How to properly use PowerShell $args[] - foreach

Question

Sometimes a PowerShell script gets called with arguments containing spaces and enclosed by double-quotes,
to ensure the string is interpreted as one single argument. A problem arises when the argument is empty and the double-quotes are passed containing nothing. PS sees the command line as non-empty, and tries to form an argument value from it. Except the double-quotes are stripped away, leaving an empty string as an argument value. Of course PS cannot actually use the empty value, so we want to validate command line arguments to disallow empty strings.
See the calling Cmd script, and the PS source, both commented. Tests 4 and 5 show the benefit of the function that scans the $args[] array and returns the first valid/non-empty argument.
Question: why can the function not be coded with just one ForEach loop?
Args.CMD:
#Echo Off
CD /D "%~dp0" & Rem Change to the folder where this script lives
CLS
Echo.
Rem PS script name is the same as this Cmd script
Set Script=%~n0.PS1
Rem Parameters of each call here are TestNumber, ExpectedExitCode, Argument1, Argument2
Rem Test 0: Try with zero arguments; expected exit code is 1
Call :TryMe 0 1
Rem Test 1: Try with non-useful arguments; expected exit code is 2
Call :TryMe 1 2 "" ""
Rem Test 2: Try with valid arguments, single words; expected exit code is 0
Call :TryMe 2 0 ArgOne ArgTwo
Rem Test 3: Try with valid arguments, multiple words; expected exit code is 0
Call :TryMe 3 0 "Arg One" "Arg Two"
Rem Test 4: Try with mixed arguments, single words; expected exit code is 0
Call :TryMe 4 0 "" ArgTwo
Rem Test 5: Try with mixed arguments, multiple words; expected exit code is 0
Call :TryMe 5 0 "" "Arg Two"
Echo All Done.
Pause
GoTo :EOF
:TryMe
Set TstNbr=%1
Set ExpCod=%2
Set Args12=%3 %4
Echo Test %TstNbr% of 5: PowerShell.EXE -F %Script% %Args12%
PowerShell.EXE -F %Script% %Args12%
Echo ErrorLevel is [%ErrorLevel%], expected was [%ExpCod%]
Pause
Echo.
GoTo :EOF
Args.ps1:
Set-StrictMode -Version 3.0 # disallow implicit variables
function ArgNonEmtFstGet( $arrArgs ) { # get the first non-empty argument
ForEach( $strArgs In $arrArgs ) { # <=== why is this double-indirection needed?
ForEach( $strArg In $strArgs ) { # scan all argument strings
$intArgLen = $strArg.Length # get length of current argument-string
If( $intArgLen -gt 0 ) { Return $strArg } # first time length is greater than zero, quit scanning and pass that argument back to caller
} # done scanning
} # <=== why is this double-indirection needed?
"" # when we finished scanning and did not return early, return a non-crashing value
} # ArgNonEmtFstGet
# Step 0: show we are alive
"Arg count: " + $args.Length # emit how many strings exist that *could be* arguments
# Step 1: look for any argument at all
If( $args.Length -eq 0 ) { # when there are zero strings that look like arguments
Write-Output "Putative argument required." # emit error message for 'no strings found'
Exit( 1 ) # abort with exit code
} # done with no strings found that *could be* arguments
"First putative argument: [" + $args[0] + "]" # emit the first argument-looking string
# Step 2: look for an argument that is actually meaningful
$strArg = ArgNonEmtFstGet( $args ) # get the first non-empty argument; that is our input
If( $strArg -eq "" ) { # when it is not a usable argument
Write-Output "Non-empty argument required." # emit error message for 'no real arguments found'
Exit( 2 ) # abort with exit code
} # done with no usable arguments found
"First non-empty argument: [$strArg]" # emit the first valid argument; this is our input

Answer 1

I can't reproduce your issue with your nested foreach-statement.
When I remove the inner loop I get the same result.
But I would suggest that you should avoid using the $args variable, because it accepts any input and doesn't allow named binding (only positional).
Just declare a param block in your PowerShell Script like below, then you don't need to loop over your arguments and additionally, you have named parameters you can use in your batch script. (Positional binding is still possible).
Sample script with param block and "ValidateScript" attribute, which validates the input to avoid null, empty or whitespace input.
param(
# first mandatory parameter. Default Position = 1 (if using positional binding)
[Parameter(Mandatory=$true)]
[ValidateScript(
{
if([String]::IsNullOrWhiteSpace($_)) {
throw [System.ArgumentException]'Your string is null, empty or whitespace!'
} else {
$true
}
}
)]
[string]
$varA,
# second mandatory parameter. Default Position = 2 (if using positional binding)
[Parameter(Mandatory=$true)]
[ValidateScript(
{
if([String]::IsNullOrWhiteSpace($_)) {
throw [System.ArgumentException]'Your string is null, empty or whitespace!'
} else {
$true
}
}
)]
[string]
$varB,
# You can just use this parameter if you want your arguments being absolutely dynamic
[Parameter(ValueFromRemainingArguments)]
[ValidateScript(
{
if([String]::IsNullOrWhiteSpace($_)) {
throw [System.ArgumentException]'Your string is null, empty or whitespace!'
} else {
$true
}
}
)]
[string[]]
$remainingArgs
)
# create ordered dictionary of mandatory arguments used for returning object
$properties= [ordered]#{
'Arg0' = "'$varA'"
'Arg1' = "'$varB'"
}
# add any optional remaining argument (if any)
$index = 1 # start index
foreach ($remainingArg in $remainingArgs) {
$index++
$properties.Add("Arg$index", "'$remainingArg'")
}
# return as object
[pscustomobject]$properties
As mentioned in the script, you can also just use the third parameter if you want your script being absolutely dynamic in terms of parameters.
More infos about attribute "ValidateScript": https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_functions_advanced_parameters?view=powershell-5.1#validatescript-validation-attribute
PS:
Sample command when using named parameters in batch:
PowerShell.EXE -F %Script% -varA "%3" -varB "%4"