Extract version string from text file with Powershell

Question

I need to pull the version # out of a Vim script that might look something like:

 " some header blah blah
 " Version: 0.1
 [a bunch of code]
 " Version: fake--use only the first version

The expected output is 0.1. If a Version # doesn't exist, it should return the empty string or something like that.

I'm new to Powershell, so this is what I have so far:

Get-Content somescript.vim -ErrorAction SilentlyContinue |
 Select-String '^" Version: (.*)' | 
 select -First 1 -ExpandProperty Matches |
 select -ExpandProperty Groups |
 select -Index 1 |
 select -ExpandProperty Value

It's just that it feels... kind of verbose. For comparison, here's my *nix version:

perl -ne '/^" Version: (.*)/ && do { print "1ドル\n"; exit }' somescript.vim 2>/dev/null

Or you could write a similarly concise awk script

Is there any hope for my Windows version being as concise?

Answer 1

In PowerShell everything is an object, so you can access properties (and properties of properties) using the dot operator just like you can in most object based languages.

$matches = Get-Content somescript.vim -ErrorAction SilentlyContinue |
 Select-String '^" Version: (.*)'
if ($matches)
{
 $matches[0].Matches.Groups[1].Value
}

• Michael Kropat: Ah, of course, just use an intermediate result. Hadn't thought of it, but totally obvious in retrospect.

Answer 2

I haven't found a terser approach using only built-ins, but having a little more confidence now in Powershell, I think I'd simply refactor out the group matching code:

filter Select-MatchingGroupValue($groupNum) {
 if (! $groupNum) {
 $groupNum = 0
 }
 select -InputObject $_ -ExpandProperty Matches |
 select -ExpandProperty Groups |
 select -Index $groupNum |
 select -ExpandProperty Value
}

Then you could use it like:

Get-Content .\somescript.vim -ErrorAction SilentlyContinue |
 Select-String '^" Version: (.*)' |
 select -First 1 |
 Select-MatchingGroupValue 1

In the end, you haven't saved that many lines, but refactoring out the tedious expansions of Matches, Groups, and Value makes the resulting code much clearer IMO.