#24 add decider and purger batchlet before the doc production chunk

Decider and purger batchlet provide a before-production enhancement. User can now have possibility to purge all index before started. However, this commit provides only the working model, method inside the batchlet is not implemented yet.

us-address/src/main/java/io/github/mincongh/batch/BeforeIndexDecider.java

@@ -0,0 +1,44 @@
+package io.github.mincongh.batch;
+
+import javax.batch.api.BatchProperty;
+import javax.batch.api.Decider;
+import javax.batch.runtime.StepExecution;
+import javax.inject.Inject;
+import javax.inject.Named;
+
+/**
+ * Decider decides the next step-execution before the start of index chunk. If
+ * user requires a index purge, then the next step should be a purge, else, 
+ * the next step will be directly the index chunk. Index purge use 
+ * IndexPurgerBatchlet. 
+ * 
+ * @author Mincong HUANG
+ */
+@Named
+public class BeforeIndexDecider implements Decider {
+
+    @Inject @BatchProperty
+    private Boolean purgeAtStart;
+    private final String PURGE_IDX = "purgeIndex";
+    private final String PRODUCE_DOC = "produceLuceneDoc";
+
+    /**
+     * Decide the next step
+     * 
+     * @param executions not used for the moment.
+     */
+    @Override
+    public String decide(StepExecution[] executions) throws Exception {
+
+        String nextStep = purgeAtStart ? PURGE_IDX : PRODUCE_DOC;
+
+        for (StepExecution se : executions) {
+            System.out.println(se.getStepName() + " "
+                    + se.getBatchStatus().toString() + " "
+                    + se.getExitStatus()
+            );
+        }
+
+        return nextStep;
+    }
+}

us-address/src/main/java/io/github/mincongh/batch/IndexPurgerBatchlet.java

@@ -0,0 +1,22 @@
+package io.github.mincongh.batch;
+
+import javax.batch.api.Batchlet;
+import javax.batch.runtime.BatchStatus;
+import javax.inject.Named;
+
+@Named
+public class IndexPurgerBatchlet implements Batchlet {
+
+    @Override
+    public String process() throws Exception {
+
+        System.out.println("purging entities ...");
+
+        return BatchStatus.COMPLETED.toString();
+    }
+
+    @Override
+    public void stop() throws Exception {
+        // TODO Auto-generated method stub
+    }
+}

us-address/src/main/java/io/github/mincongh/session/BatchSession.java

@@ -68,14 +68,19 @@ public BatchSession() {
      */
     @Asynchronous
     public void massIndex() {
+
         Long start = System.currentTimeMillis();
+
         Properties jobParams = new Properties();
         jobParams.setProperty("fetchSize", "200000");
         jobParams.setProperty("arrayCapacity", "1000");
         jobParams.setProperty("maxResults", "1000000");
         jobParams.setProperty("partitionCapacity", "250");
         jobParams.setProperty("threads", "4");
+        jobParams.setProperty("purgeAtStart", String.valueOf(true));
         Long executionId = jobOperator.start("mass-index", jobParams);
+
+        // calculate the performance
         JobExecution execution = jobOperator.getJobExecution(executionId);
         int i = 0; 
         while (!execution.getBatchStatus().equals(BatchStatus.COMPLETED) && i < 200) {

us-address/src/main/resources/META-INF/batch-jobs/mass-index.xml

@@ -3,7 +3,7 @@
     xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
     xsi:schemaLocation="http://xmlns.jcp.org/xml/ns/javaee http://xmlns.jcp.org/xml/ns/javaee/jobXML_1_0.xsd"
     version="1.0">
-  <step id="loadId" next="produceLuceneDoc">
+  <step id="loadId" next="decisionBeforeIndex">
     <batchlet ref="idProducerBatchlet">
       <properties>
         <!-- fetchSize - The number of rows returned in one chunk in criteria.
@@ -21,6 +21,16 @@
       </properties>
     </batchlet>
   </step>
+  <decision id="decisionBeforeIndex" ref="beforeIndexDecider">
+    <properties>
+      <property name="purgeAtStart" value="#{jobParameters['purgeAtStart']}?:false"/>
+    </properties>
+    <next on="purgeIndex" to="purgeIndex"/>
+    <next on="produceLuceneDoc" to="produceLuceneDoc"/>
+  </decision>
+  <step id="purgeIndex" next="produceLuceneDoc">
+    <batchlet ref="indexPurgerBatchlet"/>
+  </step>
   <step id="produceLuceneDoc">
     <chunk item-count="3">
       <reader ref="batchItemReader"/>