lucene/contrib/benchmark/build.xml

<?xml version="1.0"?>
<project name="benchmark" default="default">

    <description>
        Lucene Benchmarking Contributions
    </description>

    <import file="../contrib-build.xml"/>
    <property name="working.dir" location="work"/>


    <target name="check-files">
        <available file="temp/news20.tar.gz" property="news20.exists"/>

        <available file="${working.dir}/20_newsgroup" property="news20.expanded"/>

        <available file="temp/reuters21578.tar.gz" property="reuters.exists"/>
        <available file="${working.dir}/reuters" property="reuters.expanded"/>
        <available file="${working.dir}/reuters-out" property="reuters.extracted"/>
        <available file="temp/20news-18828.tar.gz" property="20news-18828.exists"/>
        <available file="${working.dir}/20news-18828" property="20news-18828.expanded"/>
        <available file="${working.dir}/mini_newsgroups" property="mini.expanded"/>
        
        <available file="temp/enwiki-20070527-pages-articles.xml.bz2" property="enwiki.exists"/>
        <available file="temp/enwiki-20070527-pages-articles.xml" property="enwiki.expanded"/>
        <available file="${working.dir}/enwiki.txt" property="enwiki.extracted"/>

    </target>

    <target name="enwiki-files" depends="check-files">
        <mkdir dir="temp"/>
        <antcall target="get-enwiki"/>
        <antcall target="expand-enwiki"/>
    </target>

    <target name="get-enwiki" unless="enwiki.exists">
        <get src="http://people.apache.org/~gsingers/wikipedia/enwiki-20070527-pages-articles.xml.bz2"
             dest="temp/enwiki-20070527-pages-articles.xml.bz2"/>
    </target>

    <target name="expand-enwiki"  unless="enwiki.expanded">
        <bunzip2 src="temp/enwiki-20070527-pages-articles.xml.bz2" dest="temp"/>
    </target>

    <target name="get-news-20" unless="20news-18828.exists">
        <get src="http://www-2.cs.cmu.edu/afs/cs.cmu.edu/project/theo-20/www/data/news20.tar.gz"
             dest="temp/news20.tar.gz"/>

    </target>
    <target name="get-reuters" unless="reuters.exists">

        <get src="http://www.daviddlewis.com/resources/testcollections/reuters21578/reuters21578.tar.gz"
            dest="temp/reuters21578.tar.gz"/>
    </target>

    <target name="expand-news-20"  unless="news20.expanded">
        <gunzip src="temp/news20.tar.gz" dest="temp"/>
        <untar src="temp/news20.tar" dest="${working.dir}"/>
    </target>
    <target name="expand-reuters" unless="reuters.expanded">
        <gunzip src="temp/reuters21578.tar.gz" dest="temp"/>
        <mkdir dir="${working.dir}/reuters"/>
        <untar src="temp/reuters21578.tar" dest="${working.dir}/reuters"/>
        <delete >
            <fileset dir="${working.dir}/reuters">
                <include name="*.txt"/>
            </fileset>
        </delete>

    </target>
    <target name="extract-reuters" depends="check-files" unless="reuters.extracted">
        <mkdir dir="${working.dir}/reuters-out"/>
        <java classname="org.apache.lucene.benchmark.utils.ExtractReuters" maxmemory="1024M" fork="true">
            <classpath refid="run.classpath"/>
            <arg file="${working.dir}/reuters"/>
            <arg file="${working.dir}/reuters-out"/>
        </java>
    </target>
    <target name="get-20news-18828" unless="20news-18828.exists">
        <get src="http://people.csail.mit.edu/u/j/jrennie/public_html/20Newsgroups/20news-18828.tar.gz"
             dest="temp/20news-18828.tar.gz"/>

    </target>
    <target name="expand-20news-18828" unless="20news-18828.expanded">
        <gunzip src="temp/20news-18828.tar.gz" dest="temp"/>
        <untar src="temp/20news-18828.tar" dest="${working.dir}"/>
    </target>
    <target name="get-mini-news" unless="mini.exists">
        <get src="http://kdd.ics.uci.edu/databases/20newsgroups/mini_newsgroups.tar.gz"
             dest="temp/mini_newsgroups.tar.gz"/>
    </target>
    <target name="expand-mini-news" unless="mini.expanded">
        <gunzip src="temp/mini_newsgroups.tar.gz" dest="temp"/>
        <untar src="temp/mini_newsgroups.tar" dest="${working.dir}"/>
    </target>

    <target name="get-files" depends="check-files">
        <mkdir dir="temp"/>
        <antcall target="get-reuters"/>
        <antcall target="expand-reuters"/>
        <antcall target="extract-reuters"/>
    </target>

    <path id="classpath">
        <pathelement path="${common.dir}/build/classes/java"/>
        <pathelement path="${common.dir}/build/classes/demo"/>
        <pathelement path="${common.dir}/build/contrib/highlighter/classes/java"/>
    	<fileset dir="lib">
    		<include name="**/*.jar"/>
    	</fileset>
    </path>
    <path id="run.classpath">
        <path refid="classpath"/>
        <pathelement location="${build.dir}/classes/java"/>
        <pathelement location="${benchmark.ext.classpath}"/>
    </path>

    <property name="task.alg" location="conf/micro-standard.alg"/>
    <property name="task.mem" value="140M"/>

    <target name="run-task" depends="compile,check-files,get-files" 
     description="Run compound penalty perf test (optional: -Dtask.alg=your-algorithm-file -Dtask.mem=java-max-mem)">
        <echo>Working Directory: ${working.dir}</echo>
        <java classname="org.apache.lucene.benchmark.byTask.Benchmark" maxmemory="${task.mem}" fork="true">
            <classpath refid="run.classpath"/>
            <arg file="${task.alg}"/>
        </java>
    </target>

    <target name="enwiki" depends="compile,check-files,enwiki-files">
        <echo>Working Directory: ${working.dir}</echo>
        <java classname="org.apache.lucene.benchmark.byTask.Benchmark" maxmemory="1024M" fork="true">
            <assertions>
              <enable/>
            </assertions>
            <classpath refid="run.classpath"/>
            <arg file="conf/extractWikipedia.alg"/>
        </java>
    </target>

    <target name="compile-demo">
      <subant target="compile-demo">
         <fileset dir="${common.dir}" includes="build.xml"/>
      </subant>
    </target>
    <target name="compile-highlighter">
      <subant target="compile">
         <fileset dir="${common.dir}/contrib/highlighter" includes="build.xml"/>
      </subant>
    </target>

    <target name="init" depends="common.init,compile-demo, compile-highlighter,check-files"/>

    <!-- make sure online collections (reuters) are first downloaded -->
    <target name="test" depends="init,get-files">
      <antcall target="common.test" inheritRefs="true" />
    </target>
    
</project>