRunning ? t(`${I18N_PREFIX}.run`) : t(`${I18N_PREFIX}.running`)}</span>
+        </Button>
+      </div>
+      {!fold && (
+        <div className='flex flex-col gap-3 border-t border-divider-subtle px-4 py-3'>
+          {configurations.map((config, index) => {
+            const FieldComponent = BaseField({
+              initialData,
+              config,
+            })
+            return <FieldComponent key={index} form={form} />
+          })}
+        </div>
+      )}
+    </form>
+  )
+}
+
+export default Options

dify/web/app/components/datasets/documents/create-from-pipeline/data-source/website-crawl/index.tsx

@@ -0,0 +1,206 @@
+'use client'
+import React, { useCallback, useRef, useState } from 'react'
+import { useTranslation } from 'react-i18next'
+import type { CrawlResultItem } from '@/models/datasets'
+import { CrawlStep } from '@/models/datasets'
+import Header from '../base/header'
+import Options from './base/options'
+import Crawling from './base/crawling'
+import ErrorMessage from './base/error-message'
+import CrawledResult from './base/crawled-result'
+import {
+  useDraftPipelinePreProcessingParams,
+  usePublishedPipelinePreProcessingParams,
+} from '@/service/use-pipeline'
+import { useDatasetDetailContextWithSelector } from '@/context/dataset-detail'
+import { DatasourceType } from '@/models/pipeline'
+import { ssePost } from '@/service/base'
+import type {
+  DataSourceNodeCompletedResponse,
+  DataSourceNodeErrorResponse,
+  DataSourceNodeProcessingResponse,
+} from '@/types/pipeline'
+import type { DataSourceNodeType } from '@/app/components/workflow/nodes/data-source/types'
+import { useDataSourceStore, useDataSourceStoreWithSelector } from '../store'
+import { useShallow } from 'zustand/react/shallow'
+import { useModalContextSelector } from '@/context/modal-context'
+import { useGetDataSourceAuth } from '@/service/use-datasource'
+import { useDocLink } from '@/context/i18n'
+import { ACCOUNT_SETTING_TAB } from '@/app/components/header/account-setting/constants'
+
+const I18N_PREFIX = 'datasetCreation.stepOne.website'
+
+export type WebsiteCrawlProps = {
+  nodeId: string
+  nodeData: DataSourceNodeType
+  isInPipeline?: boolean
+  onCredentialChange: (credentialId: string) => void
+}
+
+const WebsiteCrawl = ({
+  nodeId,
+  nodeData,
+  isInPipeline = false,
+  onCredentialChange,
+}: WebsiteCrawlProps) => {
+  const { t } = useTranslation()
+  const docLink = useDocLink()
+  const [totalNum, setTotalNum] = useState(0)
+  const [crawledNum, setCrawledNum] = useState(0)
+  const [crawlErrorMessage, setCrawlErrorMessage] = useState('')
+  const pipelineId = useDatasetDetailContextWithSelector(s => s.dataset?.pipeline_id)
+  const setShowAccountSettingModal = useModalContextSelector(s => s.setShowAccountSettingModal)
+  const {
+    crawlResult,
+    step,
+    checkedCrawlResult,
+    previewIndex,
+    currentCredentialId,
+  } = useDataSourceStoreWithSelector(useShallow(state => ({
+    crawlResult: state.crawlResult,
+    step: state.step,
+    checkedCrawlResult: state.websitePages,
+    previewIndex: state.previewIndex,
+    currentCredentialId: state.currentCredentialId,
+  })))
+
+  const { data: dataSourceAuth } = useGetDataSourceAuth({
+    pluginId: nodeData.plugin_id,
+    provider: nodeData.provider_name,
+  })
+
+  const dataSourceStore = useDataSourceStore()
+
+  const usePreProcessingParams = useRef(!isInPipeline ? usePublishedPipelinePreProcessingParams : useDraftPipelinePreProcessingParams)
+  const { data: paramsConfig, isFetching: isFetchingParams } = usePreProcessingParams.current({
+    pipeline_id: pipelineId!,
+    node_id: nodeId,
+  }, !!pipelineId && !!nodeId)
+
+  const isInit = step === CrawlStep.init
+  const isCrawlFinished = step === CrawlStep.finished
+  const isRunning = step === CrawlStep.running
+  const showError = isCrawlFinished && crawlErrorMessage
+  const datasourceNodeRunURL = !isInPipeline
+    ? `/rag/pipelines/${pipelineId}/workflows/published/datasource/nodes/${nodeId}/run`
+    : `/rag/pipelines/${pipelineId}/workflows/draft/datasource/nodes/${nodeId}/run`
+
+  const handleCheckedCrawlResultChange = useCallback((checkedCrawlResult: CrawlResultItem[]) => {
+    const { setWebsitePages } = dataSourceStore.getState()
+    setWebsitePages(checkedCrawlResult)
+  }, [dataSourceStore])
+
+  const handlePreview = useCallback((website: CrawlResultItem, index: number) => {
+    const { setCurrentWebsite, setPreviewIndex } = dataSourceStore.getState()
+    setCurrentWebsite(website)
+    setPreviewIndex(index)
+  }, [dataSourceStore])
+
+  const handleRun = useCallback(async (value: Record<string, any>) => {
+    const { setStep, setCrawlResult, currentCredentialId } = dataSourceStore.getState()
+
+    setStep(CrawlStep.running)
+    ssePost(
+      datasourceNodeRunURL,
+      {
+        body: {
+          inputs: value,
+          datasource_type: DatasourceType.websiteCrawl,
+          credential_id: currentCredentialId,
+          response_mode: 'streaming',
+        },
+      },
+      {
+        onDataSourceNodeProcessing: (data: DataSourceNodeProcessingResponse) => {
+          setTotalNum(data.total ?? 0)
+          setCrawledNum(data.completed ?? 0)
+        },
+        onDataSourceNodeCompleted: (data: DataSourceNodeCompletedResponse) => {
+          const { data: crawlData, time_consuming } = data
+          const crawlResultData = {
+            data: crawlData as CrawlResultItem[],
+            time_consuming: time_consuming ?? 0,
+          }
+          setCrawlResult(crawlResultData)
+          handleCheckedCrawlResultChange(isInPipeline ? [crawlData[0]] : crawlData) // default select the crawl result
+          setCrawlErrorMessage('')
+          setStep(CrawlStep.finished)
+        },
+        onDataSourceNodeError: (error: DataSourceNodeErrorResponse) => {
+          setCrawlErrorMessage(error.error || t(`${I18N_PREFIX}.unknownError`))
+          setStep(CrawlStep.finished)
+        },
+      },
+    )
+  }, [dataSourceStore, datasourceNodeRunURL, handleCheckedCrawlResultChange, isInPipeline, t])
+
+  const handleSubmit = useCallback((value: Record<string, any>) => {
+    handleRun(value)
+  }, [handleRun])
+
+  const handleSetting = useCallback(() => {
+    setShowAccountSettingModal({
+      payload: ACCOUNT_SETTING_TAB.DATA_SOURCE,
+    })
+  }, [setShowAccountSettingModal])
+
+  const handleCredentialChange = useCallback((credentialId: string) => {
+    setCrawledNum(0)
+    setTotalNum(0)
+    setCrawlErrorMessage('')
+    onCredentialChange(credentialId)
+  }, [dataSourceStore, onCredentialChange])
+
+  return (
+    <div className='flex flex-col'>
+      <Header
+        docTitle='Docs'
+        docLink={docLink('/guides/knowledge-base/knowledge-pipeline/authorize-data-source')}
+        onClickConfiguration={handleSetting}
+        pluginName={nodeData.datasource_label}
+        currentCredentialId={currentCredentialId}
+        onCredentialChange={handleCredentialChange}
+        credentials={dataSourceAuth?.result || []}
+      />
+      <div className='mt-2 rounded-xl border border-components-panel-border bg-background-default-subtle'>
+        <Options
+          variables={paramsConfig?.variables || []}
+          step={step}
+          runDisabled={!currentCredentialId || isFetchingParams}
+          onSubmit={handleSubmit}
+        />
+      </div>
+      {!isInit && (
+        <div className='relative flex flex-col'>
+          {isRunning && (
+            <Crawling
+              crawledNum={crawledNum}
+              totalNum={totalNum}
+            />
+          )}
+          {showError && (
+            <ErrorMessage
+              className='mt-2'
+              title={t(`${I18N_PREFIX}.exceptionErrorTitle`)}
+              errorMsg={crawlErrorMessage}
+            />
+          )}
+          {isCrawlFinished && !showError && (
+            <CrawledResult
+              className='mt-2'
+              list={crawlResult?.data || []}
+              checkedList={checkedCrawlResult}
+              onSelectedChange={handleCheckedCrawlResultChange}
+              usedTime={Number.parseFloat(crawlResult?.time_consuming as string) || 0}
+              previewIndex={previewIndex}
+              onPreview={handlePreview}
+              showPreview={!isInPipeline}
+              isMultipleChoice={!isInPipeline} // only support single choice in test run
+            />
+          )}
+        </div>
+      )}
+    </div>
+  )
+}
+export default React.memo(WebsiteCrawl)

dify/web/app/components/datasets/documents/create-from-pipeline/hooks.ts

@@ -0,0 +1,223 @@
+import { useTranslation } from 'react-i18next'
+import { AddDocumentsStep } from './types'
+import type { DataSourceOption } from '@/app/components/rag-pipeline/components/panel/test-run/types'
+import { useCallback, useMemo, useState } from 'react'
+import { BlockEnum, type Node } from '@/app/components/workflow/types'
+import type { DataSourceNodeType } from '@/app/components/workflow/nodes/data-source/types'
+import { useDataSourceStore, useDataSourceStoreWithSelector } from './data-source/store'
+import type { DataSourceNotionPageMap, DataSourceNotionWorkspace } from '@/models/common'
+import { useShallow } from 'zustand/react/shallow'
+import { CrawlStep } from '@/models/datasets'
+
+export const useAddDocumentsSteps = () => {
+  const { t } = useTranslation()
+  const [currentStep, setCurrentStep] = useState(1)
+
+  const handleNextStep = useCallback(() => {
+    setCurrentStep(preStep => preStep + 1)
+  }, [])
+
+  const handleBackStep = useCallback(() => {
+    setCurrentStep(preStep => preStep - 1)
+  }, [])
+
+  const steps = [
+    {
+      label: t('datasetPipeline.addDocuments.steps.chooseDatasource'),
+      value: AddDocumentsStep.dataSource,
+    },
+    {
+      label: t('datasetPipeline.addDocuments.steps.processDocuments'),
+      value: AddDocumentsStep.processDocuments,
+    },
+    {
+      label: t('datasetPipeline.addDocuments.steps.processingDocuments'),
+      value: AddDocumentsStep.processingDocuments,
+    },
+  ]
+
+  return {
+    steps,
+    currentStep,
+    handleNextStep,
+    handleBackStep,
+  }
+}
+
+export const useDatasourceOptions = (pipelineNodes: Node<DataSourceNodeType>[]) => {
+  const datasourceNodes = pipelineNodes.filter(node => node.data.type === BlockEnum.DataSource)
+
+  const options = useMemo(() => {
+    const options: DataSourceOption[] = []
+    datasourceNodes.forEach((node) => {
+      const label = node.data.title
+      options.push({
+        label,
+        value: node.id,
+        data: node.data,
+      })
+    })
+    return options
+  }, [datasourceNodes])
+
+  return options
+}
+
+export const useLocalFile = () => {
+  const {
+    localFileList,
+    currentLocalFile,
+  } = useDataSourceStoreWithSelector(useShallow(state => ({
+    localFileList: state.localFileList,
+    currentLocalFile: state.currentLocalFile,
+  })))
+  const dataSourceStore = useDataSourceStore()
+
+  const allFileLoaded = useMemo(() => (localFileList.length > 0 && localFileList.every(file => file.file.id)), [localFileList])
+
+  const hidePreviewLocalFile = useCallback(() => {
+    const { setCurrentLocalFile } = dataSourceStore.getState()
+    setCurrentLocalFile(undefined)
+  }, [dataSourceStore])
+
+  return {
+    localFileList,
+    allFileLoaded,
+    currentLocalFile,
+    hidePreviewLocalFile,
+  }
+}
+
+export const useOnlineDocument = () => {
+  const {
+    documentsData,
+    onlineDocuments,
+    currentDocument,
+  } = useDataSourceStoreWithSelector(useShallow(state => ({
+    documentsData: state.documentsData,
+    onlineDocuments: state.onlineDocuments,
+    currentDocument: state.currentDocument,
+  })))
+  const dataSourceStore = useDataSourceStore()
+
+  const currentWorkspace = documentsData[0]
+
+  const PagesMapAndSelectedPagesId: DataSourceNotionPageMap = useMemo(() => {
+    const pagesMap = (documentsData || []).reduce((prev: DataSourceNotionPageMap, next: DataSourceNotionWorkspace) => {
+      next.pages.forEach((page) => {
+        prev[page.page_id] = {
+          ...page,
+          workspace_id: next.workspace_id,
+        }
+      })
+
+      return prev
+    }, {})
+    return pagesMap
+  }, [documentsData])
+
+  const hidePreviewOnlineDocument = useCallback(() => {
+    const { setCurrentDocument } = dataSourceStore.getState()
+    setCurrentDocument(undefined)
+  }, [dataSourceStore])
+
+  const clearOnlineDocumentData = useCallback(() => {
+    const {
+      setDocumentsData,
+      setSearchValue,
+      setSelectedPagesId,
+      setOnlineDocuments,
+      setCurrentDocument,
+    } = dataSourceStore.getState()
+    setDocumentsData([])
+    setSearchValue('')
+    setSelectedPagesId(new Set())
+    setOnlineDocuments([])
+    setCurrentDocument(undefined)
+  }, [dataSourceStore])
+
+  return {
+    currentWorkspace,
+    onlineDocuments,
+    currentDocument,
+    PagesMapAndSelectedPagesId,
+    hidePreviewOnlineDocument,
+    clearOnlineDocumentData,
+  }
+}
+
+export const useWebsiteCrawl = () => {
+  const {
+    websitePages,
+    currentWebsite,
+  } = useDataSourceStoreWithSelector(useShallow(state => ({
+    websitePages: state.websitePages,
+    currentWebsite: state.currentWebsite,
+  })))
+  const dataSourceStore = useDataSourceStore()
+
+  const hideWebsitePreview = useCallback(() => {
+    const { setCurrentWebsite, setPreviewIndex } = dataSourceStore.getState()
+    setCurrentWebsite(undefined)
+    setPreviewIndex(-1)
+  }, [dataSourceStore])
+
+  const clearWebsiteCrawlData = useCallback(() => {
+    const {
+      setStep,
+      setCrawlResult,
+      setWebsitePages,
+      setPreviewIndex,
+      setCurrentWebsite,
+    } = dataSourceStore.getState()
+    setStep(CrawlStep.init)
+    setCrawlResult(undefined)
+    setCurrentWebsite(undefined)
+    setWebsitePages([])
+    setPreviewIndex(-1)
+  }, [dataSourceStore])
+
+  return {
+    websitePages,
+    currentWebsite,
+    hideWebsitePreview,
+    clearWebsiteCrawlData,
+  }
+}
+
+export const useOnlineDrive = () => {
+  const {
+    onlineDriveFileList,
+    selectedFileIds,
+  } = useDataSourceStoreWithSelector(useShallow(state => ({
+    onlineDriveFileList: state.onlineDriveFileList,
+    selectedFileIds: state.selectedFileIds,
+  })))
+  const dataSourceStore = useDataSourceStore()
+
+  const selectedOnlineDriveFileList = useMemo(() => {
+    return selectedFileIds.map(id => onlineDriveFileList.find(item => item.id === id)!)
+  }, [onlineDriveFileList, selectedFileIds])
+
+  const clearOnlineDriveData = useCallback(() => {
+    const {
+      setOnlineDriveFileList,
+      setBucket,
+      setPrefix,
+      setKeywords,
+      setSelectedFileIds,
+    } = dataSourceStore.getState()
+    setOnlineDriveFileList([])
+    setBucket('')
+    setPrefix([])
+    setKeywords('')
+    setSelectedFileIds([])
+  }, [dataSourceStore])
+
+  return {
+    onlineDriveFileList,
+    selectedFileIds,
+    selectedOnlineDriveFileList,
+    clearOnlineDriveData,
+  }
+}

dify/web/app/components/datasets/documents/create-from-pipeline/index.tsx

@@ -0,0 +1,572 @@
+'use client'
+import { useCallback, useMemo, useRef, useState } from 'react'
+import DataSourceOptions from './data-source-options'
+import type { CrawlResultItem, DocumentItem, CustomFile as File, FileIndexingEstimateResponse } from '@/models/datasets'
+import LocalFile from '@/app/components/datasets/documents/create-from-pipeline/data-source/local-file'
+import { useProviderContextSelector } from '@/context/provider-context'
+import type { NotionPage } from '@/models/common'
+import OnlineDocuments from '@/app/components/datasets/documents/create-from-pipeline/data-source/online-documents'
+import VectorSpaceFull from '@/app/components/billing/vector-space-full'
+import WebsiteCrawl from '@/app/components/datasets/documents/create-from-pipeline/data-source/website-crawl'
+import OnlineDrive from '@/app/components/datasets/documents/create-from-pipeline/data-source/online-drive'
+import Actions from './actions'
+import { useTranslation } from 'react-i18next'
+import type { Datasource } from '@/app/components/rag-pipeline/components/panel/test-run/types'
+import LeftHeader from './left-header'
+import { usePublishedPipelineInfo, useRunPublishedPipeline } from '@/service/use-pipeline'
+import { useDatasetDetailContextWithSelector } from '@/context/dataset-detail'
+import Loading from '@/app/components/base/loading'
+import type { Node } from '@/app/components/workflow/types'
+import type { DataSourceNodeType } from '@/app/components/workflow/nodes/data-source/types'
+import FilePreview from './preview/file-preview'
+import OnlineDocumentPreview from './preview/online-document-preview'
+import WebsitePreview from './preview/web-preview'
+import ProcessDocuments from './process-documents'
+import ChunkPreview from './preview/chunk-preview'
+import Processing from './processing'
+import type {
+  InitialDocumentDetail,
+  OnlineDriveFile,
+  PublishedPipelineRunPreviewResponse,
+  PublishedPipelineRunResponse,
+} from '@/models/pipeline'
+import { DatasourceType } from '@/models/pipeline'
+import { TransferMethod } from '@/types/app'
+import { useAddDocumentsSteps, useLocalFile, useOnlineDocument, useOnlineDrive, useWebsiteCrawl } from './hooks'
+import DataSourceProvider from './data-source/store/provider'
+import { useDataSourceStore } from './data-source/store'
+import { useFileUploadConfig } from '@/service/use-common'
+
+const CreateFormPipeline = () => {
+  const { t } = useTranslation()
+  const plan = useProviderContextSelector(state => state.plan)
+  const enableBilling = useProviderContextSelector(state => state.enableBilling)
+  const pipelineId = useDatasetDetailContextWithSelector(s => s.dataset?.pipeline_id)
+  const [datasource, setDatasource] = useState<Datasource>()
+  const [estimateData, setEstimateData] = useState<FileIndexingEstimateResponse | undefined>(undefined)
+  const [batchId, setBatchId] = useState('')
+  const [documents, setDocuments] = useState<InitialDocumentDetail[]>([])
+  const dataSourceStore = useDataSourceStore()
+
+  const isPreview = useRef(false)
+  const formRef = useRef<any>(null)
+
+  const { data: pipelineInfo, isFetching: isFetchingPipelineInfo } = usePublishedPipelineInfo(pipelineId || '')
+  const { data: fileUploadConfigResponse } = useFileUploadConfig()
+
+  const {
+    steps,
+    currentStep,
+    handleNextStep,
+    handleBackStep,
+  } = useAddDocumentsSteps()
+  const {
+    localFileList,
+    allFileLoaded,
+    currentLocalFile,
+    hidePreviewLocalFile,
+  } = useLocalFile()
+  const {
+    currentWorkspace,
+    onlineDocuments,
+    currentDocument,
+    PagesMapAndSelectedPagesId,
+    hidePreviewOnlineDocument,
+    clearOnlineDocumentData,
+  } = useOnlineDocument()
+  const {
+    websitePages,
+    currentWebsite,
+    hideWebsitePreview,
+    clearWebsiteCrawlData,
+  } = useWebsiteCrawl()
+  const {
+    onlineDriveFileList,
+    selectedFileIds,
+    selectedOnlineDriveFileList,
+    clearOnlineDriveData,
+  } = useOnlineDrive()
+
+  const datasourceType = useMemo(() => datasource?.nodeData.provider_type, [datasource])
+  const isVectorSpaceFull = plan.usage.vectorSpace >= plan.total.vectorSpace
+  const isShowVectorSpaceFull = useMemo(() => {
+    if (!datasource)
+      return false
+    if (datasourceType === DatasourceType.localFile)
+      return allFileLoaded && isVectorSpaceFull && enableBilling
+    if (datasourceType === DatasourceType.onlineDocument)
+      return onlineDocuments.length > 0 && isVectorSpaceFull && enableBilling
+    if (datasourceType === DatasourceType.websiteCrawl)
+      return websitePages.length > 0 && isVectorSpaceFull && enableBilling
+    if (datasourceType === DatasourceType.onlineDrive)
+      return onlineDriveFileList.length > 0 && isVectorSpaceFull && enableBilling
+    return false
+  }, [allFileLoaded, datasource, datasourceType, enableBilling, isVectorSpaceFull, onlineDocuments.length, onlineDriveFileList.length, websitePages.length])
+  const notSupportBatchUpload = enableBilling && plan.type === 'sandbox'
+
+  const nextBtnDisabled = useMemo(() => {
+    if (!datasource) return true
+    if (datasourceType === DatasourceType.localFile)
+      return isShowVectorSpaceFull || !localFileList.length || !allFileLoaded
+    if (datasourceType === DatasourceType.onlineDocument)
+      return isShowVectorSpaceFull || !onlineDocuments.length
+    if (datasourceType === DatasourceType.websiteCrawl)
+      return isShowVectorSpaceFull || !websitePages.length
+    if (datasourceType === DatasourceType.onlineDrive)
+      return isShowVectorSpaceFull || !selectedFileIds.length
+    return false
+  }, [datasource, datasourceType, isShowVectorSpaceFull, localFileList.length, allFileLoaded, onlineDocuments.length, websitePages.length, selectedFileIds.length])
+
+  const fileUploadConfig = useMemo(() => fileUploadConfigResponse ?? {
+    file_size_limit: 15,
+    batch_count_limit: 5,
+  }, [fileUploadConfigResponse])
+
+  const showSelect = useMemo(() => {
+    if (datasourceType === DatasourceType.onlineDocument) {
+      const pagesCount = currentWorkspace?.pages.length ?? 0
+      return pagesCount > 0
+    }
+    if (datasourceType === DatasourceType.onlineDrive) {
+      const isBucketList = onlineDriveFileList.some(file => file.type === 'bucket')
+      return !isBucketList && onlineDriveFileList.filter((item) => {
+        return item.type !== 'bucket'
+      }).length > 0
+    }
+  }, [currentWorkspace?.pages.length, datasourceType, onlineDriveFileList])
+
+  const totalOptions = useMemo(() => {
+    if (datasourceType === DatasourceType.onlineDocument)
+      return currentWorkspace?.pages.length
+    if (datasourceType === DatasourceType.onlineDrive) {
+      return onlineDriveFileList.filter((item) => {
+        return item.type !== 'bucket'
+      }).length
+    }
+  }, [currentWorkspace?.pages.length, datasourceType, onlineDriveFileList])
+
+  const selectedOptions = useMemo(() => {
+    if (datasourceType === DatasourceType.onlineDocument)
+      return onlineDocuments.length
+    if (datasourceType === DatasourceType.onlineDrive)
+      return selectedFileIds.length
+  }, [datasourceType, onlineDocuments.length, selectedFileIds.length])
+
+  const tip = useMemo(() => {
+    if (datasourceType === DatasourceType.onlineDocument)
+      return t('datasetPipeline.addDocuments.selectOnlineDocumentTip', { count: 50 })
+    if (datasourceType === DatasourceType.onlineDrive) {
+      return t('datasetPipeline.addDocuments.selectOnlineDriveTip', {
+        count: fileUploadConfig.batch_count_limit,
+        fileSize: fileUploadConfig.file_size_limit,
+      })
+    }
+    return ''
+  }, [datasourceType, fileUploadConfig.batch_count_limit, fileUploadConfig.file_size_limit, t])
+
+  const { mutateAsync: runPublishedPipeline, isIdle, isPending } = useRunPublishedPipeline()
+
+  const handlePreviewChunks = useCallback(async (data: Record<string, any>) => {
+    if (!datasource)
+      return
+    const {
+      previewLocalFileRef,
+      previewOnlineDocumentRef,
+      previewWebsitePageRef,
+      previewOnlineDriveFileRef,
+      currentCredentialId,
+    } = dataSourceStore.getState()
+    const datasourceInfoList: Record<string, any>[] = []
+    if (datasourceType === DatasourceType.localFile) {
+      const { id, name, type, size, extension, mime_type } = previewLocalFileRef.current as File
+      const documentInfo = {
+        related_id: id,
+        name,
+        type,
+        size,
+        extension,
+        mime_type,
+        url: '',
+        transfer_method: TransferMethod.local_file,
+        credential_id: currentCredentialId,
+      }
+      datasourceInfoList.push(documentInfo)
+    }
+    if (datasourceType === DatasourceType.onlineDocument) {
+      const { workspace_id, ...rest } = previewOnlineDocumentRef.current!
+      const documentInfo = {
+        workspace_id,
+        page: rest,
+        credential_id: currentCredentialId,
+      }
+      datasourceInfoList.push(documentInfo)
+    }
+    if (datasourceType === DatasourceType.websiteCrawl) {
+      datasourceInfoList.push({
+        ...previewWebsitePageRef.current!,
+        credential_id: currentCredentialId,
+      })
+    }
+    if (datasourceType === DatasourceType.onlineDrive) {
+      const { bucket } = dataSourceStore.getState()
+      const { id, type, name } = previewOnlineDriveFileRef.current!
+      datasourceInfoList.push({
+        bucket,
+        id,
+        name,
+        type,
+        credential_id: currentCredentialId,
+      })
+    }
+    await runPublishedPipeline({
+      pipeline_id: pipelineId!,
+      inputs: data,
+      start_node_id: datasource.nodeId,
+      datasource_type: datasourceType as DatasourceType,
+      datasource_info_list: datasourceInfoList,
+      is_preview: true,
+    }, {
+      onSuccess: (res) => {
+        setEstimateData((res as PublishedPipelineRunPreviewResponse).data.outputs)
+      },
+    })
+  }, [datasource, datasourceType, runPublishedPipeline, pipelineId, dataSourceStore])
+
+  const handleProcess = useCallback(async (data: Record<string, any>) => {
+    if (!datasource)
+      return
+    const { currentCredentialId } = dataSourceStore.getState()
+    const datasourceInfoList: Record<string, any>[] = []
+    if (datasourceType === DatasourceType.localFile) {
+      const {
+        localFileList,
+      } = dataSourceStore.getState()
+      localFileList.forEach((file) => {
+        const { id, name, type, size, extension, mime_type } = file.file
+        const documentInfo = {
+          related_id: id,
+          name,
+          type,
+          size,
+          extension,
+          mime_type,
+          url: '',
+          transfer_method: TransferMethod.local_file,
+          credential_id: currentCredentialId,
+        }
+        datasourceInfoList.push(documentInfo)
+      })
+    }
+    if (datasourceType === DatasourceType.onlineDocument) {
+      const {
+        onlineDocuments,
+      } = dataSourceStore.getState()
+      onlineDocuments.forEach((page) => {
+        const { workspace_id, ...rest } = page
+        const documentInfo = {
+          workspace_id,
+          page: rest,
+          credential_id: currentCredentialId,
+        }
+        datasourceInfoList.push(documentInfo)
+      })
+    }
+    if (datasourceType === DatasourceType.websiteCrawl) {
+      const {
+        websitePages,
+      } = dataSourceStore.getState()
+      websitePages.forEach((websitePage) => {
+        datasourceInfoList.push({
+          ...websitePage,
+          credential_id: currentCredentialId,
+        })
+      })
+    }
+    if (datasourceType === DatasourceType.onlineDrive) {
+      const {
+        bucket,
+        selectedFileIds,
+        onlineDriveFileList,
+      } = dataSourceStore.getState()
+      selectedFileIds.forEach((id) => {
+        const file = onlineDriveFileList.find(file => file.id === id)
+        datasourceInfoList.push({
+          bucket,
+          id: file?.id,
+          name: file?.name,
+          type: file?.type,
+          credential_id: currentCredentialId,
+        })
+      })
+    }
+    await runPublishedPipeline({
+      pipeline_id: pipelineId!,
+      inputs: data,
+      start_node_id: datasource.nodeId,
+      datasource_type: datasourceType as DatasourceType,
+      datasource_info_list: datasourceInfoList,
+      is_preview: false,
+    }, {
+      onSuccess: (res) => {
+        setBatchId((res as PublishedPipelineRunResponse).batch || '')
+        setDocuments((res as PublishedPipelineRunResponse).documents || [])
+        handleNextStep()
+      },
+    })
+  }, [dataSourceStore, datasource, datasourceType, handleNextStep, pipelineId, runPublishedPipeline])
+
+  const onClickProcess = useCallback(() => {
+    isPreview.current = false
+    formRef.current?.submit()
+  }, [])
+
+  const onClickPreview = useCallback(() => {
+    isPreview.current = true
+    formRef.current?.submit()
+  }, [])
+
+  const handleSubmit = useCallback((data: Record<string, any>) => {
+    if (isPreview.current)
+      handlePreviewChunks(data)
+    else
+      handleProcess(data)
+  }, [handlePreviewChunks, handleProcess])
+
+  const handlePreviewFileChange = useCallback((file: DocumentItem) => {
+    const { previewLocalFileRef } = dataSourceStore.getState()
+    previewLocalFileRef.current = file
+    onClickPreview()
+  }, [dataSourceStore, onClickPreview])
+
+  const handlePreviewOnlineDocumentChange = useCallback((page: NotionPage) => {
+    const { previewOnlineDocumentRef } = dataSourceStore.getState()
+    previewOnlineDocumentRef.current = page
+    onClickPreview()
+  }, [dataSourceStore, onClickPreview])
+
+  const handlePreviewWebsiteChange = useCallback((website: CrawlResultItem) => {
+    const { previewWebsitePageRef } = dataSourceStore.getState()
+    previewWebsitePageRef.current = website
+    onClickPreview()
+  }, [dataSourceStore, onClickPreview])
+
+  const handlePreviewOnlineDriveFileChange = useCallback((file: OnlineDriveFile) => {
+    const { previewOnlineDriveFileRef } = dataSourceStore.getState()
+    previewOnlineDriveFileRef.current = file
+    onClickPreview()
+  }, [dataSourceStore, onClickPreview])
+
+  const handleSelectAll = useCallback(() => {
+    const {
+      onlineDocuments,
+      onlineDriveFileList,
+      selectedFileIds,
+      setOnlineDocuments,
+      setSelectedFileIds,
+      setSelectedPagesId,
+    } = dataSourceStore.getState()
+    if (datasourceType === DatasourceType.onlineDocument) {
+      const allIds = currentWorkspace?.pages.map(page => page.page_id) || []
+      if (onlineDocuments.length < allIds.length) {
+        const selectedPages = Array.from(allIds).map(pageId => PagesMapAndSelectedPagesId[pageId])
+        setOnlineDocuments(selectedPages)
+        setSelectedPagesId(new Set(allIds))
+      }
+      else {
+        setOnlineDocuments([])
+        setSelectedPagesId(new Set())
+      }
+    }
+    if (datasourceType === DatasourceType.onlineDrive) {
+      const allKeys = onlineDriveFileList.filter((item) => {
+        return item.type !== 'bucket'
+      }).map(file => file.id)
+      if (selectedFileIds.length < allKeys.length)
+        setSelectedFileIds(allKeys)
+      else
+        setSelectedFileIds([])
+    }
+  }, [PagesMapAndSelectedPagesId, currentWorkspace?.pages, dataSourceStore, datasourceType])
+
+  const clearDataSourceData = useCallback((dataSource: Datasource) => {
+    if (dataSource.nodeData.provider_type === DatasourceType.onlineDocument)
+      clearOnlineDocumentData()
+    else if (dataSource.nodeData.provider_type === DatasourceType.websiteCrawl)
+      clearWebsiteCrawlData()
+    else if (dataSource.nodeData.provider_type === DatasourceType.onlineDrive)
+      clearOnlineDriveData()
+  }, [])
+
+  const handleSwitchDataSource = useCallback((dataSource: Datasource) => {
+    const {
+      setCurrentCredentialId,
+      currentNodeIdRef,
+    } = dataSourceStore.getState()
+    clearDataSourceData(dataSource)
+    setCurrentCredentialId('')
+    currentNodeIdRef.current = dataSource.nodeId
+    setDatasource(dataSource)
+  }, [dataSourceStore])
+
+  const handleCredentialChange = useCallback((credentialId: string) => {
+    const { setCurrentCredentialId } = dataSourceStore.getState()
+    clearDataSourceData(datasource!)
+    setCurrentCredentialId(credentialId)
+  }, [dataSourceStore, datasource])
+
+  if (isFetchingPipelineInfo) {
+    return (
+      <Loading type='app' />
+    )
+  }
+
+  return (
+    <div
+      className='relative flex h-[calc(100vh-56px)] w-full min-w-[1024px] overflow-x-auto rounded-t-2xl border-t border-effects-highlight bg-background-default-subtle'
+    >
+      <div className='h-full min-w-0 flex-1'>
+        <div className='flex h-full flex-col px-14'>
+          <LeftHeader
+            steps={steps}
+            title={t('datasetPipeline.addDocuments.title')}
+            currentStep={currentStep}
+          />
+          <div className='grow overflow-y-auto'>
+            {
+              currentStep === 1 && (
+                <div className='flex flex-col gap-y-5 pt-4'>
+                  <DataSourceOptions
+                    datasourceNodeId={datasource?.nodeId || ''}
+                    onSelect={handleSwitchDataSource}
+                    pipelineNodes={(pipelineInfo?.graph.nodes || []) as Node<DataSourceNodeType>[]}
+                  />
+                  {datasourceType === DatasourceType.localFile && (
+                    <LocalFile
+                      allowedExtensions={datasource!.nodeData.fileExtensions || []}
+                      notSupportBatchUpload={notSupportBatchUpload}
+                    />
+                  )}
+                  {datasourceType === DatasourceType.onlineDocument && (
+                    <OnlineDocuments
+                      nodeId={datasource!.nodeId}
+                      nodeData={datasource!.nodeData}
+                      onCredentialChange={handleCredentialChange}
+                    />
+                  )}
+                  {datasourceType === DatasourceType.websiteCrawl && (
+                    <WebsiteCrawl
+                      nodeId={datasource!.nodeId}
+                      nodeData={datasource!.nodeData}
+                      onCredentialChange={handleCredentialChange}
+                    />
+                  )}
+                  {datasourceType === DatasourceType.onlineDrive && (
+                    <OnlineDrive
+                      nodeId={datasource!.nodeId}
+                      nodeData={datasource!.nodeData}
+                      onCredentialChange={handleCredentialChange}
+                    />
+                  )}
+                  {isShowVectorSpaceFull && (
+                    <VectorSpaceFull />
+                  )}
+                  <Actions
+                    showSelect={showSelect}
+                    totalOptions={totalOptions}
+                    selectedOptions={selectedOptions}
+                    onSelectAll={handleSelectAll}
+                    disabled={nextBtnDisabled}
+                    handleNextStep={handleNextStep}
+                    tip={tip}
+                  />
+                </div>
+              )
+            }
+            {
+              currentStep === 2 && (
+                <ProcessDocuments
+                  ref={formRef}
+                  dataSourceNodeId={datasource!.nodeId}
+                  isRunning={isPending}
+                  onProcess={onClickProcess}
+                  onPreview={onClickPreview}
+                  onSubmit={handleSubmit}
+                  onBack={handleBackStep}
+                />
+              )
+            }
+            {
+              currentStep === 3 && (
+                <Processing
+                  batchId={batchId}
+                  documents={documents}
+                />
+              )
+            }
+          </div>
+        </div>
+      </div>
+      {/* Preview */}
+      {
+        currentStep === 1 && (
+          <div className='h-full min-w-0 flex-1'>
+            <div className='flex h-full flex-col pl-2 pt-2'>
+              {currentLocalFile && (
+                <FilePreview
+                  file={currentLocalFile}
+                  hidePreview={hidePreviewLocalFile}
+                />
+              )}
+              {currentDocument && (
+                <OnlineDocumentPreview
+                  datasourceNodeId={datasource!.nodeId}
+                  currentPage={currentDocument}
+                  hidePreview={hidePreviewOnlineDocument}
+                />
+              )}
+              {currentWebsite && (
+                <WebsitePreview
+                  currentWebsite={currentWebsite}
+                  hidePreview={hideWebsitePreview}
+                />
+              )}
+            </div>
+          </div>
+        )
+      }
+      {
+        currentStep === 2 && (
+          <div className='h-full min-w-0 flex-1'>
+            <div className='flex h-full flex-col pl-2 pt-2'>
+              <ChunkPreview
+                dataSourceType={datasourceType as DatasourceType}
+                localFiles={localFileList.map(file => file.file)}
+                onlineDocuments={onlineDocuments}
+                websitePages={websitePages}
+                onlineDriveFiles={selectedOnlineDriveFileList}
+                isIdle={isIdle}
+                isPending={isPending && isPreview.current}
+                estimateData={estimateData}
+                onPreview={onClickPreview}
+                handlePreviewFileChange={handlePreviewFileChange}
+                handlePreviewOnlineDocumentChange={handlePreviewOnlineDocumentChange}
+                handlePreviewWebsitePageChange={handlePreviewWebsiteChange}
+                handlePreviewOnlineDriveFileChange={handlePreviewOnlineDriveFileChange}
+              />
+            </div>
+          </div>
+        )
+      }
+    </div>
+  )
+}
+
+const CreateFormPipelineWrapper = () => {
+  return (
+    <DataSourceProvider>
+      <CreateFormPipeline />
+    </DataSourceProvider>
+  )
+}
+
+export default CreateFormPipelineWrapper

dify/web/app/components/datasets/documents/create-from-pipeline/left-header.tsx

@@ -0,0 +1,53 @@
+import React from 'react'
+import { RiArrowLeftLine } from '@remixicon/react'
+import Button from '@/app/components/base/button'
+import { useParams } from 'next/navigation'
+import Effect from '@/app/components/base/effect'
+import type { Step } from './step-indicator'
+import StepIndicator from './step-indicator'
+import Link from 'next/link'
+
+type LeftHeaderProps = {
+  steps: Array<Step>
+  title: string
+  currentStep: number
+}
+
+const LeftHeader = ({
+  steps,
+  title,
+  currentStep,
+}: LeftHeaderProps) => {
+  const { datasetId } = useParams()
+
+  return (
+    <div className='relative flex flex-col gap-y-0.5 pb-2 pt-4'>
+      <div className='flex items-center gap-x-2'>
+        <span className='system-2xs-semibold-uppercase bg-pipeline-add-documents-title-bg bg-clip-text text-transparent'>
+          {title}
+        </span>
+        <span className='system-2xs-regular text-divider-regular'>/</span>
+        <StepIndicator steps={steps} currentStep={currentStep} />
+      </div>
+      <div className='system-md-semibold text-text-primary'>
+        {steps[currentStep - 1]?.label}
+      </div>
+      {currentStep !== steps.length && (
+        <Link
+          href={`/datasets/${datasetId}/documents`}
+          replace
+        >
+          <Button
+            variant='secondary-accent'
+            className='absolute -left-11 top-3.5 size-9 rounded-full p-0'
+          >
+            <RiArrowLeftLine className='size-5 ' />
+          </Button>
+        </Link>
+      )}
+      <Effect className='left-8 top-[-34px] opacity-20' />
+    </div>
+  )
+}
+
+export default React.memo(LeftHeader)

dify/web/app/components/datasets/documents/create-from-pipeline/preview/chunk-preview.tsx

@@ -0,0 +1,238 @@
+import React, { useState } from 'react'
+import { useTranslation } from 'react-i18next'
+import PreviewContainer from '../../../preview/container'
+import { PreviewHeader } from '../../../preview/header'
+import type { CrawlResultItem, CustomFile, DocumentItem, FileIndexingEstimateResponse } from '@/models/datasets'
+import { ChunkingMode } from '@/models/datasets'
+import type { NotionPage } from '@/models/common'
+import PreviewDocumentPicker from '../../../common/document-picker/preview-document-picker'
+import { useDatasetDetailContextWithSelector } from '@/context/dataset-detail'
+import { ChunkContainer, QAPreview } from '../../../chunk'
+import { FormattedText } from '../../../formatted-text/formatted'
+import { PreviewSlice } from '../../../formatted-text/flavours/preview-slice'
+import { SkeletonContainer, SkeletonPoint, SkeletonRectangle, SkeletonRow } from '@/app/components/base/skeleton'
+import { RiSearchEyeLine } from '@remixicon/react'
+import Badge from '@/app/components/base/badge'
+import Button from '@/app/components/base/button'
+import type { OnlineDriveFile } from '@/models/pipeline'
+import { DatasourceType } from '@/models/pipeline'
+import { getFileExtension } from '../data-source/online-drive/file-list/list/utils'
+
+type ChunkPreviewProps = {
+  dataSourceType: DatasourceType
+  localFiles: CustomFile[]
+  onlineDocuments: NotionPage[]
+  websitePages: CrawlResultItem[]
+  onlineDriveFiles: OnlineDriveFile[]
+  isIdle: boolean
+  isPending: boolean
+  estimateData: FileIndexingEstimateResponse | undefined
+  onPreview: () => void
+  handlePreviewFileChange: (file: DocumentItem) => void
+  handlePreviewOnlineDocumentChange: (page: NotionPage) => void
+  handlePreviewWebsitePageChange: (page: CrawlResultItem) => void
+  handlePreviewOnlineDriveFileChange: (file: OnlineDriveFile) => void
+}
+
+const ChunkPreview = ({
+  dataSourceType,
+  localFiles,
+  onlineDocuments,
+  websitePages,
+  onlineDriveFiles,
+  isIdle,
+  isPending,
+  estimateData,
+  onPreview,
+  handlePreviewFileChange,
+  handlePreviewOnlineDocumentChange,
+  handlePreviewWebsitePageChange,
+  handlePreviewOnlineDriveFileChange,
+}: ChunkPreviewProps) => {
+  const { t } = useTranslation()
+  const currentDocForm = useDatasetDetailContextWithSelector(s => s.dataset?.doc_form)
+
+  const [previewFile, setPreviewFile] = useState<DocumentItem>(localFiles[0] as DocumentItem)
+  const [previewOnlineDocument, setPreviewOnlineDocument] = useState<NotionPage>(onlineDocuments[0])
+  const [previewWebsitePage, setPreviewWebsitePage] = useState<CrawlResultItem>(websitePages[0])
+  const [previewOnlineDriveFile, setPreviewOnlineDriveFile] = useState<OnlineDriveFile>(onlineDriveFiles[0])
+
+  return (
+    <PreviewContainer
+      header={<PreviewHeader
+        title={t('datasetCreation.stepTwo.preview')}
+      >
+        <div className='flex items-center gap-1'>
+          {dataSourceType === DatasourceType.localFile
+            && <PreviewDocumentPicker
+              files={localFiles as Array<Required<CustomFile>>}
+              onChange={(selected) => {
+                setPreviewFile(selected)
+                handlePreviewFileChange(selected)
+              }}
+              value={previewFile}
+            />
+          }
+          {dataSourceType === DatasourceType.onlineDocument
+            && <PreviewDocumentPicker
+              files={
+                onlineDocuments.map(page => ({
+                  id: page.page_id,
+                  name: page.page_name,
+                  extension: 'md',
+                }))
+              }
+              onChange={(selected) => {
+                const selectedPage = onlineDocuments.find(page => page.page_id === selected.id)
+                setPreviewOnlineDocument(selectedPage!)
+                handlePreviewOnlineDocumentChange(selectedPage!)
+              }}
+              value={{
+                id: previewOnlineDocument?.page_id || '',
+                name: previewOnlineDocument?.page_name || '',
+                extension: 'md',
+              }}
+            />
+          }
+          {dataSourceType === DatasourceType.websiteCrawl
+            && <PreviewDocumentPicker
+              files={
+                websitePages.map(page => ({
+                  id: page.source_url,
+                  name: page.title,
+                  extension: 'md',
+                }))
+              }
+              onChange={(selected) => {
+                const selectedPage = websitePages.find(page => page.source_url === selected.id)
+                setPreviewWebsitePage(selectedPage!)
+                handlePreviewWebsitePageChange(selectedPage!)
+              }}
+              value={
+                {
+                  id: previewWebsitePage?.source_url || '',
+                  name: previewWebsitePage?.title || '',
+                  extension: 'md',
+                }
+              }
+            />
+          }
+          {dataSourceType === DatasourceType.onlineDrive
+            && <PreviewDocumentPicker
+              files={
+                onlineDriveFiles.map(file => ({
+                  id: file.id,
+                  name: file.name,
+                  extension: getFileExtension(previewOnlineDriveFile?.name),
+                }))
+              }
+              onChange={(selected) => {
+                const selectedFile = onlineDriveFiles.find(file => file.id === selected.id)
+                setPreviewOnlineDriveFile(selectedFile!)
+                handlePreviewOnlineDriveFileChange(selectedFile!)
+              }}
+              value={
+                {
+                  id: previewOnlineDriveFile?.id || '',
+                  name: previewOnlineDriveFile?.name || '',
+                  extension: getFileExtension(previewOnlineDriveFile?.name),
+                }
+              }
+            />
+          }
+          {
+            currentDocForm !== ChunkingMode.qa
+            && <Badge text={t('datasetCreation.stepTwo.previewChunkCount', {
+              count: estimateData?.total_segments || 0,
+            }) as string}
+            />
+          }
+        </div>
+      </PreviewHeader>}
+      className='relative flex h-full w-full shrink-0'
+      mainClassName='space-y-6'
+    >
+      {!isPending && currentDocForm === ChunkingMode.qa && estimateData?.qa_preview && (
+        estimateData?.qa_preview.map((item, index) => (
+          <ChunkContainer
+            key={`${item.question}-${index}`}
+            label={`Chunk-${index + 1}`}
+            characterCount={item.question.length + item.answer.length}
+          >
+            <QAPreview qa={item} />
+          </ChunkContainer>
+        ))
+      )}
+      {!isPending && currentDocForm === ChunkingMode.text && estimateData?.preview && (
+        estimateData?.preview.map((item, index) => (
+          <ChunkContainer
+            key={`${item.content}-${index}`}
+            label={`Chunk-${index + 1}`}
+            characterCount={item.content.length}
+          >
+            {item.content}
+          </ChunkContainer>
+        ))
+      )}
+      {!isPending && currentDocForm === ChunkingMode.parentChild && estimateData?.preview && (
+        estimateData?.preview?.map((item, index) => {
+          const indexForLabel = index + 1
+          return (
+            <ChunkContainer
+              key={`${item.content}-${index}`}
+              label={`Chunk-${indexForLabel}`}
+              characterCount={item.content.length}
+            >
+              <FormattedText>
+                {item.child_chunks.map((child, index) => {
+                  const indexForLabel = index + 1
+                  return (
+                    <PreviewSlice
+                      key={child}
+                      label={`C-${indexForLabel}`}
+                      text={child}
+                      tooltip={`Child-chunk-${indexForLabel} · ${child.length} Characters`}
+                      labelInnerClassName='text-[10px] font-semibold align-bottom leading-7'
+                      dividerClassName='leading-7'
+                    />
+                  )
+                })}
+              </FormattedText>
+            </ChunkContainer>
+          )
+        })
+      )}
+      {isIdle && (
+        <div className='flex h-full w-full items-center justify-center'>
+          <div className='flex flex-col items-center justify-center gap-3 pb-4'>
+            <RiSearchEyeLine className='size-10 text-text-empty-state-icon' />
+            <p className='text-sm text-text-tertiary'>
+              {t('datasetCreation.stepTwo.previewChunkTip')}
+            </p>
+            <Button onClick={onPreview}>
+              {t('datasetPipeline.addDocuments.stepTwo.previewChunks')}
+            </Button>
+          </div>
+        </div>
+      )}
+      {isPending && (
+        <div className='h-full w-full space-y-6 overflow-hidden'>
+          {Array.from({ length: 10 }, (_, i) => (
+            <SkeletonContainer key={i}>
+              <SkeletonRow>
+                <SkeletonRectangle className='w-20' />
+                <SkeletonPoint />
+                <SkeletonRectangle className='w-24' />
+              </SkeletonRow>
+              <SkeletonRectangle className='w-full' />
+              <SkeletonRectangle className='w-full' />
+              <SkeletonRectangle className='w-[422px]' />
+            </SkeletonContainer>
+          ))}
+        </div>
+      )}
+    </PreviewContainer>
+  )
+}
+
+export default React.memo(ChunkPreview)

dify/web/app/components/datasets/documents/create-from-pipeline/preview/file-preview.tsx

@@ -0,0 +1,75 @@
+'use client'
+import React, { useMemo } from 'react'
+import { useTranslation } from 'react-i18next'
+import Loading from './loading'
+import type { CustomFile as File } from '@/models/datasets'
+import { RiCloseLine } from '@remixicon/react'
+import { useFilePreview } from '@/service/use-common'
+import DocumentFileIcon from '../../../common/document-file-icon'
+import { formatFileSize, formatNumberAbbreviated } from '@/utils/format'
+
+type FilePreviewProps = {
+  file: File
+  hidePreview: () => void
+}
+
+const FilePreview = ({
+  file,
+  hidePreview,
+}: FilePreviewProps) => {
+  const { t } = useTranslation()
+  const { data: fileData, isFetching } = useFilePreview(file.id || '')
+
+  const fileName = useMemo(() => {
+    if (!file)
+      return ''
+    const arr = file.name.split('.')
+    return arr.slice(0, -1).join()
+  }, [file])
+
+  return (
+    <div className='flex h-full w-full flex-col rounded-t-xl border-l border-t border-components-panel-border bg-background-default-lighter shadow-md shadow-shadow-shadow-5'>
+      <div className='flex gap-x-2 border-b border-divider-subtle pb-3 pl-6 pr-4 pt-4'>
+        <div className='flex grow flex-col gap-y-1'>
+          <div className='system-2xs-semibold-uppercase text-text-accent'>{t('datasetPipeline.addDocuments.stepOne.preview')}</div>
+          <div className='title-md-semi-bold text-tex-primary'>{`${fileName}.${file.extension || ''}`}</div>
+          <div className='system-xs-medium flex items-center gap-x-1 text-text-tertiary'>
+            <DocumentFileIcon
+              className='size-3.5 shrink-0'
+              name={file.name}
+              extension={file.extension}
+            />
+            <span className='uppercase'>{file.extension}</span>
+            <span>·</span>
+            <span>{formatFileSize(file.size)}</span>
+            {fileData && (
+              <>
+                <span>·</span>
+                <span>{`${formatNumberAbbreviated(fileData.content.length)} ${t('datasetPipeline.addDocuments.characters')}`}</span>
+              </>
+            )}
+          </div>
+        </div>
+        <button
+          type='button'
+          className='flex h-8 w-8 shrink-0 items-center justify-center'
+          onClick={hidePreview}
+        >
+          <RiCloseLine className='size-[18px]' />
+        </button>
+      </div>
+      {isFetching && (
+        <div className='grow'>
+          <Loading />
+        </div>
+      )}
+      {!isFetching && fileData && (
+        <div className='body-md-regular grow overflow-hidden px-6 py-5 text-text-secondary'>
+          {fileData.content}
+        </div>
+      )}
+    </div>
+  )
+}
+
+export default FilePreview

dify/web/app/components/datasets/documents/create-from-pipeline/preview/loading.tsx

@@ -0,0 +1,52 @@
+import React from 'react'
+import { SkeletonContainer, SkeletonRectangle } from '@/app/components/base/skeleton'
+
+const Loading = () => {
+  return (
+    <div className='flex h-full w-full flex-col gap-y-3 overflow-hidden bg-gradient-to-b from-components-panel-bg-transparent to-components-panel-bg px-6 py-5'>
+      <SkeletonContainer className='w-full gap-0'>
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-3/5' />
+      </SkeletonContainer>
+      <SkeletonContainer className='w-full gap-0'>
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-[70%]' />
+      </SkeletonContainer>
+      <SkeletonContainer className='w-full gap-0'>
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-[56%]' />
+      </SkeletonContainer>
+      <SkeletonContainer className='w-full gap-0'>
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-3/5' />
+      </SkeletonContainer>
+      <SkeletonContainer className='w-full gap-0'>
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-3/5' />
+      </SkeletonContainer>
+      <SkeletonContainer className='w-full gap-0'>
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-full' />
+        <SkeletonRectangle className='my-1.5 w-1/2' />
+      </SkeletonContainer>
+    </div>
+  )
+}
+
+export default React.memo(Loading)

dify/web/app/components/datasets/documents/create-from-pipeline/preview/online-document-preview.tsx

@@ -0,0 +1,89 @@
+'use client'
+import React, { useEffect, useState } from 'react'
+import { useTranslation } from 'react-i18next'
+import type { NotionPage } from '@/models/common'
+import { RiCloseLine } from '@remixicon/react'
+import { formatNumberAbbreviated } from '@/utils/format'
+import Loading from './loading'
+import { Notion } from '@/app/components/base/icons/src/public/common'
+import { useDatasetDetailContextWithSelector } from '@/context/dataset-detail'
+import { usePreviewOnlineDocument } from '@/service/use-pipeline'
+import Toast from '@/app/components/base/toast'
+import { Markdown } from '@/app/components/base/markdown'
+import { useDataSourceStore } from '../data-source/store'
+
+type OnlineDocumentPreviewProps = {
+  currentPage: NotionPage
+  datasourceNodeId: string
+  hidePreview: () => void
+}
+
+const OnlineDocumentPreview = ({
+  currentPage,
+  datasourceNodeId,
+  hidePreview,
+}: OnlineDocumentPreviewProps) => {
+  const { t } = useTranslation()
+  const [content, setContent] = useState('')
+  const pipelineId = useDatasetDetailContextWithSelector(state => state.dataset?.pipeline_id)
+  const { mutateAsync: getOnlineDocumentContent, isPending } = usePreviewOnlineDocument()
+  const dataSourceStore = useDataSourceStore()
+
+  useEffect(() => {
+    const { currentCredentialId } = dataSourceStore.getState()
+    getOnlineDocumentContent({
+      workspaceID: currentPage.workspace_id,
+      pageID: currentPage.page_id,
+      pageType: currentPage.type,
+      pipelineId: pipelineId || '',
+      datasourceNodeId,
+      credentialId: currentCredentialId,
+    }, {
+      onSuccess(data) {
+        setContent(data.content)
+      },
+      onError(error) {
+        Toast.notify({
+          type: 'error',
+          message: error.message,
+        })
+      },
+    })
+  }, [currentPage.page_id])
+
+  return (
+    <div className='flex h-full w-full flex-col rounded-t-xl border-l border-t border-components-panel-border bg-background-default-lighter shadow-md shadow-shadow-shadow-5'>
+      <div className='flex gap-x-2 border-b border-divider-subtle pb-3 pl-6 pr-4 pt-4'>
+        <div className='flex grow flex-col gap-y-1'>
+          <div className='system-2xs-semibold-uppercase text-text-accent'>{t('datasetPipeline.addDocuments.stepOne.preview')}</div>
+          <div className='title-md-semi-bold text-tex-primary'>{currentPage?.page_name}</div>
+          <div className='system-xs-medium flex items-center gap-x-1 text-text-tertiary'>
+            <Notion className='size-3.5' />
+            <span>{currentPage.type}</span>
+            <span>·</span>
+            <span>{`${formatNumberAbbreviated(content.length)} ${t('datasetPipeline.addDocuments.characters')}`}</span>
+          </div>
+        </div>
+        <button
+          type='button'
+          className='flex h-8 w-8 shrink-0 items-center justify-center'
+          onClick={hidePreview}
+        >
+          <RiCloseLine className='size-[18px]' />
+        </button>
+      </div>
+      {isPending && (
+        <div className='grow'>
+          <Loading />
+        </div>
+      )}
+      {!isPending && content && (
+        <div className='body-md-regular grow overflow-hidden px-6 py-5 text-text-secondary'>
+          <Markdown content={content} />
+        </div>
+      )}
+    </div>
+  )
+}
+
+export default OnlineDocumentPreview

dify/web/app/components/datasets/documents/create-from-pipeline/preview/web-preview.tsx

@@ -0,0 +1,48 @@
+'use client'
+import React from 'react'
+import { useTranslation } from 'react-i18next'
+import type { CrawlResultItem } from '@/models/datasets'
+import { RiCloseLine, RiGlobalLine } from '@remixicon/react'
+import { formatNumberAbbreviated } from '@/utils/format'
+
+type WebsitePreviewProps = {
+  currentWebsite: CrawlResultItem
+  hidePreview: () => void
+}
+
+const WebsitePreview = ({
+  currentWebsite,
+  hidePreview,
+}: WebsitePreviewProps) => {
+  const { t } = useTranslation()
+
+  return (
+    <div className='flex h-full w-full flex-col rounded-t-xl border-l border-t border-components-panel-border bg-background-default-lighter shadow-md shadow-shadow-shadow-5'>
+      <div className='flex gap-x-2 border-b border-divider-subtle pb-3 pl-6 pr-4 pt-4'>
+        <div className='flex grow flex-col gap-y-1'>
+          <div className='system-2xs-semibold-uppercase'>{t('datasetPipeline.addDocuments.stepOne.preview')}</div>
+          <div className='title-md-semi-bold text-tex-primary'>{currentWebsite.title}</div>
+          <div className='system-xs-medium flex gap-x-1  text-text-tertiary'>
+            <RiGlobalLine className='size-3.5' />
+            <span className='uppercase' title={currentWebsite.source_url}>{currentWebsite.source_url}</span>
+            <span>·</span>
+            <span>·</span>
+            <span>{`${formatNumberAbbreviated(currentWebsite.content.length)} ${t('datasetPipeline.addDocuments.characters')}`}</span>
+          </div>
+        </div>
+        <button
+          type='button'
+          className='flex h-8 w-8 shrink-0 items-center justify-center'
+          onClick={hidePreview}
+        >
+          <RiCloseLine className='size-[18px]' />
+        </button>
+      </div>
+      <div className='body-md-regular grow overflow-hidden px-6 py-5 text-text-secondary'>
+        {currentWebsite.content}
+      </div>
+    </div>
+  )
+}
+
+export default WebsitePreview

dify/web/app/components/datasets/documents/create-from-pipeline/process-documents/actions.tsx

@@ -0,0 +1,40 @@
+import React from 'react'
+import Button from '@/app/components/base/button'
+import { useTranslation } from 'react-i18next'
+import { RiArrowLeftLine } from '@remixicon/react'
+
+type ActionsProps = {
+  onBack: () => void
+  runDisabled?: boolean
+  onProcess: () => void
+}
+
+const Actions = ({
+  onBack,
+  runDisabled,
+  onProcess,
+}: ActionsProps) => {
+  const { t } = useTranslation()
+
+  return (
+    <div className='flex items-center justify-between'>
+      <Button
+        variant='secondary'
+        onClick={onBack}
+        className='gap-x-0.5'
+      >
+        <RiArrowLeftLine className='size-4' />
+        <span className='px-0.5'>{t('datasetPipeline.operations.dataSource')}</span>
+      </Button>
+      <Button
+        variant='primary'
+        disabled={runDisabled}
+        onClick={onProcess}
+      >
+        {t('datasetPipeline.operations.saveAndProcess')}
+      </Button>
+    </div>
+  )
+}
+
+export default React.memo(Actions)

dify/web/app/components/datasets/documents/create-from-pipeline/process-documents/form.tsx

@@ -0,0 +1,96 @@
+import { useAppForm } from '@/app/components/base/form'
+import BaseField from '@/app/components/base/form/form-scenarios/base/field'
+import type { BaseConfiguration } from '@/app/components/base/form/form-scenarios/base/types'
+import Toast from '@/app/components/base/toast'
+import { useCallback, useImperativeHandle } from 'react'
+import type { ZodSchema } from 'zod'
+import Header from './header'
+
+type OptionsProps = {
+  initialData: Record<string, any>
+  configurations: BaseConfiguration[]
+  schema: ZodSchema
+  onSubmit: (data: Record<string, any>) => void
+  onPreview: () => void
+  ref: React.RefObject<any>
+  isRunning: boolean
+}
+
+const Form = ({
+  initialData,
+  configurations,
+  schema,
+  onSubmit,
+  onPreview,
+  ref,
+  isRunning,
+}: OptionsProps) => {
+  const form = useAppForm({
+    defaultValues: initialData,
+    validators: {
+      onSubmit: ({ value }) => {
+        const result = schema.safeParse(value)
+        if (!result.success) {
+          const issues = result.error.issues
+          const firstIssue = issues[0]
+          const errorMessage = `"${firstIssue.path.join('.')}" ${firstIssue.message}`
+          Toast.notify({
+            type: 'error',
+            message: errorMessage,
+          })
+          return errorMessage
+        }
+        return undefined
+      },
+    },
+    onSubmit: ({ value }) => {
+      onSubmit(value)
+    },
+  })
+
+  useImperativeHandle(ref, () => {
+    return {
+      submit: () => {
+        form.handleSubmit()
+      },
+    }
+  }, [form])
+
+  const handleReset = useCallback(() => {
+    form.reset()
+  }, [form])
+
+  return (
+    <form
+      className='flex w-full flex-col rounded-lg border border-components-panel-border bg-components-panel-bg'
+      onSubmit={(e) => {
+        e.preventDefault()
+        e.stopPropagation()
+        form.handleSubmit()
+      }}
+    >
+      <form.Subscribe
+        selector={state => state.isDirty}
+        children={isDirty => (
+          <Header
+            onReset={handleReset}
+            resetDisabled={!isDirty}
+            onPreview={onPreview}
+            previewDisabled={isRunning}
+          />
+        )}
+      />
+      <div className='flex flex-col gap-3 border-t border-divider-subtle px-4 py-3'>
+        {configurations.map((config, index) => {
+          const FieldComponent = BaseField({
+            initialData,
+            config,
+          })
+          return <FieldComponent key={index} form={form} />
+        })}
+      </div>
+    </form>
+  )
+}
+
+export default Form

dify/web/app/components/datasets/documents/create-from-pipeline/process-documents/header.tsx

@@ -0,0 +1,42 @@
+import React from 'react'
+import Button from '@/app/components/base/button'
+import { useTranslation } from 'react-i18next'
+import { RiSearchEyeLine } from '@remixicon/react'
+
+type HeaderProps = {
+  onReset: () => void
+  resetDisabled: boolean
+  previewDisabled: boolean
+  onPreview?: () => void
+}
+
+const Header = ({
+  onReset,
+  resetDisabled,
+  previewDisabled,
+  onPreview,
+}: HeaderProps) => {
+  const { t } = useTranslation()
+
+  return (
+    <div className='flex items-center gap-x-1 px-4 py-2'>
+      <div className='system-sm-semibold-uppercase grow text-text-secondary'>
+        {t('datasetPipeline.addDocuments.stepTwo.chunkSettings')}
+      </div>
+      <Button variant='ghost' disabled={resetDisabled} onClick={onReset}>
+        {t('common.operation.reset')}
+      </Button>
+      <Button
+        variant='secondary-accent'
+        onClick={onPreview}
+        className='gap-x-0.5'
+        disabled={previewDisabled}
+      >
+        <RiSearchEyeLine className='size-4' />
+        <span className='px-0.5'>{t('datasetPipeline.addDocuments.stepTwo.previewChunks')}</span>
+      </Button>
+    </div>
+  )
+}
+
+export default React.memo(Header)

dify/web/app/components/datasets/documents/create-from-pipeline/process-documents/hooks.ts

@@ -0,0 +1,15 @@
+import { usePublishedPipelineProcessingParams } from '@/service/use-pipeline'
+import { useDatasetDetailContextWithSelector } from '@/context/dataset-detail'
+
+export const useInputVariables = (datasourceNodeId: string) => {
+  const pipelineId = useDatasetDetailContextWithSelector(state => state.dataset?.pipeline_id)
+  const { data: paramsConfig, isFetching: isFetchingParams } = usePublishedPipelineProcessingParams({
+    pipeline_id: pipelineId!,
+    node_id: datasourceNodeId,
+  })
+
+  return {
+    paramsConfig,
+    isFetchingParams,
+  }
+}

dify/web/app/components/datasets/documents/create-from-pipeline/process-documents/index.tsx

@@ -0,0 +1,48 @@
+import React from 'react'
+import { generateZodSchema } from '@/app/components/base/form/form-scenarios/base/utils'
+import { useInputVariables } from './hooks'
+import Form from './form'
+import Actions from './actions'
+import { useConfigurations, useInitialData } from '@/app/components/rag-pipeline/hooks/use-input-fields'
+
+type ProcessDocumentsProps = {
+  dataSourceNodeId: string
+  ref: React.RefObject<any>
+  isRunning: boolean
+  onProcess: () => void
+  onPreview: () => void
+  onSubmit: (data: Record<string, any>) => void
+  onBack: () => void
+}
+
+const ProcessDocuments = ({
+  dataSourceNodeId,
+  isRunning,
+  onProcess,
+  onPreview,
+  onSubmit,
+  onBack,
+  ref,
+}: ProcessDocumentsProps) => {
+  const { isFetchingParams, paramsConfig } = useInputVariables(dataSourceNodeId)
+  const initialData = useInitialData(paramsConfig?.variables || [])
+  const configurations = useConfigurations(paramsConfig?.variables || [])
+  const schema = generateZodSchema(configurations)
+
+  return (
+    <div className='flex flex-col gap-y-4 pt-4'>
+      <Form
+        ref={ref}
+        initialData={initialData}
+        configurations={configurations}
+        schema={schema}
+        onSubmit={onSubmit}
+        onPreview={onPreview}
+        isRunning={isRunning}
+      />
+      <Actions runDisabled={isFetchingParams || isRunning} onBack={onBack} onProcess={onProcess} />
+    </div>
+  )
+}
+
+export default React.memo(ProcessDocuments)

dify/web/app/components/datasets/documents/create-from-pipeline/processing/embedding-process/index.tsx

@@ -0,0 +1,251 @@
+import React, { useEffect, useMemo, useState } from 'react'
+import { useRouter } from 'next/navigation'
+import { useTranslation } from 'react-i18next'
+import {
+  RiAedFill,
+  RiArrowRightLine,
+  RiCheckboxCircleFill,
+  RiErrorWarningFill,
+  RiLoader2Fill,
+  RiTerminalBoxLine,
+} from '@remixicon/react'
+import cn from '@/utils/classnames'
+import Button from '@/app/components/base/button'
+import type { IndexingStatusResponse } from '@/models/datasets'
+import NotionIcon from '@/app/components/base/notion-icon'
+import PriorityLabel from '@/app/components/billing/priority-label'
+import { Plan } from '@/app/components/billing/type'
+import UpgradeBtn from '@/app/components/billing/upgrade-btn'
+import { useProviderContext } from '@/context/provider-context'
+import Tooltip from '@/app/components/base/tooltip'
+import { useInvalidDocumentList } from '@/service/knowledge/use-document'
+import DocumentFileIcon from '@/app/components/datasets/common/document-file-icon'
+import RuleDetail from './rule-detail'
+import type { IndexingType } from '@/app/components/datasets/create/step-two'
+import type { RETRIEVE_METHOD } from '@/types/app'
+import { DatasourceType, type InitialDocumentDetail } from '@/models/pipeline'
+import { useIndexingStatusBatch, useProcessRule } from '@/service/knowledge/use-dataset'
+import Divider from '@/app/components/base/divider'
+import { useDatasetApiAccessUrl } from '@/hooks/use-api-access-url'
+import Link from 'next/link'
+
+type EmbeddingProcessProps = {
+  datasetId: string
+  batchId: string
+  documents?: InitialDocumentDetail[]
+  indexingType?: IndexingType
+  retrievalMethod?: RETRIEVE_METHOD
+}
+
+const EmbeddingProcess = ({
+  datasetId,
+  batchId,
+  documents = [],
+  indexingType,
+  retrievalMethod,
+}: EmbeddingProcessProps) => {
+  const { t } = useTranslation()
+  const router = useRouter()
+  const { enableBilling, plan } = useProviderContext()
+  const [indexingStatusBatchDetail, setIndexingStatusDetail] = useState<IndexingStatusResponse[]>([])
+  const [shouldPoll, setShouldPoll] = useState(true)
+  const { mutateAsync: fetchIndexingStatus } = useIndexingStatusBatch({ datasetId, batchId })
+
+  useEffect(() => {
+    let timeoutId: ReturnType<typeof setTimeout>
+
+    const fetchData = async () => {
+      await fetchIndexingStatus(undefined, {
+        onSuccess: (res) => {
+          const indexingStatusDetailList = res.data
+          setIndexingStatusDetail(indexingStatusDetailList)
+          const isCompleted = indexingStatusDetailList.every(indexingStatusDetail => ['completed', 'error', 'paused'].includes(indexingStatusDetail.indexing_status))
+          if (isCompleted)
+            setShouldPoll(false)
+        },
+        onSettled: () => {
+          if (shouldPoll)
+            timeoutId = setTimeout(fetchData, 2500)
+        },
+      })
+    }
+
+    fetchData()
+
+    return () => {
+      clearTimeout(timeoutId)
+    }
+  }, [shouldPoll])
+
+  // get rule
+  const firstDocument = documents[0]
+  const { data: ruleDetail } = useProcessRule(firstDocument.id)
+
+  const invalidDocumentList = useInvalidDocumentList()
+  const navToDocumentList = () => {
+    invalidDocumentList()
+    router.push(`/datasets/${datasetId}/documents`)
+  }
+  const apiReferenceUrl = useDatasetApiAccessUrl()
+
+  const isEmbeddingWaiting = useMemo(() => {
+    if (!indexingStatusBatchDetail.length) return false
+    return indexingStatusBatchDetail.every(indexingStatusDetail => ['waiting'].includes(indexingStatusDetail?.indexing_status || ''))
+  }, [indexingStatusBatchDetail])
+  const isEmbedding = useMemo(() => {
+    if (!indexingStatusBatchDetail.length) return false
+    return indexingStatusBatchDetail.some(indexingStatusDetail => ['indexing', 'splitting', 'parsing', 'cleaning'].includes(indexingStatusDetail?.indexing_status || ''))
+  }, [indexingStatusBatchDetail])
+  const isEmbeddingCompleted = useMemo(() => {
+    if (!indexingStatusBatchDetail.length) return false
+    return indexingStatusBatchDetail.every(indexingStatusDetail => ['completed', 'error', 'paused'].includes(indexingStatusDetail?.indexing_status || ''))
+  }, [indexingStatusBatchDetail])
+
+  const getSourceName = (id: string) => {
+    const doc = documents.find(document => document.id === id)
+    return doc?.name
+  }
+  const getFileType = (name?: string) => name?.split('.').pop() || 'txt'
+  const getSourcePercent = (detail: IndexingStatusResponse) => {
+    const completedCount = detail.completed_segments || 0
+    const totalCount = detail.total_segments || 0
+    if (totalCount === 0)
+      return 0
+    const percent = Math.round(completedCount * 100 / totalCount)
+    return percent > 100 ? 100 : percent
+  }
+  const getSourceType = (id: string) => {
+    const doc = documents.find(document => document.id === id)
+    return doc?.data_source_type
+  }
+  const getIcon = (id: string) => {
+    const doc = documents.find(document => document.id === id)
+
+    return doc?.data_source_info.notion_page_icon
+  }
+  const isSourceEmbedding = (detail: IndexingStatusResponse) =>
+    ['indexing', 'splitting', 'parsing', 'cleaning', 'waiting'].includes(detail.indexing_status || '')
+
+  return (
+    <>
+      <div className='flex flex-col gap-y-3'>
+        <div className='system-md-semibold-uppercase flex items-center gap-x-1 text-text-secondary'>
+          {(isEmbeddingWaiting || isEmbedding) && (
+            <>
+              <RiLoader2Fill className='size-4 animate-spin' />
+              <span>
+                {isEmbeddingWaiting ? t('datasetDocuments.embedding.waiting') : t('datasetDocuments.embedding.processing')}
+              </span>
+            </>
+          )}
+          {isEmbeddingCompleted && t('datasetDocuments.embedding.completed')}
+        </div>
+        {
+          enableBilling && plan.type !== Plan.team && (
+            <div className='flex h-[52px] items-center gap-x-2 rounded-xl border-[0.5px] border-components-panel-border-subtle bg-components-panel-on-panel-item-bg p-2.5 pl-3 shadow-xs shadow-shadow-shadow-3'>
+              <div className='flex shrink-0 items-center justify-center rounded-lg border-[0.5px] border-divider-subtle bg-util-colors-blue-brand-blue-brand-500 shadow-md shadow-shadow-shadow-5'>
+                <RiAedFill className='size-4 text-text-primary-on-surface' />
+              </div>
+              <div className='system-md-medium grow text-text-primary'>
+                {t('billing.plansCommon.documentProcessingPriorityUpgrade')}
+              </div>
+              <UpgradeBtn loc='knowledge-speed-up' />
+            </div>
+          )
+        }
+        <div className='flex flex-col gap-0.5 pb-2'>
+          {indexingStatusBatchDetail.map(indexingStatusDetail => (
+            <div
+              key={indexingStatusDetail.id}
+              className={cn(
+                'relative h-[26px] overflow-hidden rounded-md bg-components-progress-bar-bg',
+                indexingStatusDetail.indexing_status === 'error' && 'bg-state-destructive-hover-alt',
+              )}
+            >
+              {isSourceEmbedding(indexingStatusDetail) && (
+                <div
+                  className='absolute left-0 top-0 h-full min-w-0.5 border-r-[2px] border-r-components-progress-bar-progress-highlight bg-components-progress-bar-progress'
+                  style={{ width: `${getSourcePercent(indexingStatusDetail)}%` }}
+                />
+              )}
+              <div className='z-[1] flex h-full items-center gap-1 pl-[6px] pr-2'>
+                {getSourceType(indexingStatusDetail.id) === DatasourceType.localFile && (
+                  <DocumentFileIcon
+                    size='sm'
+                    className='shrink-0'
+                    name={getSourceName(indexingStatusDetail.id)}
+                    extension={getFileType(getSourceName(indexingStatusDetail.id))}
+                  />
+                )}
+                {getSourceType(indexingStatusDetail.id) === DatasourceType.onlineDocument && (
+                  <NotionIcon
+                    className='shrink-0'
+                    type='page'
+                    src={getIcon(indexingStatusDetail.id)}
+                  />
+                )}
+                <div className='flex w-0 grow items-center gap-1' title={getSourceName(indexingStatusDetail.id)}>
+                  <div className='system-xs-medium truncate text-text-secondary'>
+                    {getSourceName(indexingStatusDetail.id)}
+                  </div>
+                  {
+                    enableBilling && (
+                      <PriorityLabel className='ml-0' />
+                    )
+                  }
+                </div>
+                {isSourceEmbedding(indexingStatusDetail) && (
+                  <div className='shrink-0 text-xs text-text-secondary'>{`${getSourcePercent(indexingStatusDetail)}%`}</div>
+                )}
+                {indexingStatusDetail.indexing_status === 'error' && (
+                  <Tooltip
+                    popupClassName='px-4 py-[14px] max-w-60 body-xs-regular text-text-secondary border-[0.5px] border-components-panel-border rounded-xl'
+                    offset={4}
+                    popupContent={indexingStatusDetail.error}
+                  >
+                    <span>
+                      <RiErrorWarningFill className='size-4 shrink-0 text-text-destructive' />
+                    </span>
+                  </Tooltip>
+                )}
+                {indexingStatusDetail.indexing_status === 'completed' && (
+                  <RiCheckboxCircleFill className='size-4 shrink-0 text-text-success' />
+                )}
+              </div>
+            </div>
+          ))}
+        </div>
+        <Divider type='horizontal' className='my-0 bg-divider-subtle' />
+        <RuleDetail
+          sourceData={ruleDetail}
+          indexingType={indexingType}
+          retrievalMethod={retrievalMethod}
+        />
+      </div>
+      <div className='mt-6 flex items-center gap-x-2 py-2'>
+        <Link
+          href={apiReferenceUrl}
+          target='_blank'
+          rel='noopener noreferrer'
+        >
+          <Button
+            className='w-fit gap-x-0.5 px-3'
+          >
+            <RiTerminalBoxLine className='size-4' />
+            <span className='px-0.5'>Access the API</span>
+          </Button>
+        </Link>
+        <Button
+          className='w-fit gap-x-0.5 px-3'
+          variant='primary'
+          onClick={navToDocumentList}
+        >
+          <span className='px-0.5'>{t('datasetCreation.stepThree.navTo')}</span>
+          <RiArrowRightLine className='size-4 stroke-current stroke-1' />
+        </Button>
+      </div>
+    </>
+  )
+}
+
+export default EmbeddingProcess

dify/web/app/components/datasets/documents/create-from-pipeline/processing/embedding-process/rule-detail.tsx

@@ -0,0 +1,84 @@
+import React, { useCallback } from 'react'
+import { IndexingType } from '@/app/components/datasets/create/step-two'
+import { ProcessMode, type ProcessRuleResponse } from '@/models/datasets'
+import { RETRIEVE_METHOD } from '@/types/app'
+import { useTranslation } from 'react-i18next'
+import { FieldInfo } from '@/app/components/datasets/documents/detail/metadata'
+import Image from 'next/image'
+import { indexMethodIcon, retrievalIcon } from '@/app/components/datasets/create/icons'
+
+type RuleDetailProps = {
+  sourceData?: ProcessRuleResponse
+  indexingType?: IndexingType
+  retrievalMethod?: RETRIEVE_METHOD
+}
+
+const RuleDetail = ({
+  sourceData,
+  indexingType,
+  retrievalMethod,
+}: RuleDetailProps) => {
+  const { t } = useTranslation()
+
+  const getValue = useCallback((field: string) => {
+    let value = '-'
+    switch (field) {
+      case 'mode':
+        value = !sourceData?.mode
+          ? value
+          // eslint-disable-next-line sonarjs/no-nested-conditional
+          : sourceData.mode === ProcessMode.general
+            ? (t('datasetDocuments.embedding.custom') as string)
+            // eslint-disable-next-line sonarjs/no-nested-conditional
+            : `${t('datasetDocuments.embedding.hierarchical')} · ${sourceData?.rules?.parent_mode === 'paragraph'
+              ? t('dataset.parentMode.paragraph')
+              : t('dataset.parentMode.fullDoc')}`
+        break
+    }
+    return value
+  }, [sourceData, t])
+
+  return (
+    <div className='flex flex-col gap-1'>
+      <FieldInfo
+        label={t('datasetDocuments.embedding.mode')}
+        displayedValue={getValue('mode')}
+      />
+      <FieldInfo
+        label={t('datasetCreation.stepTwo.indexMode')}
+        displayedValue={t(`datasetCreation.stepTwo.${indexingType === IndexingType.ECONOMICAL ? 'economical' : 'qualified'}`) as string}
+        valueIcon={
+          <Image
+            className='size-4'
+            src={
+              indexingType === IndexingType.ECONOMICAL
+                ? indexMethodIcon.economical
+                : indexMethodIcon.high_quality
+            }
+            alt=''
+          />
+        }
+      />
+      <FieldInfo
+        label={t('datasetSettings.form.retrievalSetting.title')}
+        displayedValue={t(`dataset.retrieval.${indexingType === IndexingType.ECONOMICAL ? 'keyword_search' : retrievalMethod}.title`) as string}
+        valueIcon={
+          <Image
+            className='size-4'
+            src={
+              retrievalMethod === RETRIEVE_METHOD.fullText
+                ? retrievalIcon.fullText
+                // eslint-disable-next-line sonarjs/no-nested-conditional
+                : retrievalMethod === RETRIEVE_METHOD.hybrid
+                  ? retrievalIcon.hybrid
+                  : retrievalIcon.vector
+            }
+            alt=''
+          />
+        }
+      />
+    </div>
+  )
+}
+
+export default React.memo(RuleDetail)

dify/web/app/components/datasets/documents/create-from-pipeline/processing/index.tsx

@@ -0,0 +1,61 @@
+'use client'
+import React from 'react'
+import { useTranslation } from 'react-i18next'
+import { RiBookOpenLine } from '@remixicon/react'
+import { useDocLink } from '@/context/i18n'
+import EmbeddingProcess from './embedding-process'
+import type { InitialDocumentDetail } from '@/models/pipeline'
+import { useDatasetDetailContextWithSelector } from '@/context/dataset-detail'
+
+type ProcessingProps = {
+  batchId: string
+  documents: InitialDocumentDetail[]
+}
+
+const Processing = ({
+  batchId,
+  documents,
+}: ProcessingProps) => {
+  const { t } = useTranslation()
+  const docLink = useDocLink()
+  const datasetId = useDatasetDetailContextWithSelector(s => s.dataset?.id)
+  const indexingType = useDatasetDetailContextWithSelector(s => s.dataset?.indexing_technique)
+  const retrievalMethod = useDatasetDetailContextWithSelector(s => s.dataset?.retrieval_model_dict.search_method)
+
+  return (
+    <div className='flex h-full w-full justify-center overflow-hidden'>
+      <div className='h-full w-3/5 overflow-y-auto pb-8 pt-10'>
+        <div className='max-w-[640px]'>
+          <EmbeddingProcess
+            datasetId={datasetId!}
+            batchId={batchId}
+            documents={documents}
+            indexingType={indexingType}
+            retrievalMethod={retrievalMethod}
+          />
+        </div>
+      </div>
+      <div className='w-2/5 pr-8 pt-[88px]'>
+        <div className='flex w-[328px] flex-col gap-3 rounded-xl bg-background-section p-6'>
+          <div className='flex size-10 items-center justify-center rounded-[10px] bg-components-card-bg shadow-lg shadow-shadow-shadow-5'>
+            <RiBookOpenLine className='size-5 text-text-accent' />
+          </div>
+          <div className='flex flex-col gap-y-2'>
+            <div className='system-xl-semibold text-text-secondary'>{t('datasetCreation.stepThree.sideTipTitle')}</div>
+            <div className='system-sm-regular text-text-tertiary'>{t('datasetCreation.stepThree.sideTipContent')}</div>
+            <a
+              href={docLink('/guides/knowledge-base/integrate-knowledge-within-application')}
+              target='_blank'
+              rel='noreferrer noopener'
+              className='system-sm-regular text-text-accent'
+            >
+              {t('datasetPipeline.addDocuments.stepThree.learnMore')}
+            </a>
+          </div>
+        </div>
+      </div>
+    </div>
+  )
+}
+
+export default Processing

dify/web/app/components/datasets/documents/create-from-pipeline/step-indicator.tsx

@@ -0,0 +1,33 @@
+import cn from '@/utils/classnames'
+import React from 'react'
+
+export type Step = {
+  label: string
+  value: string
+}
+
+type StepIndicatorProps = {
+  currentStep: number
+  steps: Step[]
+}
+
+const StepIndicator = ({
+  currentStep,
+  steps,
+}: StepIndicatorProps) => {
+  return (
+    <div className='flex gap-x-1'>
+      {steps.map((step, index) => {
+        const isActive = index === currentStep - 1
+        return (
+          <div
+            key={step.value}
+            className={cn('h-1 w-1 rounded-lg bg-divider-solid', isActive && 'w-2 bg-state-accent-solid')}
+          />
+        )
+      })}
+    </div>
+  )
+}
+
+export default React.memo(StepIndicator)

dify/web/app/components/datasets/documents/create-from-pipeline/types.tsx

@@ -0,0 +1,5 @@
+export enum AddDocumentsStep {
+  dataSource = 'dataSource',
+  processDocuments = 'processDocuments',
+  processingDocuments = 'processingDocuments',
+}