No. That, in fact, is the whole point of (?
groups.
Why do you think you need to do this?
Ben
Thank you for your attention to my question and your precious time.
In my application, I'll have reformat the telephone number in my
company.
Since they came from all over the world, we have various telephone
format.
Here's some examples:
000-000-000
(00)00-000-0000
!00!00-000-0000
0000000
0000-0000
0000 0000 000
you dont care
What my job is to check the valid telephone number and to give them
out.
In case that I meet a ')' character, I'll have to check if there's a
'(' in the front.
'!' is also valid here.
But all these characters should not be captured.
So I'll have to backreference a previous group
Expression like "((?:[!|]*)\d+\1\d+)" gives a bad result.
Any further hints or advice here?
Thanks again for your reply.
Although your not too clear about your intent,
it is obvious you *think* you can capture individual numbers
in groups, shed surrounding delimeters, validate special delemeters
(one of which is paranthetical closures) then wrap it all up and
stamp it "Special Delivery" for an array.
I dunno, maybe, but if not, an approach like this might get you started.
(this goes into my bin for future reference, if any)
-sln
---------------------------
## ph_capt.pl
##
use strict;
use warnings;
my $str = <<TELNUMS;
phone numbers:
100-junk-(00)
200-000-000-!99!
(30)-00-000-0000
!40!00-000-0000-
5000000
6000-0000
7000 0000 000
!80)-00-000-0000
TELNUMS
print $str;
my @ar = "!40!00-000-0000" =~
/^ # line start
( # begin capture group 1 ($1)
!\d+! # '!' + 1 or more digits + '!'
| # or
\(\d+\) # '(' + 1 or more digits + ')'
| # or
# defined-empty string
) # end group 1
[ \t-]* # 0 or more of these chars
( # begin capture group 2 ($2)
(?: # grouping
\d+[ \t-]* # 1 or more digits + 0 or more of these chars
)+ # end group, do 1 or more times
) # end group 2
$ # line end
/x;
print "\n-------------\n\n";
print "'$_'\n" for (@ar);
print "\n-------------\n\n";
for (split /\n/,$str)
{
next if (!length());
print "Target: '$_'\n";
if ( @ar = /^(!\d+!|\(\d+\)|)[ \t-]*((?:\d+[ \t-]*)+)$/ )
{
my @ngroups = split /[ \t-]/, $ar[1];
$ar[0] =~ s/[!()]//g;
print " ++ matched ". join (' , ',@ngroups) ."\n";
print " special prefix $ar[0]\n" if length $ar[0];
}
else {
print " -- invalid phone #\n";
}
print "\n";
}
__END__
Output:
phone numbers:
100-junk-(00)
200-000-000-!99!
(30)-00-000-0000
!40!00-000-0000-
5000000
6000-0000
7000 0000 000
!80)-00-000-0000
-------------
'!40!'
'00-000-0000'
-------------
Target: 'phone numbers:'
-- invalid phone #
Target: '100-junk-(00)'
-- invalid phone #
Target: '200-000-000-!99!'
-- invalid phone #
Target: '(30)-00-000-0000'
++ matched 00 , 000 , 0000
special prefix 30
Target: '!40!00-000-0000-'
++ matched 00 , 000 , 0000
special prefix 40
Target: '5000000'
++ matched 5000000
Target: '6000-0000'
++ matched 6000 , 0000
Target: '7000 0000 000'
++ matched 7000 , 0000 , 000
Target: '!80)-00-000-0000'
-- invalid phone #
