Question

I often have the following situation in my PowerShell code: I have a function or property that returns a collection of objects, or $null. If you push the results into the pipeline, you also handle an element in the pipeline if $null is the only element.

Example:

$Project.Features | Foreach-Object { Write-Host "Feature name: $($_.Name)" }

If there are no features ($Project.Features returns $null), you will see a single line with “Feature name:”.

I see three ways to solve this:

if ($Project.Features -ne $null)
{
  $Project.Features | Foreach-Object { Write-Host "Feature name: $($_.Name)" }
}

or

$Project.Features | Where-Object {$_ -ne $null) | Foreach-Object { 
  Write-Host "Feature name: $($_.Name)" 
}

or

$Project.Features | Foreach-Object { 
  if ($_ -ne $null) {
    Write-Host "Feature name: $($_.Name)" }
  }
}

But actually I don’t like any of these approaches, but what do you see as the best approach?

Answer 1

I don’t think anyone likes the fact that both “foreach ($a in $null) {}” and “$null | foreach-object{}” iterate once. Unfortunately there is no other way to do it than the ways you have demonstrated. You could be pithier:

$null | ?{$_} | % { ... }

the ?{$_} is shorthand for where-object {$_ -ne $null} as $null evaluated as a boolean expression will be treated as $false

I have a filter defined in my profile like this:

filter Skip-Null { $_|?{ $_ } }

Usage:

$null | skip-null | foreach { ... }

A filter is the same as a function except the default block is process {} not end {}.

UPDATE: As of PowerShell 3.0, $null is no longer iterable as a collection. Yay!

-Oisin