the code looks good for me,

try to show a minimal working (and intended) example reading text from DATA and having @terms populated.

personally I would code it with something like

$clause= join "|", @terms;
print $1 while m/UP($clause)LB/g;
[download]

(untested)

Thank you for the reply Rolf! I'm guessing I should've put all the code instead of a little snippet, as it isn't very long.

use strict;
use warnings;
use File::Slurp;

my $countfile = 1;

open (IN, 'C:\terms.txt');
my @terms = <IN>;
close IN;

open (OUT,'>>C:\result.txt');

while ($countfile < 10000) {
    
    my $text = read_file('C:\file' . $countfile . '.xml') or die "Can'
+t open file!";
            
        if($text =~ /upperboundary(.+?)lowerboundary/s){

        if(grep {$_ eq $1} @terms) {
            
        print OUT "$1\t";
        
        }
    }
    
    close IN;
    $countfile++;
    open (IN, 'C:\file' . $countfile . '.xml') or die "End of files!";
}

close OUT;
[download]

I can see the array populated if I print it, but there must be some other thing that's destroying the output at the middle...

Tip #3 from the Basic debugging checklist: chomp

open (IN, 'C:\terms.txt');
my @terms = <IN>;
close IN;
chomp @terms;
[download]

Ok, I was overlooking something. I was comparing a whole set of free text to an array of terms, no wonder why it didn't return anything.

Can the solution be splitting that free text into an array, and then comparing that to the terms I already have, printing the matches?

my @adv = split /\s+/, $1;
[download]

How would I compare if each term in @adv matches one on @terms, and then print it if it exists?

@terms = qw/.A. .C./;

my $clause= join "|", map quotemeta ,@terms;
while (my $line = <DATA>){
   print $1 while $line =~m/UP($clause)LB/g;
}

__DATA__
  UP.A.LB
  UP.B.LB
  UP.C.LB
[download]

I noticed you used the s modifier in your match and that you seem to be trying to match across the whole file.

This example might not be the most efficient way but is a start.

use warnings;
use strict;

my @terms = qw( Maple Oak walnut pine juniper );

my $text;
{
    #slurp the whole file
    local $/;
    $text = <DATA>;
}
if ( $text =~ /upperboundary(.+)lowerboundary/s ) {
    $text = $1;
    print $text, "\n\n";
}
else {
    die "no match";
}
foreach my $term (@terms) {
    if ( $text =~ /($term)/is ) {
        print "$1\n";
    }
}

__DATA__
upperboundary
The book Hiking the Red, A Complete Trail Guide To Kentucky's Red Rive
+r Gorge 
made it a point to describe the habitat and the diverse species of tre
+es that 
one is likely to encounter at the Red. According to the authors, the s
+pecies 
of trees found in the Daniel Boone National Forest includes beech, sug
+ar 
maples, white pines, hemlock, several types of oak, and hickory. These
+ trees 
provide habitat for an estimated 67 different species of reptiles and 
amphibians, 46 species of mammals, and 100 species of birds. 
lowerboundary
[download]

Edit: The text is from http://www.redrivergorge.com/dbnf.html.

Hi Lotus,

I used the s modifier because i needed the .+ to match new lines too. Is that the way to go?

Yes, I used the s modifier in both of the regex's in my example. I changed my node a little to make clear I was saying my example might not be the most efficient. I wasn't referring to the use of the s modifier.