Initial commit

.gitignore

@@ -0,0 +1,15 @@
+.DS_Store
+.classpath
+.project
+.settings/
+.local.*
+build
+target
+src/gen
+build.sh
+*.iml
+*.dot
+*.hprof
+trash
+private
+junit*.properties

README.md

@@ -1,4 +1,4 @@
 text-similarity
 ===============
 
-Source code for blog post series on text features for similarity calculation
+Source code for blog post series on text features for similarity calculation.

build.properties

@@ -0,0 +1,37 @@
+#
+# Copyright (c) 2013 Scale Unlimited
+# 
+# All rights reserved.
+#
+name=${ant.project.name}
+version=1.0-SNAPSHOT
+
+jar.name=${ant.project.name}-${version}.jar
+job.name=${ant.project.name}-job-${version}.jar
+
+main.src.dir=src/main/java
+test.src.dir=src/test/java
+
+main.res.dir=src/main/resources
+test.res.dir=src/test/resources
+
+lib.dir=lib
+
+build.dir=build
+build.dir.main-classes=${build.dir}/classes-main
+build.dir.test-classes=${build.dir}/classes-test
+
+build.dir.main-classes-eclipse=${build.dir}/classes-main-eclipse
+build.dir.test-classes-eclipse=${build.dir}/classes-test-eclipse
+
+build.dir.test-reports=${build.dir}/test
+build.dir.dist=${build.dir}/${name}-dist
+build.release.file=${build.dir}/${name}.tgz
+
+javac.debug=on
+javac.optimize=on
+javac.deprecation=off
+javac.version=1.6
+javac.args=
+javac.args.warnings=-Xlint:none
+build.encoding=UTF-8

build.xml

@@ -0,0 +1,295 @@
+<!--
+Copyright (c) 2013 Scale Unlimited
+
+All rights reserved.
+-->
+
+<project name="text-similarity" default="test">
+
+    <property name="root.dir" value="${basedir}" />
+    <property file="${root.dir}/build.properties" />
+	<property name="target.dir" location="target"/>
+
+    <!-- ================================================================== -->
+    <!-- General cleaning sources                                           -->
+    <!-- ================================================================== -->
+
+    <target name="clean" description="--> clean the project">
+        <echo>cleaning ${ant.project.name}</echo>
+        <delete includeemptydirs="true" failonerror="false">
+            <fileset dir="${build.dir}" excludes="classes-*-eclipse/" />
+            <fileset dir="${target.dir}" />
+        </delete>
+    </target>
+
+	<!-- ================================================================== -->
+	<!-- Maven                                                              -->
+	<!-- ================================================================== -->
+
+	<property name="maven-artifact-ant.bootstrap.jar"
+			  location="${target.dir}/maven-ant-tasks-2.1.3.jar"/>
+
+    <target name="check-get-maven-artifact-ant">
+        <available property="get-maven-artifact-ant-done"
+				   file="${maven-artifact-ant.bootstrap.jar}"/>
+    </target>
+
+    <target name="get-maven-artifact-ant"
+		    depends="check-get-maven-artifact-ant"
+		    unless="get-maven-artifact-ant-done">
+        <mkdir dir="${target.dir}"/>
+        <get src="http://repo1.maven.org/maven2/org/apache/maven/maven-ant-tasks/2.1.3/maven-ant-tasks-2.1.3.jar"
+		     dest="${maven-artifact-ant.bootstrap.jar}"/>
+    </target>
+
+	<target name="mvn-init" 
+	        unless="mvn.compile.classpath" 
+	        depends="get-maven-artifact-ant"
+	        xmlns:artifact="urn:maven-artifact-ant">
+		<path id="maven.ant.tasks.classpath" path="${maven-artifact-ant.bootstrap.jar}" />
+
+        <typedef resource="org/apache/maven/artifact/ant/antlib.xml"
+	             uri="urn:maven-artifact-ant"
+                 classpathref="maven.ant.tasks.classpath"/>
+
+	    <condition property="maven.repo.local"
+	               value="${maven.repo.local}"
+	               else="${user.home}/.m2/repository">
+            <isset property="maven.repo.local"/>
+	    </condition>
+
+	    <artifact:localRepository id="local.repository" path="${maven.repo.local}"/>
+	    <artifact:pom file="pom.xml" id="maven.project"/>
+	    <artifact:dependencies pathId="compile.classpath" filesetId="compile.fileset" useScope="compile">
+	        <pom refid="maven.project"/>
+	        <localRepository refid="local.repository"/>
+	    </artifact:dependencies>
+	    <artifact:dependencies pathId="test.classpath" filesetId="test.fileset" useScope="test">
+	        <pom refid="maven.project"/>
+	        <localRepository refid="local.repository"/>
+	    </artifact:dependencies>
+	    <artifact:dependencies pathId="runtime.classpath" filesetId="runtime.fileset" useScope="runtime">
+	        <pom refid="maven.project"/>
+	        <localRepository refid="local.repository"/>
+	    </artifact:dependencies>
+	</target>
+
+    <!-- ================================================================== -->
+    <!-- Build sources                                                      -->
+    <!-- ================================================================== -->
+
+    <target name="compile" 
+        depends="mvn-init"
+		description="--> compile main classes">
+        <mkdir dir="${build.dir.main-classes}" />
+        <javac encoding="${build.encoding}"
+			   srcdir="${main.src.dir}"
+			   includes="**/*.java"
+               destdir="${build.dir.main-classes}"
+               debug="${javac.debug}"
+               optimize="${javac.optimize}"
+               target="${javac.version}"
+               source="${javac.version}"
+               includeantruntime="false"
+               deprecation="${javac.deprecation}">
+            <compilerarg line="${javac.args} ${javac.args.warnings}" />
+            <classpath refid="compile.classpath" />
+        </javac>
+    </target>
+
+
+    <!-- ================================================================== -->
+    <!-- Unit Tests                                                         -->
+    <!-- ================================================================== -->
+
+    <target name="compile-test" depends="compile">
+        <echo>*** Building Unit Tests Sources ***</echo>
+        <mkdir dir="${build.dir.test-classes}" />
+        <path id="test.path">
+            <pathelement location="${build.dir.main-classes}" />
+        </path>
+
+        <javac encoding="${build.encoding}"
+               srcdir="${test.src.dir}"
+               includes="**/*.java"
+               destdir="${build.dir.test-classes}"
+               debug="${javac.debug}"
+               optimize="${javac.optimize}"
+               target="${javac.version}"
+               source="${javac.version}"
+               includeantruntime="false"
+               deprecation="${javac.deprecation}">
+            <compilerarg line="${javac.args} ${javac.args.warnings}" />
+            <classpath refid="test.classpath" />
+            <classpath refid="test.path" />
+        </javac>
+    </target>
+
+    <target name="test" depends="compile-test" description="--> run unit tests">
+        <delete dir="${build.dir.test-reports}" />
+        <mkdir dir="${build.dir.test-reports}" />
+
+        <junit showoutput="false"
+               printsummary="yes"
+               haltonfailure="no"
+               fork="yes"
+               maxmemory="256m"
+               dir="${basedir}"
+               errorProperty="tests.failed"
+               failureProperty="tests.failed">
+			<sysproperty key="java.security.krb5.realm" value="" />
+			<sysproperty key="java.security.krb5.kdc" value="" />
+	        <classpath>
+	            <pathelement location="${build.dir.main-classes}" />
+	            <pathelement location="${build.dir.test-classes}" />
+	            <pathelement location="${test.res.dir}" />
+	            <pathelement location="${main.res.dir}" />
+	            <path refid="test.classpath" />
+	        </classpath>
+	        <formatter type="plain" />
+	        <batchtest fork="yes" todir="${build.dir.test-reports}" unless="testcase">
+	            <fileset dir="${test.src.dir}">
+	                <include name="**/*Test.java" unless="testcase" />
+	                <exclude name="**/Abstract*.java" unless="testcase" />
+	                <include name="${testcase}" if="testcase" />
+	            </fileset>
+	        </batchtest>
+	        <batchtest fork="yes" todir="${build.dir.test-reports}" if="testcase">
+	            <fileset dir="${test.src.dir}" includes="**/${testcase}.java" />
+	        </batchtest>
+        </junit>
+        <fail if="tests.failed">Tests failed!</fail>
+    </target>
+
+	<!-- ================================================================== -->
+	<!-- Distribution                                                       -->
+	<!-- ================================================================== -->
+
+	<target name="dist"
+	        depends="test"
+	        description="--> create a tarball distribution">
+
+		<delete dir="${build.dir.dist}" />
+
+		<!-- create target directory -->
+		<mkdir dir="${build.dir.dist}" />
+		<mkdir dir="${build.dir.dist}/doc" />
+		<mkdir dir="${build.dir.dist}/lib" />
+		<mkdir dir="${build.dir.dist}/src" />
+
+		<!-- copy libs -->
+		<copy todir="${build.dir.dist}/lib" flatten="true">
+			<fileset dir="${basedir}/lib" />
+			<path refId="compile.classpath" />
+			<path refId="runtime.classpath" />
+			<path refId="test.classpath" />
+		</copy>
+
+		<!-- copy src -->
+		<copy todir="${build.dir.dist}/src">
+			<fileset dir="${basedir}/src" />
+		</copy>
+
+		<!-- copy documents -->
+		<copy todir="${build.dir.dist}/doc">
+			<fileset dir="${basedir}/doc" />
+		</copy>
+
+		<!-- copy the build files -->
+		<copy todir="${build.dir.dist}">
+			<fileset file="${basedir}/build.xml" />
+		</copy>
+
+		<!-- copy the build properties -->
+		<copy todir="${build.dir.dist}">
+			<fileset file="${basedir}/build.properties" />
+		</copy>
+
+		<tar longfile="gnu" compression="gzip" destfile="${build.release.file}">
+			<tarfileset dir="${build.dir.dist}" prefix="${name}/">
+			    <include name="**" />
+			</tarfileset>
+		</tar>
+	</target>
+
+
+    <!-- ================================================================== -->
+    <!-- Hadoop job jar                                                     -->
+    <!-- ================================================================== -->
+
+    <target name="job"
+            depends="compile"
+            description="--> create a Hadoop ready jar with all dependencies">
+
+        <!-- Make sure lib/ dir starts out empty, so we don't get multiple
+		     copies of jars with slightly different versions
+		 -->
+        <delete dir="${build.dir}/lib" />
+        <mkdir dir="${build.dir}/lib" />
+
+        <copy todir="${build.dir}/lib" flatten="true">
+            <path refid="runtime.classpath" />
+        </copy>
+
+        <jar destfile="${build.dir}/${job.name}" compress="true">
+            <fileset dir="${build.dir.main-classes}" />
+            <fileset dir="${main.res.dir}" />
+            <fileset dir="${build.dir}" includes="lib/" />
+        </jar>
+    </target>
+
+
+    <!-- ================================================================== -->
+    <!-- Generating eclipse file                                            -->
+    <!-- ================================================================== -->
+
+	<target name="eclipse"
+		depends="mvn-init, clean-eclipse"
+		description="--> create the Eclipse project files">
+
+		<taskdef name="eclipse"
+		         classname="prantl.ant.eclipse.EclipseTask"
+		         classpathref="compile.classpath" />
+		<mkdir dir="${build.dir.main-classes-eclipse}" />
+		<mkdir dir="${build.dir.test-classes-eclipse}" />
+
+		<eclipse>
+			<settings>
+				<jdtcore compilercompliance="6.0" />
+				<resources encoding="UTF-8" />
+			</settings>
+			<project name="${ant.project.name}" />
+			<classpath>
+				<container path="org.eclipse.jdt.launching.JRE_CONTAINER/org.eclipse.jdt.internal.debug.ui.launcher.StandardVMType/JavaSE-1.6" />
+
+				<source path="${main.src.dir}"
+				        output="${build.dir.main-classes-eclipse}" />
+
+				<!-- TODO use build.properties (with more consistent names) for these paths -->
+				<source path="${basedir}/src/main/resources"
+				        output="${build.dir.main-classes-eclipse}" />
+				<source path="${basedir}/src/test/java"
+				        output="${build.dir.test-classes-eclipse}" />
+				<source path="${basedir}/src/test/resources"
+				        output="${build.dir.test-classes-eclipse}" />
+
+				<output path="${build.dir.main-classes-eclipse}" />
+				<library pathref="test.classpath" exported="false" />
+			</classpath>
+		</eclipse>
+
+		<concat destfile="${root.dir}/.settings/org.eclipse.jdt.core.prefs"
+		        append="true">
+			<filelist dir="${root.dir}/doc/" files="eclipse-formatter.properties" />
+		</concat>
+	</target>
+
+    <target name="clean-eclipse" description="--> clean the Eclipse project files">
+        <delete file=".classpath" />
+        <delete file=".eclipse" />
+        <delete file=".project" />
+        <delete dir=".settings" />
+    </target>
+
+
+</project>