Your source string needs to be in single quotes, otherwise Perl will try to interpolate '@begin' and '@end' as arrays. You also don't want to put your replacement strings inside square brackets, that groups them as a single array reference inside of @replace, rather than as individual strings.

Anyway, this should get you started:

use strict;
use warnings;

my $string='
hi hello @begintobe
replace @end asas x x x x x zxzxax
hi hello hi hellooo
@beginthis has tobe
replaced @end ...';

my @replace= ("replacestring1","replacestring2");

$string =~ s|\@begin.*?\@end
            |shift @replace // '[null]'
            |sexg;

print $string;
[download]

Usually regular expression substitutions are written as:

But in my code above, the vertical bar (or pipe) symbol (|) is used instead. It's also split over 3 lines which is allowed when you use the "x" modifier.

The "e" modifier means that instead of substitution text, the given code will be executed for each @begin ... @end match. The code shown, removes the first element of the array and returns it for the current substitution. If the @array becomes empty, the literal value '[null]' will be used instead.

The "g" modifier causes the substitution to be done as many times as there are matches found in the string -- consuming another array element each time. Without it only the first substitution would be done.

And lastly the "s" modifier means that the period character in the pattern, will match newline characters since those appear in your example input string.

Much more information can be found in the perlre and perlretut documentation.

Slightly off-topic, but when writing such a templating engine, I prefer to deal with missing replacement values by leaving the template parameters as-is (that is, replacing the template parameter by itself instead of a (missing) value):

$string =~ s|(\@begin.*?\@end)
            |shift @replace // 'Missing parameter for ' . $1
            |sexg;
[download]

I find this helps me debug the template better, as I easily see which template expression failed. This still does not help me see why a value was missing, but at least I know where the value was missing.

I know what I mean. Why don't you?.

That's your second post on the subject (the first here), so by now you should have some code. Go on and try something, and present some runnable code that fails to do what you want, and then you can ask for advice. But just don't expect perlmonks to magically sprout whatever solution you need without any visible effort on your side.

What code have you written and where do you have problems with it?

How does your code fail?

Have you looked at using one of the established template modules instead of rolling your own template?

Here's another approach. This needs Perl version 5.10+ for  \K (if you have < 5.10, let me know; I can supply adjusted code).

c:\@Work\Perl>perl -wMstrict -le
"use 5.010;
 ;;
 my $s = qq{xxx \@B grab \n this \@E yy zzzz \@B and \n this \n too \@
+E qq};
 print qq{[[$s]]};
 ;;
 my $start = qr{ \@B }xms;
 my $stop  = qr{ \@E }xms;
 ;;
 my $replace = [ 'ZOT', 'ZONK' ];
 ;;
 my $i = 0;
 $s =~ s{ $start \K .*? (?= $stop) }
        { $replace->[ $i++ % @$replace ] }xmsg;
 print qq{<<$s>>};
"
[[xxx @B grab
 this @E yy zzzz @B and
 this
 too @E qq]]
<<xxx @B ZOT @E yy zzzz @B ZONK @E qq>>
[download]

Pay particular attention to Loops's and Corion's remarks here and here about what should happen when the  $replace array is exhausted (I just wrap around and start re-using replacement strings), and also here about using a real templating engine rather than trying to roll your own.

Update: Removed a needless capture group:  .*? vice  (.*?)

i had a string which has a pattern @begin...@end ,...

you don't have it anymore because double strings interpolate... you should post an actual short perl program, perlintro can help you get started, this is the beginning where you have to start

 #!/usr/bin/perl
use strict;
use File::Find;
use warnings;
my @file=("hh","yy","zz","aa","bb","hh","yy","zz","aa","bb","hh","yy",
+"zz","aa","bb");
my $i=0;
my $string='hi hello @begintobe 
replace @end asas x x x x x zxzxax
hi hello hi hellooo
@beginthis has tobe 
replaced @end';
my $x=$file[$i];
        while ($string =~ m/\@begin([\w\W]*?)\@end/g) {
        $i++;
        $string =~ s/\@begin([\w\W]*?)\@end/$x/s;
        $x=$file[$i];
         print $i."\n";
        }
print $string;
[download]

ok, my program looks as below

#!/usr/bin/perl
use strict;
use File::Find;
use warnings;
my @file=("hh","yy","zz","aa","bb","hh","yy","zz","aa","bb","hh","yy",
+"zz","aa","bb");
my $i=0;
$string="
hi hello @begintobe 
replace @end asas x x x x x zxzxax
hi hello hi hellooo
@beginthis has tobe 
replaced @end";
my $x=$file[$i];
        while ($string =~ s/\@begin([\w\W]*?)\@end/$x/ge) {
        $i++;
        $x=$file[$i];
         print $i."\n";
        }
print $string;
[download]

and i am getting output as below so far<\p>

##output
hi hello hh asas x x x x x zxzxax
hi hello hi hellooo
hh
[download]

only the first element in the arrray is getting replaced in every match...

This does not even compile, and even if it did, $string does not contain either @begin or @end (try printing it without doing any replacement). Have a look at Loops's post on that point.

If strict gives you errors (when you actually use it, and don't just put it when you update your code on perlmonks), it's nearly always a sign that the code is wrong, and doesn't mean what you expect. Same goes for warnings.