On Jan 28, 2013, at 11:57 AM, Angela Barone wrote:
> Hello,
>
> I'm trying to abbreviate ordinals(?) that occur only in the middle of
> an address and I'm having a problem. The line below works:
>
> $test_data =~ s/(\S) North (\S)/$1 N. $2/i;
>
> however, if the address is something like 901 North St., it abbreviates that
> as well. I'm wanting it to work only on an address like
> 53 North Benson St. -> 53 N. Benson St.
>
> Is there a way to know whether or not 'North' is a street name as
> opposed to a direction, or am I asking too much? I was thinking of counting
> the number of "words" between "North' and 'St.' (and if it's zero then don't
> do it), but I wouldn't know how to do that or even if that's the best way to
> do it.
You want to change 'North' to 'N.' unless it is followed directly by 'St'. This
is known as a "negative lookahead assertion" and uses the (?!...) construct:
$test_data =~ s/North (?!St)/N. /;
Of course, you have more cases to handle. Here is a short program to get you
started:
#!/usr/bin/perl
use strict;
use warnings;
my @a = (
'101 South Curtis Dr',
'701 North St',
'901 North Benson Rd',
'802 West Ave',
);
my @streets = qw( St Ave Dr Rd Ln );
my $streets = join('|',@streets);
my @directions = qw( north south east west );
my $directions = join('|',@directions);
for my $s ( @a ) {
$s =~ s{ \b ($directions) \b \s+ (?!$streets) }{ uc substr($1,0,1) . '. '
}iex;
print "$s\n";
}
--
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
http://learn.perl.org/
