Perl Regex Extract first two section of windows path - regex

Question

I want to write a method to extract first two sections of windows path in Perl.
For example,
'D:\git_root_tfs\WorkStation\Projects\InterACT\Tools\server-rule-checker'
Extract as:
'D:\git_root_tfs\WorkStation'
sub Split_Location_as_VMPath {
my $location = shift;
# ^([d-z]:\\.+?\\.+?)\\
# ^(?:\\.*\\.*)\\
if($location ~~ m/^(?:\\.*\\.*)\\/){ # the path drive start from D to E;
# print "VMPath=$1\n";
# push #$vmPathList, $1;
return Convert_to_Lowercase($1);
}
return "Invalid Path $location";
}
How to write the regex?
Test case:
{
my $item = Split_Location_as_VMPath('D:\VM\ia7-BGCDev8.1\test.vhd');
my $expected = Convert_to_Lowercase('D:\VM\ia7-BGCDev8.1');
ok( $item eq $expected, "Test Split_Location_as_VMPath=$item");
$item = Split_Location_as_VMPath('E:\Hyper-V-2\ia-int-7.1Beta\test.vhd');
$expected = Convert_to_Lowercase('E:\Hyper-V-2\ia-int-7.1Beta');
ok( $item eq $expected, "Test Split_Location_as_VMPath=$item");
$item = Split_Location_as_VMPath('D:\VM\ia7-int-7.1\test.vhd');
$expected = Convert_to_Lowercase('D:\VM\ia7-int-7.1');
ok( $item eq $expected, "Test Split_Location_as_VMPath=$item");
$item = Split_Location_as_VMPath('D:\VM\ia7-int-8.1B153\test.vhd');
$expected = Convert_to_Lowercase('D:\VM\ia7-int-8.1B153');
ok( $item eq $expected, "Test Split_Location_as_VMPath=$item");
$item = Split_Location_as_VMPath('D:\Hyper-v\IA5-SDE-WIN2K3(Feng Tong)\test.vhd');
$expected = Convert_to_Lowercase('D:\Hyper-v\IA5-SDE-WIN2K3(Feng Tong)');
ok( $item eq $expected, "Test Split_Location_as_VMPath=$item");
$item = Split_Location_as_VMPath('D:\git_root_tfs\WorkStation\Projects\InterACT\Tools\server-rule-checker');
$expected = Convert_to_Lowercase('D:\git_root_tfs\WorkStation');
ok( $item eq $expected, "Test Split_Location_as_VMPath=$item");
}

Answer 1

Don't use a regex for file processing.
Instead use a module like File::Spec or Path::Tiny.
use strict;
use warnings;
use File::Spec;
while (<DATA>) {
my ($vol, $dir, $file) = File::Spec->splitpath($_);
my #dirs = File::Spec->splitdir($dir);
#dirs = #dirs[0..2] if #dirs > 3;
$dir = File::Spec->catdir(#dirs);
my $path = File::Spec->catpath($vol, $dir);
print "$path\n";
}
__DATA__
D:\VM\ia7-BGCDev8.1\test.vhd
E:\Hyper-V-2\ia-int-7.1Beta\test.vhd
D:\VM\ia7-int-7.1\test.vhd
D:\VM\ia7-int-8.1B153\test.vhd
D:\Hyper-v\IA5-SDE-WIN2K3(Feng Tong)\test.vhd
D:\git_root_tfs\WorkStation\Projects\InterACT\Tools\server-rule-checker
Outputs:
D:\VM\ia7-BGCDev8.1
E:\Hyper-V-2\ia-int-7.1Beta
D:\VM\ia7-int-7.1
D:\VM\ia7-int-8.1B153
D:\Hyper-v\IA5-SDE-WIN2K3(Feng Tong)
D:\git_root_tfs\WorkStation

Answer 2

Correct regex is ^([d-z]:\.+?\.+?)\.
sub Split_Location_as_VMPath {
my $location = shift;
# ^([d-z]:\\.+?\\.+?)\\
# ^(?:\\.*\\.*)\\
if($location ~~ m/^([D-Z]:\\.+?\\.+?)\\/){ # the path drive start from D to E;
# print "VMPath=$1\n";
# push #$vmPathList, $1;
return Convert_to_Lowercase($1);
}
return "Invalid Path $location";
}

Answer 3

Using regex in this context is an interesting homework for students. Outside school, you should use the standard modules dedicated for this task:
use File::Spec;
sub Split_Location_as_VMPath {
my $location = shift;
my ($volume, $directories, $file) = File::Spec->splitpath($location);
my #dirs = File::Spec->splitdir($directories);
return "Invalid Path $location" unless #dirs > 2;
return lc File::Spec->catpath($volume, File::Spec->catdir(#dirs[0..2]));
}