An easy way to create large dataset when playing/demonstrating Couchbase -or any other NoSQL engine- is to inject Twitter feed into your database.
For this small application I am using:
- Couchbase Server 2.0 Server
- Couchbase Java SDK (will be installed by Maven)
- Twitter4J (will be installed by Maven)
- Twitter Streaming API called using Twitter4J
In this example I am using Java to inject Tweets into Couchbase, you can obviously use another langage if you want to.
The sources of this project are available on my Github repository Twitter Injector for Couchbase you can also download the Binary version here, and execute the application from the command line, see Run The Application paragraph. Do not forget to create your Twitter oAuth keys (see next paragraph)
Create oAuth Keys
The first thing to do to be able to use the Twitter API is to create a set of keys. If you want to learn more about all these keys/tokens take a look to the oAuth protocol : http://oauth.net/
1. Log in into the Twitter Development Portal : https://dev.twitter.com/
2. Create a new Application
Click on the ‘Create an App’ link or go into the ‘User Menu > My Applications > Create a new application’
3. Enter the Application Details information
4. Click ‘Create Your Twitter Application’ button
Your application’s OAuth settings are now available :
5- Go down on the Application Settings page and click on the ‘Create My Access Token’ button
You have now all the necessary information to create your application:
- Consumer key
- Consumer secret
- Access token
- Access token secret
These keys will be uses in the twitter4j.properties file when running the Java application from the command line see
Create the Java Application
The following code is the main code of the application:
package com.couchbase.demo;
import com.couchbase.client.CouchbaseClient;
import org.json.JSONException;
import org.json.JSONObject;
import twitter4j.*;
import twitter4j.json.DataObjectFactory;
import java.io.InputStream;
import java.net.URI;
import java.util.ArrayList;
import java.util.List;
import java.util.Properties;
public class TwitterInjector {
public final static String COUCHBASE_URIS = "couchbase.uri.list";
public final static String COUCHBASE_BUCKET = "couchbase.bucket";
public final static String COUCHBASE_PASSWORD = "couchbase.password";
private List<URI> couchbaseServerUris = new ArrayList<URI>();
private String couchbaseBucket = "default";
private String couchbasePassword = "";
public static void main(String[] args) {
TwitterInjector twitterInjector = new TwitterInjector();
twitterInjector.setUp();
twitterInjector.injectTweets();
}
private void setUp() {
try {
Properties prop = new Properties();
InputStream in = TwitterInjector.class.getClassLoader().getResourceAsStream("twitter4j.properties");
if (in == null) {
throw new Exception("File twitter4j.properties not found");
}
prop.load(in);
in.close();
if (prop.containsKey(COUCHBASE_URIS)) {
String[] uriStrings = prop.getProperty(COUCHBASE_URIS).split(",");
for (int i=0; i<uriStrings.length; i++) {
couchbaseServerUris.add( new URI( uriStrings[i] ) );
}
} else {
couchbaseServerUris.add( new URI("http://127.0.0.1:8091/pools") );
}
if (prop.containsKey(COUCHBASE_BUCKET)) {
couchbaseBucket = prop.getProperty(COUCHBASE_BUCKET);
}
if (prop.containsKey(COUCHBASE_PASSWORD)) {
couchbasePassword = prop.getProperty(COUCHBASE_PASSWORD);
}
} catch (Exception e) {
System.out.println( e.getMessage() );
System.exit(0);
}
}
private void injectTweets() {
TwitterStream twitterStream = new TwitterStreamFactory().getInstance();
try {
final CouchbaseClient cbClient = new CouchbaseClient( couchbaseServerUris , couchbaseBucket , couchbasePassword );
System.out.println("Send data to : "+ couchbaseServerUris +"/"+ couchbaseBucket );
StatusListener listener = new StatusListener() {
@Override
public void onStatus(Status status) {
String twitterMessage = DataObjectFactory.getRawJSON(status);
// extract the id_str from the JSON document
// see : https://dev.twitter.com/docs/twitter-ids-json-and-snowflake
try {
JSONObject statusAsJson = new JSONObject(twitterMessage);
String idStr = statusAsJson.getString("id_str");
cbClient.add( idStr ,0, twitterMessage );
System.out.print(".");
} catch (JSONException e) {
e.printStackTrace();
}
}
@Override
public void onDeletionNotice(StatusDeletionNotice statusDeletionNotice) {
}
@Override
public void onTrackLimitationNotice(int numberOfLimitedStatuses) {
}
@Override
public void onScrubGeo(long userId, long upToStatusId) {
}
@Override
public void onException(Exception ex) {
ex.printStackTrace();
}
};
twitterStream.addListener(listener);
twitterStream.sample();
} catch (Exception e) {
e.printStackTrace();
}
}
}Some basic explanation:
- The
setUp()method simply reads thetwitter4j.propertiesfile from the classpath to build the Couchbase connection string. - The injectTweets opens the Couchbase connection -line 76- and calls the TwitterStream API.
- A
Listeneris created and will receive all theonStatus(Status status) from Twitter. The most important method is onStatus() that receive the message and save it into Couchbase. - One interesting thing : since
Couchbaseis a JSON Document database it allows your to just take the JSON String and save it directly.
cbClient.add(idStr,0 ,twitterMessage);
Packaging
To be able to execute the application directly from the Jar file, I am using the assembly plugin with the following informations from the
pom.xml :
... <archive> <manifest> <mainclass>com.couchbase.demo.TwitterInjector</mainclass> </manifest> <manifestentries> <class-path>.</class-path> </manifestentries> </archive> ...
Some information:
- The mainClass entry allows you to set which class to execute when running java -jar command.
- The Class-Path entry allows you to set the current directory as part of the classpath where the program will search for the twitter4j.properties file.
- The assembly file is also configure to include all the dependencies (Twitter4J, Couchbase client SDK, …)
If you do want to build it from the sources, simply run :
mvn clean package
This will create the following Jar file . /target/CouchbaseTwitterInjector.jar
Run the Java Application
Before running the application you must create a twitter4j.properties file with the following information :
twitter4j.jsonStoreEnabled=true oauth.consumerKey=[YOUR CONSUMER KEY] oauth.consumerSecret=[YOUR CONSUMER SECRET KEY] oauth.accessToken=[YOUR ACCESS TOKEN] oauth.accessTokenSecret=[YOUR ACCESS TOKEN SECRET] couchbase.uri.list=http://127.0.0.1:8091/pools couchbase.bucket=default couchbase.password=
Save the properties file and from the same location run:
jar -jar [path-to-jar]/CouchbaseTwitterInjector.jar
This will inject Tweets into your Couchbase Server. Enjoy !
Reference: Couchbase : Create a large dataset using Twitter and Java from our JCG partner Tugdual Grall at the Tug’s Blog blog.
