Question

I want to write a cmdlet that reads multiple records from a database and puts them onto the pipeline.

I think I can do either a single WriteObject(Enumerable<rec>, true) or I can loop myself and call WriteObject multiple times.

What’s the difference between these two?

Answer 1

Here is the documentation: Cmdlet.WriteObject Method (Object, Boolean)

And here is the example:

# Writes objects one by one
function Test1
{
    [CmdletBinding()]param()
    $data | %{ $PSCmdlet.WriteObject($_) }
}

# Writes the collection and allows the core to enumerate it one level.
# Numbers of written objects is the collection item count.
function Test2
{
    [CmdletBinding()]param()
    $PSCmdlet.WriteObject($data, $true)
}

# Writes the collection as a single object.
# Numbers of written objects is 1.
function Test3
{
    [CmdletBinding()]param()
    $PSCmdlet.WriteObject($data, $false)
}

function Test
{
    (Test1).GetType().Name
    (Test2).GetType().Name
    (Test3).GetType().Name
}

$data = New-Object System.Collections.ArrayList

Write-Host "1 item"
$null = $data.Add('hello')
Test

Write-Host "2+ items"
$null = $data.Add('world')
Test

Output:

1 item
String
String
ArrayList
2+ items
Object[]
Object[]
ArrayList

Thus, calling WriteObject(item) for each item in a collection is basically the same as WriteObject(items, true); in both cases the collection itself has gone.

WriteObject(items, false) is different; it returns a reference to the collection and the caller can use that effectively depending on a scenario. For example, if a collection is a DataTable object (not unrolled set of DataRow items) then a caller can operate on DataTable members of the returned object.