SOLR-5201 - AnalysisEngines are now created in the factory and passed to the processors with a JCas pool

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1520239 13f79535-47bb-0310-9956-ffa450edef68

solr/contrib/uima/src/java/org/apache/solr/uima/processor/UIMAUpdateRequestProcessor.java

@@ -17,49 +17,49 @@ package org.apache.solr.uima.processor;
  * limitations under the License.
  */
 
+import java.io.IOException;
+import java.util.Map;
+import java.util.Map.Entry;
+
 import org.apache.solr.common.SolrException;
 import org.apache.solr.common.SolrException.ErrorCode;
 import org.apache.solr.common.SolrInputDocument;
 import org.apache.solr.schema.SchemaField;
 import org.apache.solr.uima.processor.SolrUIMAConfiguration.MapField;
-import org.apache.lucene.analysis.uima.ae.AEProvider;
-import org.apache.lucene.analysis.uima.ae.AEProviderFactory;
 import org.apache.solr.update.AddUpdateCommand;
 import org.apache.solr.update.processor.UpdateRequestProcessor;
 import org.apache.uima.analysis_engine.AnalysisEngine;
 import org.apache.uima.analysis_engine.AnalysisEngineProcessException;
 import org.apache.uima.jcas.JCas;
 import org.apache.uima.resource.ResourceInitializationException;
+import org.apache.uima.util.JCasPool;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
-import java.io.IOException;
-import java.util.Map;
-
 /**
  * Update document(s) to be indexed with UIMA extracted information
- *
+ * 
  */
 public class UIMAUpdateRequestProcessor extends UpdateRequestProcessor {
-
+  
-  private final Logger log = LoggerFactory.getLogger(UIMAUpdateRequestProcessor.class);
+  private final Logger log = LoggerFactory
-
+      .getLogger(UIMAUpdateRequestProcessor.class);
+  
   SolrUIMAConfiguration solrUIMAConfiguration;
-
+  
-  private AEProvider aeProvider;
+  private AnalysisEngine ae;
-
+  
-  public UIMAUpdateRequestProcessor(UpdateRequestProcessor next, String coreName,
+  private JCasPool pool;
-      SolrUIMAConfiguration config) {
+  
+  public UIMAUpdateRequestProcessor(UpdateRequestProcessor next,
+      String coreName, SolrUIMAConfiguration config, AnalysisEngine ae,
+      JCasPool pool) {
     super(next);
-    initialize(coreName, config);
+    this.ae = ae;
-  }
+    this.pool = pool;
-
-  private void initialize(String coreName, SolrUIMAConfiguration config) {
     solrUIMAConfiguration = config;
-    aeProvider = AEProviderFactory.getInstance().getAEProvider(coreName,
-            solrUIMAConfiguration.getAePath(), solrUIMAConfiguration.getRuntimeParameters());
   }
-
+  
   @Override
   public void processAdd(AddUpdateCommand cmd) throws IOException {
     String text = null;
@@ -72,54 +72,66 @@ public class UIMAUpdateRequestProcessor extends UpdateRequestProcessor {
       for (String currentText : texts) {
         text = currentText;
         if (text != null && text.length() > 0) {
-          /* process the text value */
+          /* create a JCas which contain the text to analyze */
-          JCas jcas = processText(text);
+          JCas jcas = pool.getJCas(0);
+          try {
+            /* process the text value */
+            processText(text, jcas);
 
-          UIMAToSolrMapper uimaToSolrMapper = new UIMAToSolrMapper(solrInputDocument, jcas);
+            UIMAToSolrMapper uimaToSolrMapper = new UIMAToSolrMapper(
-          /* get field mapping from config */
+                solrInputDocument, jcas);
-          Map<String, Map<String, MapField>> typesAndFeaturesFieldsMap = solrUIMAConfiguration
+            /* get field mapping from config */
-              .getTypesFeaturesFieldsMapping();
+            Map<String,Map<String,MapField>> typesAndFeaturesFieldsMap = solrUIMAConfiguration
-          /* map type features on fields */
+                .getTypesFeaturesFieldsMapping();
-          for (String typeFQN : typesAndFeaturesFieldsMap.keySet()) {
+            /* map type features on fields */
-            uimaToSolrMapper.map(typeFQN, typesAndFeaturesFieldsMap.get(typeFQN));
+            for (Entry<String,Map<String,MapField>> entry : typesAndFeaturesFieldsMap
+                .entrySet()) {
+              uimaToSolrMapper.map(entry.getKey(), entry.getValue());
+            }
+          } finally {
+            pool.releaseJCas(jcas);
           }
         }
       }
     } catch (Exception e) {
       String logField = solrUIMAConfiguration.getLogField();
-      if(logField == null){
+      if (logField == null) {
-        SchemaField uniqueKeyField = cmd.getReq().getSchema().getUniqueKeyField();
+        SchemaField uniqueKeyField = cmd.getReq().getSchema()
-        if(uniqueKeyField != null){
+            .getUniqueKeyField();
+        if (uniqueKeyField != null) {
           logField = uniqueKeyField.getName();
         }
       }
-      String optionalFieldInfo = logField == null ? "." :
+      String optionalFieldInfo = logField == null ? "."
-        new StringBuilder(". ").append(logField).append("=")
+          : new StringBuilder(". ")
-        .append((String)cmd.getSolrInputDocument().getField(logField).getValue())
+              .append(logField)
-        .append(", ").toString();
+              .append("=")
+              .append(
+                  (String) cmd.getSolrInputDocument().getField(logField)
+                      .getValue()).append(", ").toString();
       int len;
       String debugString;
       if (text != null && text.length() > 0) {
         len = Math.min(text.length(), 100);
-        debugString = new StringBuilder(" text=\"").append(text.substring(0, len)).append("...\"").toString();
+        debugString = new StringBuilder(" text=\"")
-      }
+            .append(text.substring(0, len)).append("...\"").toString();
-      else {
+      } else {
         debugString = " null text";
       }
       if (solrUIMAConfiguration.isIgnoreErrors()) {
-        log.warn("skip the text processing due to {}",new StringBuilder()
+        log.warn(
-          .append(e.getLocalizedMessage()).append(optionalFieldInfo)
+            "skip the text processing due to {}",
-          .append(debugString));
+            new StringBuilder().append(e.getLocalizedMessage())
+                .append(optionalFieldInfo).append(debugString));
       } else {
-        throw new SolrException(ErrorCode.SERVER_ERROR,
+        throw new SolrException(ErrorCode.SERVER_ERROR, new StringBuilder(
-            new StringBuilder("processing error ")
+            "processing error ").append(e.getLocalizedMessage())
-              .append(e.getLocalizedMessage()).append(optionalFieldInfo)
+            .append(optionalFieldInfo).append(debugString).toString(), e);
-              .append(debugString).toString(), e);
       }
     }
     super.processAdd(cmd);
   }
-
+  
   /*
    * get the texts to analyze from the corresponding fields
    */
@@ -130,30 +142,31 @@ public class UIMAUpdateRequestProcessor extends UpdateRequestProcessor {
     if (merge) {
       StringBuilder unifiedText = new StringBuilder("");
       for (String aFieldsToAnalyze : fieldsToAnalyze) {
-        unifiedText.append(String.valueOf(solrInputDocument.getFieldValue(aFieldsToAnalyze)));
+        unifiedText.append(String.valueOf(solrInputDocument
+            .getFieldValue(aFieldsToAnalyze)));
       }
       textVals = new String[1];
       textVals[0] = unifiedText.toString();
     } else {
       textVals = new String[fieldsToAnalyze.length];
       for (int i = 0; i < fieldsToAnalyze.length; i++) {
-        textVals[i] = String.valueOf(solrInputDocument.getFieldValue(fieldsToAnalyze[i]));
+        textVals[i] = String.valueOf(solrInputDocument
+            .getFieldValue(fieldsToAnalyze[i]));
       }
     }
     return textVals;
   }
-
+  
-  /* process a field value executing UIMA the CAS containing it as document text */
+  /*
-  private JCas processText(String textFieldValue) throws ResourceInitializationException,
+   * process a field value executing UIMA on the JCas containing it as document
-          AnalysisEngineProcessException {
+   * text
+   */
+  private void processText(String textFieldValue, JCas jcas)
+      throws ResourceInitializationException, AnalysisEngineProcessException {
     if (log.isDebugEnabled()) {
       log.debug("Analyzing text");
     }
-    /* get the UIMA analysis engine */
-    AnalysisEngine ae = aeProvider.getAE();
 
-    /* create a JCas which contain the text to analyze */
-    JCas jcas = ae.newJCas();
     jcas.setDocumentText(textFieldValue);
 
     /* perform analysis on text field */
@@ -161,7 +174,6 @@ public class UIMAUpdateRequestProcessor extends UpdateRequestProcessor {
     if (log.isDebugEnabled()) {
       log.debug("Text processing completed");
     }
-    return jcas;
   }
-
+  
 }

solr/contrib/uima/src/java/org/apache/solr/uima/processor/UIMAUpdateRequestProcessorFactory.java

@@ -17,20 +17,29 @@ package org.apache.solr.uima.processor;
  * limitations under the License.
  */
 
+import org.apache.lucene.analysis.uima.ae.AEProvider;
+import org.apache.lucene.analysis.uima.ae.AEProviderFactory;
+import org.apache.solr.common.SolrException;
 import org.apache.solr.common.util.NamedList;
 import org.apache.solr.request.SolrQueryRequest;
 import org.apache.solr.response.SolrQueryResponse;
 import org.apache.solr.update.processor.UpdateRequestProcessor;
 import org.apache.solr.update.processor.UpdateRequestProcessorFactory;
+import org.apache.uima.analysis_engine.AnalysisEngine;
+import org.apache.uima.resource.ResourceInitializationException;
+import org.apache.uima.util.JCasPool;
 
 /**
  * Factory for {@link UIMAUpdateRequestProcessor}
  * 
- *
+ * 
  */
-public class UIMAUpdateRequestProcessorFactory extends UpdateRequestProcessorFactory {
+public class UIMAUpdateRequestProcessorFactory extends
+    UpdateRequestProcessorFactory {
 
   private NamedList<Object> args;
+  private AnalysisEngine ae;
+  private JCasPool pool;
 
   @SuppressWarnings("unchecked")
   @Override
@@ -39,10 +48,26 @@ public class UIMAUpdateRequestProcessorFactory extends UpdateRequestProcessorFac
   }
 
   @Override
-  public UpdateRequestProcessor getInstance(SolrQueryRequest req, SolrQueryResponse rsp,
+  public UpdateRequestProcessor getInstance(SolrQueryRequest req,
-          UpdateRequestProcessor next) {
+      SolrQueryResponse rsp, UpdateRequestProcessor next) {
+    SolrUIMAConfiguration configuration = new SolrUIMAConfigurationReader(args)
+        .readSolrUIMAConfiguration();
+    synchronized (this) {
+      if (ae == null && pool == null) {
+        AEProvider aeProvider = AEProviderFactory.getInstance().getAEProvider(
+            req.getCore().getName(), configuration.getAePath(),
+            configuration.getRuntimeParameters());
+        try {
+          ae = aeProvider.getAE();
+          pool = new JCasPool(10, ae);
+        } catch (ResourceInitializationException e) {
+          throw new SolrException(SolrException.ErrorCode.SERVER_ERROR, e);
+        }
+      }
+    }
+    
     return new UIMAUpdateRequestProcessor(next, req.getCore().getName(),
-            new SolrUIMAConfigurationReader(args).readSolrUIMAConfiguration());
+        configuration, ae, pool);
   }
-
+  
 }

solr/contrib/uima/src/test/org/apache/solr/uima/processor/UIMAUpdateRequestProcessorTest.java

@@ -93,7 +93,6 @@ public class UIMAUpdateRequestProcessorTest extends SolrTestCaseJ4 {
 
   @Test
   public void testProcessing() throws Exception {
-
     addDoc("uima", adoc(
             "id",
             "2312312321312",
@@ -185,6 +184,13 @@ public class UIMAUpdateRequestProcessorTest extends SolrTestCaseJ4 {
     }
   }
 
+  @Test
+  public void testMultiplierProcessing() throws Exception {
+    for (int i = 0; i < RANDOM_MULTIPLIER; i++) {
+      testProcessing();
+    }
+  }
+
   private void addDoc(String chain, String doc) throws Exception {
     Map<String, String[]> params = new HashMap<String, String[]>();
     params.put(UpdateParams.UPDATE_CHAIN, new String[] { chain });