User-Defined Types

Cassandra 2.1 introduced user-defined types (UDTs).

Background

Given

a running cassandra cluster with schema:

CREATE KEYSPACE simplex WITH replication = {'class': 'SimpleStrategy', 'replication_factor': 3};
CREATE TYPE simplex.address (street text, zipcode int);
CREATE TYPE simplex.check_in (
  location frozen <address>,
  time timestamp,
  data frozen <tuple<int, text, float>>
);
CREATE TABLE simplex.users (id int PRIMARY KEY, location frozen<address>);

Using User-Defined Types with prepared statements

Given

the following example:

require 'cassandra'

cluster = Cassandra.cluster(consistency: :all)
session = cluster.connect('simplex')
insert  = session.prepare('INSERT INTO users (id, location) VALUES (?, ?)')

session.execute(insert, arguments: [0, Cassandra::UDT.new(street: '123 Main St.', zipcode: 78723)], consistency: :all)
session.execute('SELECT * FROM users').each do |row|
  location = row['location']
  puts "Location: #{location}"
end

When

it is executed

Then

its output should contain:

Location: { street: "123 Main St.", zipcode: 78723 }

Using User-Defined Types with raw CQL

Given

the following example:

require 'cassandra'

cluster = Cassandra.cluster(consistency: :all)
session = cluster.connect('simplex')

session.execute("INSERT INTO users (id, location) VALUES (0, {street: '123 Main St.', zipcode: 78723})", consistency: :all)
session.execute('SELECT * FROM users').each do |row|
  location = row['location']
  puts "Location: #{location}"
end

When

it is executed

Then

its output should contain:

Location: { street: "123 Main St.", zipcode: 78723 }

Inspecting User-Defined Types

Given

the following example:

require 'cassandra'

cluster = Cassandra.cluster

cluster.keyspace('simplex').each_type do |type|
  puts type.to_cql
end

When

it is executed

Then

its output should contain:

CREATE TYPE simplex.address (
  street text,
  zipcode int
);
CREATE TYPE simplex.check_in (
  location frozen <address>,
  time timestamp,
  data frozen <tuple<int, text, float>>
);

Inserting a partially-complete User-Defined Type

Given

the following example:

require 'cassandra'

cluster = Cassandra.cluster(consistency: :all)
session = cluster.connect('simplex')
insert  = session.prepare('INSERT INTO users (id, location) VALUES (?, ?)')

session.execute(insert, arguments: [0, Cassandra::UDT.new(zipcode: 78723)], consistency: :all)
session.execute('SELECT * FROM users').each do |row|
  location = row['location']
  puts "Location: #{location}"
end

When

it is executed

Then

its output should contain:

Location: { street: nil, zipcode: 78723 }

Nesting a User-Defined Type

Given

the following example:

require 'cassandra'

cluster = Cassandra.cluster(consistency: :all)
session = cluster.connect('simplex')

session.execute("CREATE TABLE registration (id int PRIMARY KEY, info frozen<check_in>)")
insert  = session.prepare('INSERT INTO registration (id, info) VALUES (?, ?)')

location = Cassandra::UDT.new(street: '123 Main St.', zipcode: 78723)
tuple = Cassandra::Tuple.new(42, 'math', 3.14)
input = Cassandra::UDT.new(location: location, time: Time.at(1358013521, 123000), data: tuple)

session.execute(insert, arguments: [0, input], consistency: :all)
session.execute('SELECT * FROM registration').each do |row|
  info = row['info']
  puts "Info: {street: #{info.location.street}, zipcode: #{info.location.zipcode}}, #{info.time.httpdate}, #{info.data}"
end

When

it is executed

Then

its output should contain:

Info: {street: 123 Main St., zipcode: 78723}, Sat, 12 Jan 2013 17:58:41 GMT, (42, math, 3.140000104904175)