note LanX Thanks!<P> IMHO sounds like far less than 1‰ of all CPAN modules might rely on this.<P> <hr>UPDATE:<P> OK found another use-case, I just recently had the need in my [id://898097|ORG-Parser] to distinguish the range delimiters and the range "body" with a flip-flop-operator, like for ORG's "BEGIN/END"-blocks.<P> This can be simplified (i.e. more DRY), if one has access to the last successful pattern:<P> <c> DB<114> for(0..99) {print if (/10/../20/ and not //)} 111213141516171819 DB<115> for(0..99) {print if (/10/../20/)} 1011121314151617181920 DB<116> for(0..99) {print if (/10/../20/ and //)} 1020 </c><P> I think that's a more frequent application, I even slightly remember seeing it in Friedl's book.<P> But I'd rather prefer an explicit special varą, something like <c>$PATTERN</c> or <c>$&&</c> (in analogy to <c>$MATCH</c> resp. <c>$&</c>) .<P> <!-- Node text goes above. Div tags should contain sig only --> <div class="pmsig"><div class="pmsig-708738"> <p>Cheers Rolf </div></div><P> 1) an special var has the advantage that the regex itself can be accessed, e.g. printed. 899784 899822