Not necessarily. It won't work for

print($regexp)  Expect        Gives
--------------  ------------  ----------
\Q*\E\Q*\E      \*\*          \*\Q*\E
\Q**LOL**       \*\*LOL\*\*   \Q**LOL**
\\Quit\\Exit    \\Quit\\Exit  \Quit\Exit
\Qfoo\\E        foo\\\\E      foo\\
[download]

Solution:

my $in_quote = 0;

$regexp =~ s/([^\\]|\\.)/
   if ($in_quote) {
      if ($1 eq '\\E') {
         $in_quote = 0;
         ''
      } else {
         quotemeta($1)
      }
   } else {
      if ($1 eq '\\Q') {
         $in_quote = 1;
         ''
      } else {
         $1
      }
   }
/eg;
[download]

Update: Alternative:

$regexp =~ s/
   \G
   (
      (?:[^\\]|\\[^Q])*
   )
   (?:
      \\Q
      (?:[^\\]|\\[^E])*
      (?:\\E)?
   )?
/
   $1 . (defined($2) ? quotemeta($2) : '')
/xge;
[download]

Neither snippet is fully tested. In fact, both are known to be unable to handle regexps in which (?{...}) or (?{{...}}) are used. What's wrong with what I suggested in my earlier post?

But what's wrong with what I suggested in my earlier post?


The regexes are in a DB. In the beginning, all of them where full regexes (no need for \Q...\E, etc...). But now, some of them are created by an user with a GUI. These "new" regexes are "plain text", and I want to ignore the metacharacters of *them* only, and not of all the regexes. If I "quotemeta" any regex i see, I can end up quoting real regexes :-(

Oh, and the interface is not in Perl, so I don't have quotemeta there. All I can do is surround with \Q...\E and hope it works, but as I found out (the bad way), it doesn't :-(
-- 6x9=42

quotemeta should be easy to implement in any language. Adding a slash in front of any characters other than [a-zA-Z0-9] should do the trick.