Fwd: slice of an arrayref

[shawn wilson]

i am having an issue here, i think the bulk of it revolves around this;
        print join(',', @uword{ $seen{ $key }[ 1 ] }[ 1 .. $#{ @uword{
$seen( $key ) } } ] );

however, given how that looks, i'm thinking i've got others.

so, pretty much, what i do is go through each line and take out the words
and take out the duplicate words. i take the hash of those words and put the
full text from the lines they came from and push it onto the array reference
to that hashref. i then go and search a database for each word and take out
the duplicates of the results (keeping the search terms with the results -
though there's a minor logic issue with this i think - if two searches bring
up the same text, i'll just know about the first). the result string and the
word that was searched to find the string are stored to the same type of
array reference to a hashref. that all works fine (exept my many to many
problem i noted above). my issue is how to print out a slice of an arrayref
of a hashref?

my underlying problem (i think) is that there has got to be a better way?

i've got half a mind to create a temp table in the db to do this, but that
doesn't seem like my 'better way' either.

thanks

#!/usr/bin/perl

##### WHAT #
# What names return per search
##########

use strict;
use warnings;
use DBI;

        my $searcher = "owner.manager, owner.owner, owner.manown";

my $dbh = DBI->connect('DBI:mysql:db;host=localhost',
                        'user', 'word')
                or die "Database connection: $!";

open( FILE, "< $ARGV[0]" );

my %uword = ();
my %seen = ();
my $count = 0;
my @data;
my $key;
my $i = 0;

while ( <FILE> ) {

        my $line = $_;

        chomp ($line);
        my @word = split / /, $line;


        $count = 0;
        while ( $word[ $count ] ) {
                $word[ $count ] =~ tr/^[\-a-zA-Z]//;
                $word[ $count ] =~ s/\'/\\\'/g;
                $count++;
        }

        foreach my $string ( @word ) {
                if ( $uword{ $string }[ 0 ] == 1 ) {
                        push @{ $uword{ $string } }, $line;
                        next;
                }
                $uword{ $string }[ 0 ] = 1;
                push @{ $uword{ $string } }, $line;
        }
}


for my $key ( keys %uword ) {

        my ( $number, $owner, $manown, $manager );

        my $select =    qq/SELECT $searcher /;
        my $from =              qq/FROM owner, spd /;
        my $where =             qq/WHERE MATCH( $searcher ) AGAINST('+$key'
IN BOOLEAN MODE) /;
        my $join =              qq/AND owner.num = spd.num/;

        my $query = $select . $from . $where . $join;

        print "SQL: $query\n";
        my $sth = $dbh->prepare( $query );
        $sth->execute;

        $sth->bind_columns( \$manager, \$owner, \$manown );

        while ( $sth->fetch ) {

                if ( defined( $owner ) ) {
                        $data[$count] = $owner;
                        $count++;
                }
                if ( defined( $manown ) ) {
                        $data[$count] = $manown;
                        $count++;
                }
                if ( defined( $manager ) ) {
                        $data[$count] = $manager;
                        $count++
                }
        }

# @data holds full names of data.

        foreach my $string ( @data ) {   # dedupe data and sanity check.
                next if !defined ($string);   # should never be true.
                next if $seen{ $string }[ 0 ] == 1; # check %seen hash /
array for dupe
                $seen{ $string }[ 0 ] = 1;    # define hash of array and
assign check to it.
                $seen{ $string }[ 1 ] = $key; # add word from line to hash
for reference.
        }
}

$dbh->disconnect;

# $key                             - (below) is the deduped sql string
# $seen { $key }[ 1 ]              - is the word searched in sql
# $uword { '$seen{ $key }[ 1 ]' }  - should be the line(s) of test the
search string came from

for my $key ( keys %seen ) {

        print "$seen{ $key }[ 1 ], $key,";
        print join(',', @uword{ $seen{ $key }[ 1 ] }[ 1 .. $#{ @uword{
$seen( $key ) } } ] );
        print "\n";
}