How to remove all XML children nodes but not attributes in Powershell - powershell

Question

From this :
<products id="col">
<product code="id1" />
<product code="id2" />
</products>
I want to get this :
<products id="col">
</products>
If I use RemoveAll() it will remove products id attribute also so I will get this
<products>
</products>
That's not what I want. What method could I use to keep my id attribute in products node ?

Answer 1

Instead of calling RemoveAll(), you need to find the nodes you want to remove and then remove each of them manually, with RemoveChild():
$XmlDocument = [xml]#'
<products id="col">
<product code="id1" />
<product code="id2" />
</products>
'#
$ChildNodes = $XmlDocument.SelectNodes('//product')
foreach($Child in $ChildNodes){
[void]$Child.ParentNode.RemoveChild($Child)
}
$XmlDocument.Save("C:\path\to\doc.xml")
If you don't know the name of the immediate child nodes, but only the parent node name, you can use XPath to select them anyways: '//ParentNodeName/child::*'
$ChildNodes = $XmlDocument.SelectNodes('//products/child::*')

Answer 2

My answer doesn't rely on XPath to accomplish this. The script assumes $XmlDocument is a variable containing an XmlDocument instance. The script also assumes that only <product> elements live inside <products>.
if ($XmlDocument.products -is [Xml.XmlElement]) {
$XmlDocument.products.product | %{ $_.ParentNode.RemoveChild($_) | Out-Null }
}
This code also works if <products> does not contain any child elements (i.e., it is a text element).