如何修改Selenium类生成与Chrome检查工具一致的完整XPath

尝试基于Selenium.WebElement生成完整XPath时，生成结果与谷歌浏览器检查工具的「复制完整XPath」不一致。例如在finance.yahoo.com历史价格数据页面中，Chrome生成的路径为：

/html/body/div[2]/main/section/section/section/article/section[1]/div[2]/div[1]/section/div/section[1]/div[2]

而自定义类生成的路径为：

/html/body/div[3]/main/section[1]/section[1]/section[2]/article/section[1]/div[14]/div[1]/section[1]/div[1]/section[1]/div[2]

目标是修改SeleniumXPath类，使其生成的XPath与Chrome工具结果匹配。原类代码如下：

Option Explicit
Private pLabel As String
Private pXPath As String
Private pHasAnId As Boolean
Private pTimeStamp As Date
Public Property Let Label(aLabel As String)
  'Set in factory
  pLabel = aLabel
End Property
Public Property Get Label() As String
  Label = pLabel
End Property
Public Property Let WebElement(anElem As Selenium.WebElement)
  'Set in factory
  Call FindFullXPath(anElem)
  pTimeStamp = SNow
End Property
Public Property Get XPath() As String
  XPath = pXPath
End Property
Public Property Get hasAnId() As Boolean
  hasAnId = pHasAnId
End Property
Public Property Get IsStale() As Boolean
  If (SNow - pTimeStamp > 1) Then
    IsStale = True
  End If
End Property
Public Property Get Id() As String
  Id = XPath
End Property
Private Sub FindFullXPath(ByVal anElem As Selenium.WebElement)

  'A helpful discussion can be found in
  '  https://www.youtube.com/watch?v=lr7CFZEI2YA

  Dim anId As String
  Dim finished As Boolean
  Dim aTag As String
  Dim ancestorElem As Selenium.WebElement
  Dim anXPath As String
  Dim anElemArr As Selenium.WebElements
  Dim aCount As Long
  Dim k2 As Long

  On Error Resume Next

  'Attempt to retrieve the WebElement unique ID
  anId = anElem.Attribute("id")

  If (anId = "") Then

    'A unique ID was not found so, from the given element
    '  traverse the html tree backwards to map the full path

    Do While (Not finished)
      aTag = anElem.tagName
      If (aTag = "html") Then
        finished = True
      Else
        Set ancestorElem = anElem.FindElementByXPath("..")
        If (anXPath <> "") Then
          anXPath = "/" & anXPath
        End If
        Set anElemArr = ancestorElem.FindElementsByTag(aTag)
        aCount = anElemArr.Count
        If (aCount > 1) Then
          For k2 = 1 To aCount
            If (anElemArr.item(k2).Equals(anElem)) Then
              Exit For
            End If
          Next k2
          anXPath = aTag & "[" & k2 & "]" & anXPath
        Else
          anXPath = aTag & anXPath
        End If
        Set anElem = ancestorElem
      End If
    Loop
    pXPath = "/html/" & anXPath

  Else

    'A unique id was found, use it instead of the full
    '  XPath

    pHasAnId = True
    pXPath = anId

  End If

  Call Err.Clear

End Sub

差异原因分析 #

1. 子节点范围错误：原代码通过FindElementsByTag获取父元素下所有层级的同标签后代元素，但Chrome的完整XPath仅统计父元素的直接子节点中同标签的数量，这是索引偏差的核心原因。
2. 标签大小写不一致：Chrome生成的XPath标签全为小写，原代码直接使用tagName返回的可能是大写（如HTML、BODY），导致路径格式不匹配。
3. 元素匹配逻辑偏差：原代码遍历所有后代元素匹配目标节点，顺序可能与DOM文档的直接子节点顺序不一致。

修改后的代码 #

仅需替换原FindFullXPath方法，其余类成员保持不变：

Private Sub FindFullXPath(ByVal anElem As Selenium.WebElement)
    Dim anId As String
    Dim finished As Boolean
    Dim aTag As String
    Dim ancestorElem As Selenium.WebElement
    Dim anXPath As String
    Dim directChildren As Selenium.WebElements
    Dim childIndex As Long
    Dim currentChild As Selenium.WebElement
    
    On Error Resume Next
    
    ' 获取元素ID，优先使用标准ID XPath格式
    anId = anElem.Attribute("id")
    If anId <> "" Then
        pHasAnId = True
        pXPath = "//*[@id='" & anId & "']"
        GoTo Cleanup
    End If
    
    ' 从目标元素向上遍历DOM树，构建完整XPath
    finished = False
    Do While Not finished
        aTag = LCase(anElem.tagName) ' 统一转小写，匹配Chrome格式
        If aTag = "html" Then
            finished = True
        Else
            Set ancestorElem = anElem.FindElementByXPath("..")
            ' 获取父元素的所有直接子节点（而非所有后代）
            Set directChildren = ancestorElem.FindElementsByXPath("./child::*")
            
            ' 遍历直接子节点，找到当前元素的索引（从1开始）
            childIndex = 0
            For Each currentChild In directChildren
                childIndex = childIndex + 1
                If currentChild.Equals(anElem) Then
                    Exit For
                End If
            Next currentChild
            
            ' 统计父元素下同标签的直接子节点数量，仅当数量>1时添加索引
            Dim sameTagCount As Long
            sameTagCount = ancestorElem.FindElementsByXPath("./child::" & aTag).Count
            
            If sameTagCount > 1 Then
                anXPath = aTag & "[" & childIndex & "]/" & anXPath
            Else
                anXPath = aTag & "/" & anXPath
            End If
            
            Set anElem = ancestorElem
        End If
    Loop
    
    ' 拼接根节点，移除末尾多余的斜杠
    pXPath = "/html/" & Left(anXPath, Len(anXPath) - 1)
    
Cleanup:
    Call Err.Clear
    Set ancestorElem = Nothing
    Set directChildren = Nothing
    Set currentChild = Nothing
End Sub

修改说明 #

1. 直接子节点统计：使用./child::*获取父元素的直接子节点，确保仅统计当前层级的同标签元素，和Chrome的逻辑一致。
2. 标签小写转换：通过LCase(anElem.tagName)将标签名统一转为小写，匹配Chrome的XPath格式。
3. 索引逻辑优化：先找到当前元素在父节点直接子节点中的位置，再统计同标签直接子节点的数量，仅当数量大于1时添加索引，和Chrome的省略规则一致。
4. ID处理优化：将原直接返回ID字符串的逻辑，修改为标准的ID XPath格式//*[@id='xxx']，更符合XPath规范。

内容的提问来源于stack exchange，提问作者KC2