In the previous tip we split a chunk of multi-line text into individual lines and removed any empty lines.

However, when a line isn’t really empty but contains whitespace (spaces or tabulators), it is still returned:

# $data is a single string and contains blank lines
$data = @'

Server1

   
Server2
Cluster4


'@

$data = $data.Trim()

# split in single lines and remove empty lines
$regex = '[\r\n]{1,}'
$array = $data -split $regex

$array.Count

$c = 0
Foreach ($_ in $array)
{
    '{0:d2} {1}' -f $c, $_
    $c++
}

Here, we have added a few spaces to the line right above “Server2” (which you obviously can’t see in the listing). This is the result:

 
00 Server1
01    
02 Server2
03 Cluster4  
 

Since we are splitting at any number of CR and LF characters, a space would break that pattern.

Rather than turning the regular expression in an even more complex beast, for such things you may want to append a simple Where-Object to do the fine polishing:

# $data is a single string and contains blank lines
$data = @'

Server1

   
Server2
Cluster4


'@

$data = $data.Trim()

# split in single lines and remove empty lines
$regex = '[\r\n]{1,}'
$array = $data -split $regex | 
  Where-Object { [string]::IsNullOrWhiteSpace($_) -eq $false }

$array.Count

$c = 0
Foreach ($_ in $array)
{
    '{0:d2} {1}' -f $c, $_
    $c++
}

[string]::IsNullOrEmpty() identifies the situation we are after, so lines that qualify are removed by Where-Object. The result is what is needed:

 
00 Server1
01 Server2
02 Cluster4  
 




Twitter This Tip! ReTweet this Tip!

Anonymous