Result paging

Starting with Cassandra native protocol v2 (used by Apache Cassandra 2.0), paging through query results is allowed.

Page size can be specified by setting the $pageSize attribute of execution-options or cluster-wide, using Cassandra\Cluster\Builder::withDefaultPageSize().

Once a Cassandra\Rows object has been received, next page can be retrieved using Cassandra\Rows::nextPage() or Cassandra\Rows::nextPageAsync() methods for synchronous and asynchronous next page retrieval accordingly.

To check if the next page is available, use Cassandra\Rows::isLastPage()

Background

Given

a running Cassandra cluster

And

the following schema:

CREATE KEYSPACE simplex WITH replication = {
  'class': 'SimpleStrategy',
  'replication_factor': 1
};
USE simplex;
CREATE TABLE paging_entries (key text, value int, PRIMARY KEY(key, value));
INSERT INTO paging_entries (key, value) VALUES ('a', 0);
INSERT INTO paging_entries (key, value) VALUES ('b', 1);
INSERT INTO paging_entries (key, value) VALUES ('c', 2);
INSERT INTO paging_entries (key, value) VALUES ('d', 3);
INSERT INTO paging_entries (key, value) VALUES ('e', 4);
INSERT INTO paging_entries (key, value) VALUES ('f', 5);
INSERT INTO paging_entries (key, value) VALUES ('g', 6);
INSERT INTO paging_entries (key, value) VALUES ('h', 7);
INSERT INTO paging_entries (key, value) VALUES ('i', 8);
INSERT INTO paging_entries (key, value) VALUES ('j', 9);
INSERT INTO paging_entries (key, value) VALUES ('k', 10);
INSERT INTO paging_entries (key, value) VALUES ('l', 11);
INSERT INTO paging_entries (key, value) VALUES ('m', 12);

Paging through results synchronously

Given

the following example:

<?php
$cluster   = Cassandra::cluster()
               ->withContactPoints('127.0.0.1')
               ->build();
$session   = $cluster->connect("simplex");
$options   = array('page_size' => 5);
$rows      = $session->execute("SELECT * FROM paging_entries", $options);

while (true) {
    echo "entries in page: " . $rows->count() . "\n";

    foreach ($rows as $row) {
        echo "key: " . $row['key'] . ", value: " . $row['value'] . "\n";
    }

    if ($rows->isLastPage()) {
        break;
    }

    $rows = $rows->nextPage();
}

When

it is executed

Then

its output should contain:

entries in page: 5
key: a, value: 0
key: c, value: 2
key: m, value: 12
key: f, value: 5
key: g, value: 6
entries in page: 5
key: e, value: 4
key: d, value: 3
key: h, value: 7
key: l, value: 11
key: j, value: 9
entries in page: 3
key: i, value: 8
key: k, value: 10
key: b, value: 1

Accessing page info after loading next one

Given

the following example:

<?php
$cluster   = Cassandra::cluster()
               ->withContactPoints('127.0.0.1')
               ->build();
$session   = $cluster->connect("simplex");

$firstPageRows = $session->execute("SELECT * FROM paging_entries", array('page_size' => 10));
echo "Page 1: " . ($firstPageRows->isLastPage() ? "last" : "not last") . PHP_EOL;

$secondPageRows = $firstPageRows->nextPage();
echo $firstPageRows->isLastPage() ? "1: last\n" : "1: not last\n";
echo $secondPageRows->isLastPage() ? "2: last\n" : "2: not last\n";

echo "entries in page 1: " . $firstPageRows->count() . "\n";
echo "entries in page 2: " . $secondPageRows->count() . "\n";

When

it is executed

Then

its output should contain:

1: not last
1: not last
2: last
entries in page 1: 10
entries in page 2: 3

Use paging state token to get next result

Given

the following example:

<?php
$cluster   = Cassandra::cluster()
               ->withContactPoints('127.0.0.1')
               ->build();
$session   = $cluster->connect("simplex");
$query     = "SELECT * FROM paging_entries";
$options = array('page_size' => 2);
$result = $session->execute($query, $options);

foreach ($result as $row) {
  printf("key: '%s' value: %d\n", $row['key'], $row['value']);
}

while ($result->pagingStateToken()) {
    $options = array(
        'page_size' => 2,
        'paging_state_token' => $result->pagingStateToken()
    );

    $result = $session->execute($query, $options);

    foreach ($result as $row) {
      printf("key: '%s' value: %d\n", $row['key'], $row['value']);
    }
}

When

it is executed

Then

its output should contain:

key: 'a' value: 0
key: 'c' value: 2
key: 'm' value: 12
key: 'f' value: 5
key: 'g' value: 6
key: 'e' value: 4
key: 'd' value: 3
key: 'h' value: 7
key: 'l' value: 11
key: 'j' value: 9
key: 'i' value: 8
key: 'k' value: 10
key: 'b' value: 1