Slow array operations after DB resultset - arrays

Question

I need help optimizing my PowerShell script.
$sorted = #()
$firsttime = 0
$j = 0
$zaehler = $results.Count-1
for ($i=0; $i -le $results.Count-1; $i++) {
$j = $i+1
while ($results.GUID[$i] -eq $results.GUID[$j]) {
$klassen = ""
$rec = $results | where {$_.GUID -eq $results.GUID[$i]}
if ($firsttime -eq 0 -or !$sorted.GUID.contains($rec[0].GUID)) {
$firsttime = 1
foreach ($item in $rec.Klasse) {
if ($klassen -eq "") {
$klassen += $item
} else {
if (!$klassen.Contains($item)) {
$klassen += "," + $item
}
}
}
$rec[0].Klasse = $klassen
$sorted += $rec[0]
}
$j = $j+1
}
Write-Host ($i/$zaehler).ToString("P") "von Schule" $schule
}
if (!$sorted) {
$results
} else {
$sorted
}
Basically in my resultset ($results) I got duplicate lines of teachers and the only difference is the class ("Klasse/Klassen") they are teaching at.
To minimize the output I am checking if the first GUID is the same as the second and then the script appends the second class to the first one. So the $sorted array has just one line per teacher with a comma-seperated string which shows all classes.
Sample line of $results:
#{
GUID={1234567-1234-1234-1234-1234567};
userrole=teacher;
Vorname=Max;
Nachname=Mustermann;
Geburtstag=01.01.2000;
Klasse=9A;
Schule=123456
}
#{
GUID={1234567-1234-1234-1234-1234567};
userrole=teacher;
Vorname=Max;
Nachname=Mustermann;
Geburtstag=01.01.2000;
Klasse=9B;
Schule=123456
}
Sample line of $sorted[0]:
#{
GUID={1234567-1234-1234-1234-1234567};
userrole=teacher;
Vorname=Max;
Nachname=Mustermann;
Geburtstag=01.01.2000;
Klasse=9A,9B,9C,5A;
Schule=123456
}
The sorting process (check if contains, foreach $item, add to $klassen) is pretty slow.
I would be very grateful for any kind of ideas how to optimize the script.

Answer 1

Maybe something like this would work:
$results | ForEach-Object {
New-Object -Type PSObject -Property $_
} | Group-Object GUID | ForEach-Object {
$teacher = $_.Group
$teacher | Select-Object -First 1 -Exclude Klasse -Property *, #{n='Klasse';e={
$teacher | Select-Object -Expand Klasse
}}
}
Convert your hashtables into custom objects, group them by GUID, then replace the original "Klasse" property with a new one containing an array of the values from all objects in that group, and select the first result.
Should $results already be a list of objects you can skip the first ForEach-Object:
$results | Group-Object GUID | ForEach-Object {
$teacher = $_.Group
$teacher | Select-Object -First 1 -Exclude Klasse -Property *, #{n='Klasse';e={
$teacher | Select-Object -Expand Klasse
}}
}