How do I quote the special character + and a one-letter-plus-two-digit string?

I have the following text to modify in Perl:

"tti_d01_DA_" + it + "H"

In my replacement text, I want to replace d01 with a variable ${Dom} and DA with ${simul}, so that I have:

"tti_${Dom}_${simul}_" + it + "H"

I have tried the following searches for the original string:

s/"tti_.\d{2}_\w+_" \+ \w+ \+ "H"/"tti_${Dom}_${simul}_" + it + "H"/;
s/"tti_(?=.*[a-z])(?=.*[0-9]) (?=.*[_]) \w+ (?=.*[_]) (?=.*["])\s(?=.*[+])\s\w+\s(?=.*[+])\s"H"/"tti_${Dom}_${simul}_" + it + "H"/;

I thought this should be a very simple issue, but somehow I am not able to get it. Both the above do not find what I am looking for.

This works in my tests

s/"tti_\K d[0-9]{2}_\w+_"(?= \+ it \+ "H")/${Dom}_${simul}_/

In the original same string there is DA pattern, which as comments say can be NODA instead. That is captured by \w+ above, following the attempt in the question. However, if it can only be either DA or NODA, as a comment says, then one may want to use DA|NODA (instead of \w+), so that the match fails for anything else. Or something in between, like [A-Z]+.

This is a choice to make that depends on the context -- either make a pattern as restrictive as possible, to protect from unexpected (and unacceptable) input, or make it more permissive to catch more of what is acceptable.

The question doesn't say what happens with shown attermps but here are some comments.

The only outright problem that I readily see is the $s in the replacement side, which have to be escaped if they are to be used as literal characters, otherwise it will look for variables $Dom and $simul.^† (But then strict wouldn't let the program compile because there are no such variables -- do you have use strict; in your program?)

Next, I use \K lookbehind so that all that's been matched up to that point is dropped (from $& and so kept in the string) and doesn't have to be re-entered in the replacement side.

In the same vein, I use a lookahead (?=...) for patterns that follow what needs to be replaced. That is an assertion (zero-length), which doesn't consume anything, so those sub patterns don't need to be replaced into the string either.

These two tweaks 'lighten' the replacement side (but are not necessary).

Test program

use warnings;
use strict;
use feature 'say';

my $str = shift// q("tti_d01_DA_" + it + "H");
#say $str; 

$str =~ s/"tti_\Kd[0-9]{2}_\w+_"(?= \+ it \+ "H")/${Dom}_${simul}_/;

say $str; 

^† It appears from a comment that the problem may have been with variable naming: there are arrays @Dom and @simul in the program, and the replacements need be $Dom[index] (etc).

If that indeed is the whole problem then having use strict; at the beginning would've caught it. It would not let the program even compile and it would report that the variable $Dom does not exist.

Having use strict; and use warnings; is directly helpful, and really necessary I'd say. No wonder that they are enabled by default by many tools and, finally, in newer Perls.

The regexp pattern you are using is correct.

However, $ is special in double-quote string literals, which include the second part of s///. Just escape them.

Always use use strict; use warnings; or equivalent as it would caught this error!

In other words, all you need to do is replace

s/"tti_.\d{2}_\w+_" \+ \w+ \+ "H"/"tti_${Dom}_${simul}_" + it + "H"/;

with

s/"tti_.\d{2}_\w+_" \+ \w+ \+ "H"/"tti_${Dom}_${simul}_" + it + "H"/;

Demo:

$ perl -Mv5.10 -we'
   $_ = q{"tti_d01_DA_" + it + "H"};
   s/"tti_.\d{2}_\w+_" \+ \w+ \+ "H"/"tti_${Dom}_${simul}_" + it + "H"/;
   say;
'
"tti_${Dom}_${simul}_" + it + "H"