Example about how to use map cql type with DataStax java driver

Question

I am trying to use the datastax java driver to update and query a column family that has a map field. Does anyone an example about how to use cql collections with the Datastax Java Driver?

Thanks

Answer 1

I will add some examples of using the CQL collections with both simple and prepared statements to the current Java driver doc.

You can use CQL collections with prepared statements. There's an example in the Java driver doc in the quick start section.

http://docs.datastax.com/en/developer/java-driver/2.0/java-driver/quick_start/qsSimpleClientBoundStatements_t.html

Step 4 binds a Java HashSet object to a CQL set column in the songs table.

Answer 2

Normally I'd ask what you've tried, but I know that this isn't in the DataStax documentation for the Java Driver. I'll go through what worked for me.

A couple of things to note:

• The DataStax Cassandra Java class directed me to put my variables directly into the CQL text string (instead of binding the map). I'm guessing that binding collections wasn't working at the time of production (for the class videos).
• Collections can't be queried using DevCenter, so you'll need to check their values via the command line with cqlsh if you want to see what they are outside your app.

To update an existing row (in the "users" table which has a Map<varchar,varchar> phone_numbers), try something like this:

String cqlQuery = "UPDATE users SET phone_numbers = phone_numbers + ";
cqlQuery += "{'" + phoneType + "':'" + phoneNumber+ "'} ";
cqlQuery += "WHERE username = ?";

PreparedStatement preparedStatement = getSession().prepare(cqlQuery);
BoundStatement boundStatement = preparedStatement.bind(user);
getSession().execute(boundStatement);

The better way to do this (assuming a Map<String,String> phoneNumbers), is to bind the collection to the prepared statement, like this:

String cqlQuery = "UPDATE users SET phone_numbers = ? ";
cqlQuery += "WHERE username = ?";

PreparedStatement preparedStatement = getSession().prepare(cqlQuery);
BoundStatement boundStatement = preparedStatement.bind(phoneNumbers,user);
getSession().execute(boundStatement);

Likewise, to read it back out:

String cqlQuery2 = "SELECT phone_numbers FROM users WHERE username = ?";

PreparedStatement preparedStatement2 = getSession().prepare(cqlQuery2);
BoundStatement boundStatement2 = preparedStatement2.bind(user);
ResultSet result2 = getSession().execute(boundStatement2);

Map<String,String> newMap = result2.one().getMap("phone_numbers", String.class, String.class);

They just covered this today in the (free) CAS101J class on DataStax Academy.

Answer 3

Here is how I did it; This has mapping for a tuple column in Cassandra as well as the map column in Cassandra. I was using Scala with the DataStax Cassandra Java Driver

Needed imports

import java.lang
import java.text.SimpleDateFormat

import com.datastax.driver.core._
import com.datastax.driver.core.querybuilder.QueryBuilder

import scala.collection.Map
import org.apache.spark.SparkContext
import org.apache.spark.SparkConf
import com.datastax.spark.connector.cql.CassandraConnector
import org.apache.spark.rdd.RDD
import scala.collection.JavaConversions._

Code snippet

val simpleDateFormat: SimpleDateFormat = new SimpleDateFormat("dd-MM-yyyy H:mm:ss")
val start_date: java.util.Date = simpleDateFormat.parse(val("StartTime").replaceAll(":(\\s+)", ":"))

val b_tuple= session.getCluster().getMetadata().newTupleType(DataType.cint(), DataType.cint(), DataType.text())
val time_tuple = session.getCluster().getMetadata().newTupleType(DataType.timestamp(), DataType.timestamp())

val time_tuple_value = time_tuple.newValue(start_date, end_date)
val b_tuple_value = bin_cell_tuple.newValue(b._1: lang.Integer, b._2: lang.Integer, val("xxx"))

val statement_2: Statement = QueryBuilder.insertInto("keyspace", "table_name")
          .value("b_key", bin_cell_tuple_value)
          .value("time_key", time_tuple_value)
          .value("some_map", mapAsJavaMap(my_scala_map))
session.executeAsync(statement_2)