<?xml version="1.0"?>

<project name="LingPipe Tutorial - Language ID" default="compile" basedir=".">

  <property name="data.dir"
            location="e:\data\"/>
  <property name="corpus.dir"
            location="${data.dir}/leipzig/munged"/>
  <property name="model.file"
            location="../../models/langid-leipzig.classifier"/>
  <property name="training.size"
            value="2000000"/>
  <property name="nGram"
            value="5"/>
  <property name="token.count.min"
            value="20"/>
  <property name="test.size"
            value="32"/>
  <property name="test.num"
            value="10000"/>

  <path id="classpath.standard">
    <pathelement location="languageId.jar"/>
    <pathelement location="../../../lingpipe-4.1.0.jar"/>
  </path>

  <target name="clean"
          description="Clean build and distribution directories.">
    <delete dir="build"/>
    <delete file="languageId.jar"/>
  </target>

  <target name="compile"
          depends=""
          description="Compile the src code into the build directory.">
    <mkdir dir="build/classes"/>
    <javac includeAntRuntime="false"   optimize="on"
           debug="yes"
           debuglevel="source,lines,vars"
           destdir="build/classes">
      <compilerarg value="-Xlint:all"/>
      <classpath refid="classpath.standard"/>
      <src path="src/"/>
    </javac>
  </target>

  <target name="jar"
          depends="compile">
    <jar destfile="languageId.jar">
      <fileset dir="build/classes"
               includes="**/*.class"/>
    </jar>
  </target>


  <target name="train"
          depends="jar"
          description="Train language identifier">
    <java classname="TrainLanguageId"
          fork="true">
      <classpath refid="classpath.standard"/>
      <arg value="${corpus.dir}"/>
      <arg value="${model.file}"/>
      <arg value="${nGram}"/>
      <arg value="${training.size}"/>
      <arg value="${token.count.min}"/>
    </java>
  </target>

  <target name="eval"
          depends="jar">
    <java classname="EvalLanguageId"
          fork="true">
      <classpath refid="classpath.standard"/>
      <arg value="${corpus.dir}"/>
      <arg value="${model.file}"/>
      <arg value="${training.size}"/>
      <arg value="${test.size}"/>
      <arg value="${test.num}"/>
    </java>
  </target>

  <target name="run"
          depends="jar">
    <java classname="RunLanguageId"
          fork="true">
      <classpath refid="classpath.standard"/>
      <arg value="${model.file}"/>
      <arg value="Per poder jutjar l'efectivitat d'una novetat Acs imprescindible deixar"/>
      <arg value="Michael"/>
      <arg value="Maria"/>
    </java>
  </target>

<target name="unpack">
  <echo message="Please be patient, this is lots of data."/>
  <mkdir dir="${dir.unpacked}/"/>
  <unzip dest="${dir.unpacked}/">
    <fileset dir="${dir.dist}">
      <include name="**/*.zip"/>
    </fileset>
  </unzip>
</target>

<target name="munge"
        depends="jar">
  <mkdir dir="${dir.munged}"/>
  <java classname="Munge"
        classpath="build/classes"
        fork="true">
    <arg value="${dir.unpacked}"/>
    <arg value="${dir.munged}"/>
  </java>
</target>



</project>
