Beyond Key-Value Lookup - Learning Apache Cassandra - page 81

Paginating over rows in a partition
As our users create more and more status updates, we'll build pagination functionality into
MyStatus so that the information on the page doesn't overwhelm readers. For the sake of
convenience, let's say that each page will only contain three status updates.
To retrieve the first page, we'll use the LIMIT keyword that we first encountered in
Chapter 2 , The First Table :
SELECT "id", DATEOF("id"), "body"
FROM "user_status_updates"
WHERE "username" = 'alice'
LIMIT 3;
As expected, Cassandra will give us the first three rows in ascending order of id :
Now, we'll ask for the collection of rows where the id value is strictly greater than the last
id we saw:
SELECT "id", DATEOF("id"), "body"
FROM "user_status_updates"
WHERE "username" = 'alice'
AND id > 3f9df710-e8f7-11e3-9211-5f98e903bf02
LIMIT 3;
This is similar to the pagination query for users that we made in Chapter 2 , The First
Table , but it's in some ways simpler. We keep the restriction of rows to alice 's partition,
and then add a simple greater than restriction for id .
Unlike in the users table, we don't need to use the TOKEN function, because id is not a
partition key. Since clustering columns are stored in semantic order rather than token order,