More efficient way of using Import-CSV in powershell - powershell

Question

I am trying to figure out a more efficient way of using Import-CSV (powershell) to place values into an array of csv files. The problem is that some of these files have several hundred thousand lines and running this script in conjunction with other lines of code is what appears to be a big bottle neck. Do you guys have any suggestions of how to make this code more efficient and faster?
foreach($csv in $csvfiles)
{
$csvname = $csv.name;
$paygroup = $csvname.substring(4,3);
$batch = $csvname.substring(14,4);
write-host "Writing $csvname";
$csvimportdata = Import-CSV $CurrentPath"\$csvname";
foreach($record in $csvimportdata)
{
$record.chartfield1 = $paygroup;
$record.chartfield2 = $batch;
$record.chartfield3 = $record.line_descr.substring(0,6);
}
$csvimportdata | Export-CSV $CurrentPath"\$csvname" -NoTypeInformation
};

Answer 1

If your CSVs are large then loading into memory is probably not a good idea. How about something like this:
foreach($csv in $csvfiles)
{
$csvname = $csv.name
$paygroup = $csvname.substring(4,3)
$batch = $csvname.substring(14,4)
write-host "Writing $csvname"
Get-Content $CurrentPath"\$csvname" -Readcount 1 | % {
# Regex below assumes a three column CSV
$_ -replace '^([^,]+,[^,]+,[^,]{6}).*$', '$1'
} | Set-Content $CurrentPath"\$csvname"
}