Installing the java api

This page describes how to start interacting with the Wikipedia Miner toolkit from within Java code. It assumes you have already obtained the data (both XML dump and corresponding CSV summaries).

##Unpack data

Place the uncompressed XML file and all of the extracted csv summaries into a single directory

##Create configuration file

Copy the wikipedia-template.xml file from the configs directory included in the toolkit. It is likely that you will reuse this file every time you interact with Wikipedia miner, so name it something memorable and put it somewhere safe (the configs directory is a good place).

Fill in the file with appropriate details. The file is heavily commented, so it should be self explanatory. At a minimum, you should at least enter the fields that are mandatory: langCode (e.g. "en") and databaseDirectory (e.g. "db", which will be created in the next step). Also fill in the dataDirectory field to refer to the folder containing an uncompressed wikipedia dump and extracted summaries.

##Build the berkeley database

Edit the ant/build.properties file to include a reference to the configuration file you just created. Add a line like this:

wiki.conf=path/to/conf/file

The path needs to be fully qualified, or relative to the root folder of the wikipedia miner project.

Now run the ant target build-database

ant build-database

If you did everything right, you will get lots of status updates, and the whole process should be over in an hour or so depending on the size of the dump and the specs of your computer.

##Start playing around

To verify that everything is working ok, try creating a new java project that we can play around with.

Start by using the ant target package

    ant package

This will create the file bin/jar/wikipedia-miner.jar

Version 1.2 contains a bugthat prevents the jar file from being created correctly. It can be fixed by replacing the ant/build.xml with the version found here This patch will be included in the next version.

Using your preferred programming environment (Eclipse, emacs, butterflies, whatever), create a new project, whose classpath includes wikipedia-miner.jar and all of the libraries in the lib directory of wikipedia miner.

Then try creating the WikipediaDefiner class:

import java.io.File;

import org.wikipedia.miner.model.Article;
import org.wikipedia.miner.model.Wikipedia;
import org.wikipedia.miner.util.WikipediaConfiguration;


public class WikipediaDefiner {

    public static void main(String args[]) throws Exception {
		
        WikipediaConfiguration conf = new WikipediaConfiguration(new File("path/to/conf")) ;
			
        Wikipedia wikipedia = new Wikipedia(conf, false) ;
	    
        Article article = wikipedia.getArticleByTitle("Wikipedia") ;
	    
        System.out.println(article.getSentenceMarkup(0)) ;
	    
        wikipedia.close() ;
    }
}

If everything is working well, you should be able to run this and print out a pithy definition of what Wikipedia is.

Don't be worried if this first call takes a while to run. It seems to take a while for the database behind the toolkit to iron out once it is first built. It will soon be running quickly.

Check out Using the Java API for more ideas, and the JavaDoc for the gritty details.

Java API is very much useful Java Training in Chennai