quickstore

gistfile1.txt

<?php
/*

create table store(
 id int primary key,
 value blob -- 65535 bytes
);

*/

class quickStore
{
        private $db;
        private $zerofill;

        public function __construct($dbname='tmp', $user='root', $pass='')
        {
                $this->db = new PDO('mysql:host=localhost;dbname='.$dbname, $user, $pass);
                $this->zerofill = str_repeat("\0", 65535);
        }

        // initialise new key
        public function initKey($key, $list=false)
        {
                $stmt = $this->db->prepare('insert into store(id, value) values(:k, :v)');
                $stmt->execute([':k' => $key, ':v' => $list?$list:$this->zerofill]);
        }

        public function insert($key, $idx, $value)
        {
                $idx = ($idx*4);

echo "Inserting key: $key at idx: $idx\n";
                $value = pack('L', (int)$value);
                $stmt = $this->db->prepare('update store set value = concat(substring(value, 1, '.$idx.'), :v, substring(value, '.$idx.'+5, length(value))) where id = :k');
                $stmt->bindParam(':k', $key);
                $stmt->bindParam(':v', $value, PDO::PARAM_LOB);
                $stmt->execute();
        }

        public function select($key, $idx, $range=1)
        {
                $idx = ($idx*4);
                $range = ($range*4);
echo "Fetching key: $key at idx: $idx with range: $range\n";
                $stmt = $this->db->prepare('select substring(value, '.$idx.'+1, '.$range.') as value from store where id = :k');
                $stmt->execute([':k' => $key]);
                $row = $stmt->fetch(PDO::FETCH_ASSOC);
                return unpack('L*', $row['value']);
        }

}

$qs = new quickStore();


$qs->initKey(1);
$qs->initKey(2);
$qs->initKey(3);

$qs->insert(1, 1, 42);
$qs->insert(2, 2, 43);
$qs->insert(2, 3, 44);
$qs->insert(3, 3, 45);

print_r($qs->select(1, 1));
print_r($qs->select(2, 2, 2));
print_r($qs->select(3, 3));

Hi Sandeep

MySQL will manage the caching of the BLOB as and when needed. Updates will also be managed in MySQL effectively so you shouldn't need to worry about your application using large amounts of memory.

I've added the ability to read ranges to the example code. When you say 10k updates .. how quickly do they need to be processed?

regards
Andrew