If I understand your question properly...

$string = "il asdfasdfasdf";
$string =~ s/([\S]{2})//;
$tag = "<$1>";

$string = $tag . "$string ". $tag
[download]

Would do what you're asking for. I'm sure a true perl master could condense that into 1-2 lines, but I'm just a little perl footsoldier right now...

Peace,
LassiLantar

I'm sure a true perl master could condense that into 1-2 lines.

But a true Perl master would go for clarity over conciseness unless the circumstances dictate otherwise.

Cheers,
Ovid

New address of my CGI Course.

True, true. Again, I am outclassed =)

Peace,
LassiLantar

I'm pretty sure you can make this code shorter if you really want to, but I am curious why you chose to substitute all lines. That, to me, looks like a lot of useless hassle ;)

Anyways, for the OP, my €0,02:

while(<DATA>) { 
   print "<$1>$1$2<$1>" if $_ =~ m|(\w{2})(.*)|;
}
__DATA__
il yadayadayada
df yadayadayada
[download]

--
b10m

All code is usually tested, but rarely trusted.

<snappy comeback> Well, he said each line had a tag on the beginning of it, so I figured it would be extra to deal with checking whether it did. </snappy comeback>

<real excuse> Didn't think of it =) </real excuse>

Peace,
LA

I must not be a true perl master, because I hacked on your program and it got BIGGER!

#!/usr/bin/perl

# you have to use strict and warnings unless you
# have a really good reason not to.
use strict;
use warnings;

my $string = "il asdfasdfasdf";
my $tag = "";

# use matching here instead of substitution
# all of the string should appear in the output
# also, don't need square brackets in match
if ($string =~ m/(\S{2})/)
{
    $tag = "<$1>";
}

# you don't need to concatenate, just interpolate the lot
$string = "$tag $string $tag";
print "string = $string\n";

__END__
[download]

I must not be a true perl master, because I hacked on your program and it got BIGGER!

Gwuahaha! I am superior! (read: I am too lazy to write in use strict/use warnings on PM). I agree with you, use strict and warnings are totally necessary. I'm so lazy I even sometimes try to circumvent use strict by redeclaring my variables in random places, but really they're improving the way I write code. (As is sparring with the monks).

Peace,
LassiLantar

I'm working on tagging a large linguistic corpus

What I need to do is add a tag around each line (<il> or <df> in the above cases) where the contents of the tag match the two character string at the head of each line:

<il> il yadayada <il>

There are two things you should consider (maybe ask others in your group/research community to get their suggestions):

1. The tags you add should be paired like this:

   <tag> text content ... </tag>
   [download]

   Note the slash character in the second tag that marks the end of the region -- that's required.

2. If the initial "token" on each is really a classifier (i.e. an annotation that someone has added to the corpus data, rather than being part of the original spoken or written corpus content), then the XML tags ought to replace the classifier, rather than simply being placed around it.

On the second point, I could see wanting to leave the 2-letter code in the line, just to make sure you put the tags in the right way, but there are better ways to validate your process.

If I'm guessing right about what you really should be doing, your regex should just put angle brackets around the initial 2-character token, then make a copy of it at the end of the line with a slash added as needed. Something like this:

  s{^(\w{2})(.*)}{<$1>$2 </$1>};
[download]

(P.S.: Welcome to the Monastery!)

If you just want this on the command line to read from one file and write to STDOUT (great for seeing that it works):

#!/usr/bin/perl

use strict;
use warnings;

# read in the data line by line
while (my $line = <DATA>)
{
    # chomp off the newline
    chomp $line;
    # see if we have a match of two letters at the start
    # of the line
    if ($line =~ m|^(\w{2})|)
    {
        # if it matched, add tags
        my $tag = $1;
        print "<$tag> $line <$tag>\n";
    }
    else
    {
        # if it didn't match, just print the line
        print "$line\n";
    }
}

__DATA__
il yadayadayada

df yadayadayada
[download]

my $x='il yadayadayada'; 
$x=~s/^(\w{2})(.*)/<$1> $1$2 <$1>/;
print $x;

-- output --
<il> il yadayadayada <il>
[download]

Dude, the output is supposed to be:

<il> il yadayada <il>
[download]

. Therefore, you should have written this:

$x=~s/^(\w{2})(.*)/<$1> $1$2 <$1>/;
[download]

</nitpick>

My code is,

while (<DATA>){
   s#^(\w{2}).*#<$1>$&<\/$1># && print
}

__DATA__
lg alkjslkjs
sl slksjlkjslkjs
slkjslkjs
[download]