Cannot convert value "version="1.0" encoding="utf-8"" to type "Sy-CodePudding

I have a sample XML file test.xml that looks like this:

<?xml version="1.0" encoding="utf-8"?>
<Response>
  <Update>
    <foo>1.1.1</foo>
    <bar>12345</bar>
  </Update>
</Response>

I'm trying to use PowerShell to get the value of foo. I can download and save the file into text.xml and then access foo. This works:

Invoke-RestMethod $myurl | Out-File test.xml
[xml]$abc = (Get-Content test.xml) 
$abc.Response.Update.foo

However, I receive errors when I attempt to save the XML content into the $abc variable like this:

[xml]$abc = (Invoke-RestMethod $myurl)

Error:

MetadataError: Cannot convert value "<?xml version="1.0" encoding="utf-8"?>
<Response>
  <Update>
[...snip...]
  </Update>
</Response>" to type "System.Xml.XmlDocument". Error: "The specified node cannot be inserted as the valid child of this node, because the specified node is the wrong type."

What am I doing wrong here?

Update #1

This produces no output and no errors:

$abc = Invoke-RestMethod $myurl
$abc.Response
$abc.Response.Update           
$abc.Response.Update.foo

Changing to Invoke-WebRequest:

[xml]$abc = Invoke-WebRequest $myurl

Error is the same, but now there is three question marks in the beginning. So it's possible that the problem is somehow related to the file's encoding.

MetadataError: Cannot convert value "???<?xml version="1.0" encoding="utf-8"?>
<Response>
  <Update>
[...snip...]
  </Update>
</Response>" to type "System.Xml.XmlDocument". Error: "The specified node cannot be inserted as the valid child of this node, because the specified node is the wrong type."

CodePudding user response：

Your symptom suggests that there are three unrecognized characters preceding the XML text returned by your web service.

While three characters could generally indicate what constitutes a mis-decoded UTF-8 (pseudo) BOM,
- (a) such a BOM is usually only present in files on disk, not in web-service responses
- (b) even if it were mis-decoded as the default encoding used by the web cmdlets up to PowerShell 7.2.x, ISO-8859-1, it would print as ï»¿, not as ???

Pragmatically speaking, assuming that the remaining text is properly decoded, you can simply skip the three characters as follows:

[xml] (Invoke-WebRequest $myUrl).Content.Substring(3)

If that doesn't help, use a text editor to examine the downloaded file and determine its de-facto encoding, then try to decode the raw bytes as a string with that encoding, as shown in this answer.

CodePudding user response：

Using Xml Linq :

using assembly System.Xml.Linq 

$xml = @"
<?xml version="1.0" encoding="utf-8"?>
<Response>
  <Update>
    <foo>1.1.1</foo>
    <bar>12345</bar>
  </Update>
</Response>
"@
$doc = [System.Xml.Linq.XDocument]::Parse($xml)
$update = $doc.Descendants("Update").Foreach([System.Xml.Linq.XElement])
$foo = $update.Element("foo").Value
$bar = $update.Element("bar").Value
Write-Host "foo = " $foo "bar = " $bar