docs/attachments/LUCENE-1773/LUCENE-1773.patch

Index: contrib/benchmark/conf/highlight-vs-vector-highlight.alg
===================================================================
--- contrib/benchmark/conf/highlight-vs-vector-highlight.alg	(revision 0)
+++ contrib/benchmark/conf/highlight-vs-vector-highlight.alg	(revision 0)
@@ -0,0 +1,78 @@
+#/**
+# * Licensed to the Apache Software Foundation (ASF) under one or more
+# * contributor license agreements.  See the NOTICE file distributed with
+# * this work for additional information regarding copyright ownership.
+# * The ASF licenses this file to You under the Apache License, Version 2.0
+# * (the "License"); you may not use this file except in compliance with
+# * the License.  You may obtain a copy of the License at
+# *
+# *     http://www.apache.org/licenses/LICENSE-2.0
+# *
+# * Unless required by applicable law or agreed to in writing, software
+# * distributed under the License is distributed on an "AS IS" BASIS,
+# * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# * See the License for the specific language governing permissions and
+# * limitations under the License.
+# */
+# -------------------------------------------------------------------------------------
+
+ram.flush.mb=flush:32:32
+compound=cmpnd:true:false
+
+analyzer=org.apache.lucene.analysis.standard.StandardAnalyzer
+directory=FSDirectory
+
+doc.stored=true
+doc.tokenized=true
+doc.term.vector=true
+doc.term.vector.offsets=true
+doc.term.vector.positions=true
+log.step=2000
+
+docs.dir=reuters-out
+
+content.source=org.apache.lucene.benchmark.byTask.feeds.EnwikiContentSource
+docs.file=temp/enwiki-20070527-pages-articles.xml
+
+# Use LUCENE-1770 WikipediaQueryMaker
+query.maker=org.apache.lucene.benchmark.byTask.feeds.WikipediaQueryMaker
+#query.maker=org.apache.lucene.benchmark.byTask.feeds.ReutersQueryMaker
+
+# task at this depth or less would print when they start
+task.max.depth.log=2
+
+log.queries=true
+# -------------------------------------------------------------------------------------
+{ "Populate"
+        CreateIndex
+        { "MAddDocs" AddDoc } : 20000
+        Optimize
+        CloseIndex
+}
+{
+        OpenReader
+          { "WarmTV" SearchTravRetVectorHighlight(maxFrags[3],fields[body]) > : 100
+        CloseReader
+}
+{
+	"Rounds"
+
+        ResetSystemSoft
+
+        OpenReader
+          { "SearchVecHlgtSameRdr" SearchTravRetVectorHighlight(maxFrags[3],fields[body]) > : 200
+        CloseReader
+
+        ResetSystemSoft
+
+        OpenReader
+          { "SearchHlgtSameRdr" SearchTravRetHighlight(maxFrags[3],fields[body]) > : 200
+        CloseReader
+
+        RepSumByPref Search
+
+        NewRound
+} : 4
+
+RepSumByNameRound
+RepSumByName
Index: contrib/benchmark/conf/vector-highlight-profile.alg
===================================================================
--- contrib/benchmark/conf/vector-highlight-profile.alg	(revision 0)
+++ contrib/benchmark/conf/vector-highlight-profile.alg	(revision 0)
@@ -0,0 +1,68 @@
+#/**
+# * Licensed to the Apache Software Foundation (ASF) under one or more
+# * contributor license agreements.  See the NOTICE file distributed with
+# * this work for additional information regarding copyright ownership.
+# * The ASF licenses this file to You under the Apache License, Version 2.0
+# * (the "License"); you may not use this file except in compliance with
+# * the License.  You may obtain a copy of the License at
+# *
+# *     http://www.apache.org/licenses/LICENSE-2.0
+# *
+# * Unless required by applicable law or agreed to in writing, software
+# * distributed under the License is distributed on an "AS IS" BASIS,
+# * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# * See the License for the specific language governing permissions and
+# * limitations under the License.
+# */
+# -------------------------------------------------------------------------------------
+# multi val params are iterated by NewRound's, added to reports, start with column name.
+
+ram.flush.mb=flush:32:32
+compound=cmpnd:true:false
+
+analyzer=org.apache.lucene.analysis.standard.StandardAnalyzer
+directory=FSDirectory
+
+doc.stored=true
+doc.tokenized=true
+doc.term.vector=true
+doc.term.vector.offsets=true
+doc.term.vector.positions=true
+log.step=2000
+
+docs.dir=reuters-out
+
+content.source=org.apache.lucene.benchmark.byTask.feeds.ReutersContentSource
+
+query.maker=org.apache.lucene.benchmark.byTask.feeds.ReutersQueryMaker
+
+# task at this depth or less would print when they start
+task.max.depth.log=2
+
+log.queries=true
+# -------------------------------------------------------------------------------------
+{ "Populate"
+        CreateIndex
+        { "MAddDocs" AddDoc } : 20000
+        Optimize
+        CloseIndex
+    }
+{ "Rounds"
+
+    ResetSystemSoft
+
+
+    OpenReader
+      { "SearchVecHlgtSameRdr" SearchTravRetVectorHighlight(maxFrags[10],fields[body]) > : 1000
+
+    CloseReader
+
+    RepSumByPref MAddDocs
+
+    NewRound
+
+} : 4
+
+RepSumByNameRound
+RepSumByName
+RepSumByPrefRound MAddDocs
Index: contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/SearchTravRetVectorHighlightTask.java
===================================================================
--- contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/SearchTravRetVectorHighlightTask.java	(revision 0)
+++ contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/SearchTravRetVectorHighlightTask.java	(revision 0)
@@ -0,0 +1,141 @@
+package org.apache.lucene.benchmark.byTask.tasks;
+
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+import org.apache.lucene.analysis.Analyzer;
+import org.apache.lucene.benchmark.byTask.PerfRunData;
+import org.apache.lucene.document.Document;
+import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.search.Query;
+import org.apache.lucene.search.vectorhighlight.FastVectorHighlighter;
+import org.apache.lucene.search.vectorhighlight.FieldQuery;
+
+import java.util.Set;
+import java.util.Collection;
+import java.util.HashSet;
+import java.util.Collections;
+
+/**
+ * Search and Traverse and Retrieve docs task.  Highlight the fields in the retrieved documents by using FastVectorHighlighter.
+ *
+ * <p>Note: This task reuses the reader if it is already open.
+ * Otherwise a reader is opened at start and closed at the end.
+ * </p>
+ *
+ * <p>Takes optional multivalued, comma separated param string as: size[&lt;traversal size&gt;],highlight[&lt;int&gt;],maxFrags[&lt;int&gt;],mergeContiguous[&lt;boolean&gt;],fields[name1;name2;...]</p>
+ * <ul>
+ * <li>traversal size - The number of hits to traverse, otherwise all will be traversed</li>
+ * <li>highlight - The number of the hits to highlight.  Will always be less than or equal to traversal size.  Default is Integer.MAX_VALUE (i.e. hits.length())</li>
+ * <li>maxFrags - The maximum number of fragments to score by the highlighter</li>
+ * <li>fragSize - The length of fragments</li>
+ * <li>fields - The fields to highlight.  If not specified all fields will be highlighted (or at least attempted)</li>
+ * </ul>
+ * Example:
+ * <pre>"SearchVecHlgtSameRdr" SearchTravRetVectorHighlight(size[10],highlight[10],maxFrags[3],fields[body]) > : 1000
+ * </pre>
+ *
+ * Fields must be stored and term vector offsets and positions in order must be true for this task to work.
+ *
+ * <p>Other side effects: counts additional 1 (record) for each traversed hit,
+ * and 1 more for each retrieved (non null) document and 1 for each fragment returned.</p>
+ */
+public class SearchTravRetVectorHighlightTask extends SearchTravTask {
+
+  protected int numToHighlight = Integer.MAX_VALUE;
+  protected int maxFrags = 2;
+  protected int fragSize = 100;
+  protected Set paramFields = Collections.EMPTY_SET;
+  protected FastVectorHighlighter highlighter;
+
+  public SearchTravRetVectorHighlightTask(PerfRunData runData) {
+    super(runData);
+  }
+
+  public void setup() throws Exception {
+    super.setup();
+    //check to make sure either the doc is being stored
+    PerfRunData data = getRunData();
+    if (data.getConfig().get("doc.stored", false) == false){
+      throw new Exception("doc.stored must be set to true");
+    }
+    if (data.getConfig().get("doc.term.vector.offsets", false) == false){
+      throw new Exception("doc.term.vector.offsets must be set to true");
+    }
+    if (data.getConfig().get("doc.term.vector.positions", false) == false){
+      throw new Exception("doc.term.vector.positions must be set to true");
+    }
+  }
+
+  public boolean withRetrieve() {
+    return true;
+  }
+
+  public int numToHighlight() {
+    return numToHighlight;
+  }
+  
+  protected BenchmarkHighlighter getBenchmarkHighlighter(Query q){
+    highlighter = new FastVectorHighlighter( false, false );
+    final FieldQuery fq = highlighter.getFieldQuery( q );
+    return new BenchmarkHighlighter(){
+      public int doHighlight(IndexReader reader, int doc, String field,
+          Document document, Analyzer analyzer, String text) throws Exception {
+        String[] fragments = highlighter.getBestFragments(fq, reader, doc, field, fragSize, maxFrags);
+        return fragments != null ? fragments.length : 0;
+      }
+    };
+  }
+
+  public int maxNumFragments() {
+    return maxFrags;
+  }
+
+  protected Collection/*<String>*/ getFieldsToHighlight(Document document) {
+    Collection result = super.getFieldsToHighlight(document);
+    //if stored is false, then result will be empty, in which case just get all the param fields
+    if (paramFields.isEmpty() == false && result.isEmpty() == false) {
+      result.retainAll(paramFields);
+    } else {
+      result = paramFields;
+    }
+    return result;
+  }
+
+  public void setParams(String params) {
+    String [] splits = params.split(",");
+    for (int i = 0; i < splits.length; i++) {
+      if (splits[i].startsWith("size[") == true){
+        traversalSize = (int)Float.parseFloat(splits[i].substring("size[".length(),splits[i].length() - 1));
+      } else if (splits[i].startsWith("highlight[") == true){
+        numToHighlight = (int)Float.parseFloat(splits[i].substring("highlight[".length(),splits[i].length() - 1));
+      } else if (splits[i].startsWith("maxFrags[") == true){
+        maxFrags = (int)Float.parseFloat(splits[i].substring("maxFrags[".length(),splits[i].length() - 1));
+      } else if (splits[i].startsWith("fragSize[") == true){
+        fragSize = (int)Float.parseFloat(splits[i].substring("fragSize[".length(),splits[i].length() - 1));
+      } else if (splits[i].startsWith("fields[") == true){
+        paramFields = new HashSet();
+        String fieldNames = splits[i].substring("fields[".length(), splits[i].length() - 1);
+        String [] fieldSplits = fieldNames.split(";");
+        for (int j = 0; j < fieldSplits.length; j++) {
+          paramFields.add(fieldSplits[j]);          
+        }
+
+      }
+    }
+  }
+}
\ No newline at end of file

Property changes on: contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/SearchTravRetVectorHighlightTask.java
___________________________________________________________________
Name: svn:keywords
   + Date Author Id Revision HeadURL
Name: svn:eol-style
   + native

Index: contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/SearchTravRetHighlightTask.java
===================================================================
--- contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/SearchTravRetHighlightTask.java	(revision 799981)
+++ contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/SearchTravRetHighlightTask.java	(working copy)
@@ -17,9 +17,20 @@
  * limitations under the License.
  */
 
+import org.apache.lucene.analysis.Analyzer;
+import org.apache.lucene.analysis.TokenStream;
 import org.apache.lucene.benchmark.byTask.PerfRunData;
 import org.apache.lucene.document.Document;
+import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.search.Query;
+import org.apache.lucene.search.highlight.Highlighter;
+import org.apache.lucene.search.highlight.InvalidTokenOffsetsException;
+import org.apache.lucene.search.highlight.QueryScorer;
+import org.apache.lucene.search.highlight.SimpleHTMLFormatter;
+import org.apache.lucene.search.highlight.TextFragment;
+import org.apache.lucene.search.highlight.TokenSources;
 
+import java.io.IOException;
 import java.util.Set;
 import java.util.Collection;
 import java.util.HashSet;
@@ -57,7 +68,7 @@
   protected boolean mergeContiguous;
   protected int maxFrags = 2;
   protected Set paramFields = Collections.EMPTY_SET;
-  
+  protected Highlighter highlighter;
 
   public SearchTravRetHighlightTask(PerfRunData runData) {
     super(runData);
@@ -79,7 +90,22 @@
   public int numToHighlight() {
     return numToHighlight;
   }
+  
+  protected BenchmarkHighlighter getBenchmarkHighlighter(Query q){
+    highlighter = new Highlighter(new SimpleHTMLFormatter(), new QueryScorer(q));
+    return new BenchmarkHighlighter(){
+      public int doHighlight(IndexReader reader, int doc, String field,
+          Document document, Analyzer analyzer, String text) throws Exception {
+        TokenStream ts = TokenSources.getAnyTokenStream(reader, doc, field, document, analyzer);
+        TextFragment[] frag = highlighter.getBestTextFragments(ts, text, mergeContiguous, maxFrags);
+        return frag != null ? frag.length : 0;
+      }
+    };
+  }
 
+  /**
+   * @deprecated
+   */
   public boolean isMergeContiguousFragments() {
     return mergeContiguous;
   }
Index: contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/BenchmarkHighlighter.java
===================================================================
--- contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/BenchmarkHighlighter.java	(revision 0)
+++ contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/BenchmarkHighlighter.java	(revision 0)
@@ -0,0 +1,27 @@
+package org.apache.lucene.benchmark.byTask.tasks;
+
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+import org.apache.lucene.analysis.Analyzer;
+import org.apache.lucene.document.Document;
+import org.apache.lucene.index.IndexReader;
+
+public abstract class BenchmarkHighlighter {
+  public abstract int doHighlight( IndexReader reader, int doc, String field,
+      Document document, Analyzer analyzer, String text ) throws Exception ;
+}

Property changes on: contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/BenchmarkHighlighter.java
___________________________________________________________________
Name: svn:keywords
   + Date Author Id Revision HeadURL
Name: svn:eol-style
   + native

Index: contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/ReadTask.java
===================================================================
--- contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/ReadTask.java	(revision 799981)
+++ contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/ReadTask.java	(working copy)
@@ -121,13 +121,12 @@
             boolean retrieve = withRetrieve();
             int numHighlight = Math.min(numToHighlight(), scoreDocs.length);
             Analyzer analyzer = getRunData().getAnalyzer();
-            Highlighter highlighter = null;
+            BenchmarkHighlighter highlighter = null;
             int maxFrags = 1;
             if (numHighlight > 0) {
-              highlighter = getHighlighter(q);
+              highlighter = getBenchmarkHighlighter(q);
               maxFrags = maxNumFragments();
             }
-            boolean merge = isMergeContiguousFragments();
             for (int m = 0; m < traversalSize; m++) {
               int id = scoreDocs[m].doc;
               res++;
@@ -139,8 +138,7 @@
                   for (Iterator iterator = fieldsToHighlight.iterator(); iterator.hasNext();) {
                     String field = (String) iterator.next();
                     String text = document.get(field);
-                    TokenStream ts = TokenSources.getAnyTokenStream(ir, id, field, document, analyzer);
-                    res += doHighlight(ts, text, highlighter, merge, maxFrags);
+                    res += highlighter.doHighlight(ir, id, field, document, analyzer, text);
                   }
                 }
               }
@@ -241,9 +239,16 @@
     return 0;
   }
 
+  /**
+   * @deprecated Use {@link #getBenchmarkHighlighter(Query)}
+   */
   protected Highlighter getHighlighter(Query q){
     return new Highlighter(new SimpleHTMLFormatter(), new QueryScorer(q));
   }
+  
+  protected BenchmarkHighlighter getBenchmarkHighlighter(Query q){
+    return null;
+  }
 
   /**
    *
@@ -256,11 +261,15 @@
   /**
    *
    * @return true if the highlighter should merge contiguous fragments
+   * @deprecated
    */
   public boolean isMergeContiguousFragments(){
     return false;
   }
 
+  /**
+   * @deprecated
+   */
   protected int doHighlight(TokenStream ts, String text,  Highlighter highlighter, boolean mergeContiguous, int maxFragments) throws IOException, InvalidTokenOffsetsException {
     TextFragment[] frag = highlighter.getBestTextFragments(ts, text, mergeContiguous, maxFragments);
     return frag != null ? frag.length : 0;
Index: contrib/benchmark/build.xml
===================================================================
--- contrib/benchmark/build.xml	(revision 799981)
+++ contrib/benchmark/build.xml	(working copy)
@@ -105,6 +105,7 @@
         <pathelement path="${common.dir}/build/classes/java"/>
         <pathelement path="${common.dir}/build/classes/demo"/>
         <pathelement path="${common.dir}/build/contrib/highlighter/classes/java"/>
+        <pathelement path="${common.dir}/build/contrib/fast-vector-highlighter/classes/java"/>
     	<fileset dir="lib">
     		<include name="**/*.jar"/>
     	</fileset>
@@ -148,8 +149,13 @@
          <fileset dir="${common.dir}/contrib/highlighter" includes="build.xml"/>
       </subant>
     </target>
+    <target name="compile-vector-highlighter">
+      <subant target="compile">
+         <fileset dir="${common.dir}/contrib/fast-vector-highlighter" includes="build.xml"/>
+      </subant>
+    </target>
 
-    <target name="init" depends="common.init,compile-demo, compile-highlighter,check-files"/>
+    <target name="init" depends="common.init,compile-demo,compile-highlighter,compile-vector-highlighter,check-files"/>
 
     <!-- make sure online collections (reuters) are first downloaded -->
     <target name="test" depends="init,get-files">