Powershell script to get a part of XML node's content

To achieve what you have asked. I think we have two main steps here:

1. Extract the content of XML nodes.
2. Trim the content and take what you need only.

I'm not really familiar with your existing scripts so I will explain all two steps here. The first step is optional to you.

Extract content of XML nodes

My example XML document:

<Corpus>
    <CorpusLink>../Metadata/A_short_autobiography_of_Herculino_Alves.xml</CorpusLink>
    <CorpusLink>../Metadata/Wordlist_and_phrases_-_modifiers.xml</CorpusLink>
    <CorpusLink>../desano-silva-0151/Metadata/Wordlist_fruits_and_cultural_items.xml</CorpusLink>
    <CorpusLink>../desano-silva-0151/Metadata/The_Turtle_and_the_Deer.xml</CorpusLink>
    <CorpusLink>../desano-silva-0151/Metadata/Wordlist_and_phrases_parts_of_a_tree.xml</CorpusLink>
    <CorpusLink>../desano-silva-0151/Metadata/Wordlist_and_phrases_.xml</CorpusLink>
</Corpus>

PS script to get the content:

[xml] $XmlDocument = Get-Content D:\Path_To_Your_File
$XmlDocument.Corpus.CorpusLink # Content of the nodes you need

Trim the content

There are many methods but I think I will go with regex. Simply loop through all the contents and run the regex.

$XmlDocument2.Corpus.CorpusLink | Foreach-Object {
    if ($_ -match "\.\.\/.*?\/") {
        $Matches.Values
    }    
}

About the regex, it matches any character except for line terminators between ..\ and /:

\.\.  # Escape for 2 dots `..`
\/    # Escapefor slash `/`
.*?   # Takes any character except for line terminators in between other listed characters (above and below)
\/    # Escape for slash `/`

I imply the structure of these strings is stable like that, hence the regex.