Re: duplicate entries for ->fetch?

[csarnows]

On Dec 3, 2004, at 8:44 AM, Chad Kellerman wrote:
eval{
    $dbh{'info'} = $dbh{'conn'}->prepare(qq{
                        SELECT User, Host from db
                        WHERE Db = ?
                        }) || die $DBI::errstr;
    $dbh{'info'}->bind_param(1, $db, SQL_VARCHAR);
    $dbh{'info'}->execute();
    $dbh{'info'}->bind_columns(\$dbuser, \$host);
    while($dbh{'info'}->fetch){
        $drop{'user'} = $dbuser;
        $drop{'host'} = $host;
You're overwriting the hash elements $drop{'user'} and $drop{'host'}
each time you go through this loop. So you'll end up only with the last 
one.

    }
    $dbh{'info'}->finish();
No need to call finish after fetching with a loop.
};
But when I issue:
foreach $dbuser(sort keys %drop){
    print "$drop{'user'} \-\> $drop{'host'}\n";
}
You're not actually using the hash key you've stored in $dbuser.
The keys in %drop will be 'user' and 'host'.
Since there are 2 keys, the loop will run twice, so the result below is 
expected.

I get
ckell -> localhost
ckell -> localhost
You could use either an array of hashes or a hash of hashes.
e.g.
while ($dbh{'info'}->fetch) {
  $drop{$dbuser . $host}{'user'} = $dbuser;
  $drop{$dbuser . $host}{'host'} = $host;
}
...
foreach $dbuser (sort keys %drop) {
  print $drop{$dbuser}{'user'}, ' -> ', $drop{$dbuser}{'host'}, "\n";
}
Note that the method I used to generate the key has a fault: if you have
 a dbuser 'Bob' at host 'obob', and a dbuser 'Bobo' at host 'bob', the 
keys
will collide, and the later values will overwrite the earlier values.