Understanding $args in PowerShell
-
Ok, still finding this a bit difficult. I thought the $args was created automatically so is it necessary to include the $args[0] as in the example?
-
@chutestrate said:
Ok, still finding this a bit difficult. I thought the $args was created automatically so is it necessary to include the $args[0] as in the example?
$args is created when the user, not the programmer, passes arguments to the script. In my example the argument was my name, Scott, which I passed in the line when I ran the program.
To use the argument that I passed in, the program has to reference it. The only way to reference it in the script is to use $args[0] which is a reference to the first argument passed. If you don't use $args[0] the argument will be ignored as nothing in the script looks for it.
-
@scottalanmiller said:
To use the argument that I passed in, the program has to reference it. The only way to reference it in the script is to use $args[0] which is a reference to the first argument passed. If you don't use $args[0] the argument will be ignored as nothing in the script looks for it.
That's not accurate, actually. PowerShell very much will try to display it as best it can. So if you do:
.\test.ps1 Scott
And your script is
Write-Host "Hi " $Args
It'll work just fine. If you run it: .\test.ps1 Scott,Martin you'll get:
Hi Scott Martin
Notice no comma, PowerShell is just squashing the array together. Change the script to:
Write-Host "Hi " $Args[1] " and " $Args[0]"
And run it: .\test.ps1 Scott,Martin and you'll get:
Hi Martin and Scott
-
Let's try another program with arguments. This time we will make the program just a little bit longer:
foreach ($argument in $args) { write-host $argument }Run it like this (assuming that you also save it as argtest.ps1)...
.\argtest.ps1 this is a testAnd it prints out...
this is a test -
Well, I"m just not going to get this easily. I'm getting singular concepts but it's not flowing together cohesively.
-
@Martin9700 ah yes, you can reference the array itself as a whole rather than the individual elements. Makes sense.
-
@chutestrate said:
Well, I"m just not going to get this easily. I'm getting singular concepts but it's not flowing together cohesively.
It's important to remember though, in PowerShell no one uses $Args. We use named Parameters, just so we can avoid the confusion of $Args. Also makes the script easier to read:
Param( [string]$First, [string]$Second ) Write "Hi $First and $Second" -
@chutestrate said:
Well, I"m just not going to get this easily. I'm getting singular concepts but it's not flowing together cohesively.
Try actually running the script that I just made where it takes an unlimited number of arguments, stores them in the $args array and prints them out one at a time. Play with that for a minute actually running it with different arguments passed it. I think that it will help once you see it a few times.
-
@Martin9700 that makes learning it so much harder though. If a single array is confusing, doing that makes it even more abstract.
-
which script the args or param
-
@scottalanmiller said:
@Martin9700 that makes learning it so much harder though. If a single array is confusing, doing that makes it even more abstract.
Then using named parameters? I don't see how that's more confusing then some nameless value you HOPE is right. In this example you have two ways you could run the script:
.\test.ps1 Scott Martinor
.\test.ps1 -First Scott -Second MartinNotice in the second example the script is now resembling a proper cmdlet.
-
And once you get into named parameters, you can start using the parameter decorators:
Param ( [ValidateSet("Martin","Scott")] [string]$First, [ValidateSet("Martin","Chutestrate")] [string]$Second ) Write-Host "Hi $First and $Second!"Try .\test.ps1 -First Rob -Second Scott
-
@chutestrate said:
which script the args or param
Using param to abstract args is going to make learning what an array is less simple.
-
@Martin9700 said:
@scottalanmiller said:
@Martin9700 that makes learning it so much harder though. If a single array is confusing, doing that makes it even more abstract.
Then using named parameters? I don't see how that's more confusing then some nameless value you HOPE is right. In this example you have two ways you could run the script:
.\test.ps1 Scott Martinor
.\test.ps1 -First Scott -Second MartinNotice in the second example the script is now resembling a proper cmdlet.
That's much more complex because you are using a more complex structure where you now have to have names. It is easier to use when you are dealing with arbitrary inputs, but that isn't the goal here. John is trying to understand what an argument is and what an array is.
-
@scottalanmiller said:
@chutestrate said:
which script the args or param
Using param to abstract args is going to make learning what an array is less simple.
OK, now I see where you're going with this. I suppose that's an argument, but in my view you should know what an array is and how to manipulate it before you start messing with $Args. Otherwise your picking a tough path to walk. Besides, learning best practices early instead of breaking bad habits later.
-
Sorry martin9700 I'm not even close to understanding what you are trying demonstrate.
-
Here is a slightly more complex example but it gives more illustrative output.
for ($i=0; $i -lt $args.length; $i++) { 'This is $args[' + $i + "], which is: " + $args[$i] }Save and run with any number of arguments that you want.
> .\argsdemo.ps1 I can put in a lot of arguments and it works just fine for a demo. This is $args[0], which is: I This is $args[1], which is: can This is $args[2], which is: put This is $args[3], which is: in This is $args[4], which is: a This is $args[5], which is: lot This is $args[6], which is: of This is $args[7], which is: arguments This is $args[8], which is: and This is $args[9], which is: it This is $args[10], which is: works This is $args[11], which is: just This is $args[12], which is: fine This is $args[13], which is: for This is $args[14], which is: a This is $args[15], which is: demo. -
@chutestrate said:
Sorry martin9700 I'm not even close to understanding what you are trying demonstrate.
Martin is correct, that trying to tackle $args when you don't have a solid understanding of everyday arrays first is going to be confusing.
-
@chutestrate said:
Sorry martin9700 I'm not even close to understanding what you are trying demonstrate.
LOL, which is Scott's point.
I'm not a fan of using $Args because I always preach (and you can check the PowerShell forums over at Spiceworks to back this up) that I always write scripts for the next guy and using $Args in code make it very hard to decipher what the original writer is trying to do. Named parameters are self documenting (assuming you are using meaningful variable names). But to Scott's point, they are a more advanced technique.
-
Oh, I have an idea.
Think of my way of doing $args as being like MadLibs. But you just give a string of words and they get put in in order. Your program or script is like a MadLibs where you feed it the list of words and it says funny things. In fact, making an actual MadLibs script is a common and fun scripting project to do for exactly this reasons.
Now Martin's way is more like MadLibs where you know some basics like name, age, color, size, etc. and you put in words associated with those things so that you get a simpler matching of words to where they are used.
The approach that I am demonstrating is just taking things in order one after another. His is to identify what they are and label them so that they are more useful. Both are important to understanding, but one after another. In the real world of script writing his approach is way, way more common and practical. Not every single time, but the vast majority. But understanding my way first is important because you need to know what is happening.
