Power Automate中XPath提取XML请求ID纯数值的问题求助

Here are two reliable methods to get all pure numeric IDs from your XML's request nodes:

Method 1: Use XPath + String Manipulation #

1. First, retrieve all attribute nodes with your existing XPath expression:

   xpath(xml(outputs('Compose')), '//request/@id')
   

   This returns an array of attribute nodes (each formatted like "id="31998"").
2. Add a Select action to process this array:
   • In the "From" field, select the output from the XPath step.
   • In the "Map" field, use this expression to extract the numeric ID:

     substringBefore(substringAfter(string(item()), '="'), '"')
     

   This works by:
   • Converting the attribute node to a string (string(item()))
   • Grabbing everything after '="'' (substringAfter(...))
   • Taking everything before the closing quote (substringBefore(...))

Method 2: Direct XML Object Access (Simpler) #

If your XML has a predictable structure (e.g., a root node containing multiple request nodes), you can skip XPath entirely:

1. Parse the XML into an object using:

   xml(outputs('Compose'))
   

2. Access the array of request nodes (adjust the path to match your XML root):
   For example, if your XML starts with <requests>...</requests>, use:

   outputs('Parse_XML')?['requests']?['request']
   

3. Add a Select action to extract the @id property from each request object:
   • In "From", select the request array from the previous step.
   • In "Map", use:

     item()?['@id']
     

   This directly gives you the pure numeric ID values as strings (convert to numbers with int(item()?['@id']) if needed).

Why Your Previous Attempts Failed #

• string(//request/@id) only returns the first attribute value because XPath 1.0 (used by Power Automate) converts a node-set to string by taking the first node.
• //request/@id/text() returns null because attributes don't have text nodes—their value is the attribute itself.

内容的提问来源于stack exchange，提问作者Fixermister