Paging

Do not pull a whole table into memory. page() and pageSize() slice a query into pages, and the query tracks the totals you need to render page controls.

Paging is 1-indexed: the first page is page 1.

Page a query

PikaList<User> users = User.find()
        .where("active = :a", "a", true)
        .page(3)
        .pageSize(20)
        .fetchList();

Leave off pageSize() and Pika uses the default, which is 20 (configurable below).

Page metadata

When you run a paged query, Pika also runs a COUNT(*) with the same where conditions, so it can tell you how many rows and pages exist. Read the totals straight off the query:

PikaClassQuery<User> q = User.find().where("active = :a", "a", true).page(2);
PikaList<User> users = q.fetchList();

long total    = q.totalCount();    // matching rows across all pages
long pages    = q.totalPages();    // total pages at the current page size
boolean first = q.isFirstPage();
boolean last  = q.isLastPage();

Page links

Building “next” and “previous” links means rewriting the page= query parameter by hand. Pika does it for you:

String url  = "/users?sort=desc&page=2";
String next = q.nextPageURL(url);       // /users?sort=desc&page=3
String prev = q.previousPageURL(url);   // /users?sort=desc&page=1

Pass a second argument if your page parameter is not named page.

Defaults

Set the default page size when you build the ORM. If your database has unusual offset syntax, override that too.

PikaORM orm = new PikaORM("jdbc:sqlite:app.db")
        .withDefaultPageSize(100)
        .withOffsetClause("LIMIT {0} OFFSET {1}")  // {0} = limit, {1} = offset
        .makeDefaultORM();