More Like This Query: values of a multi-value fields are compared at the same level.

Previously, More Like This would create a new mlt query for each value of a
multi-value field. This could result in all the values of the field to be
selected, which defeats the purpose of More Like This. Instead, the correct
behavior is to generate only one mlt query for all the values of the field.
This commit provides the correct behavior for More Like This DSL. The fix for
More Like This API will be coming in another commit.

Closes #6310

src/main/java/org/elasticsearch/common/lucene/search/MoreLikeThisQuery.java

@@ -21,7 +21,6 @@ package org.elasticsearch.common.lucene.search;
 
 import org.apache.lucene.analysis.Analyzer;
 import org.apache.lucene.index.IndexReader;
-import org.apache.lucene.queries.mlt.MoreLikeThis;
 import org.apache.lucene.search.BooleanClause;
 import org.apache.lucene.search.BooleanQuery;
 import org.apache.lucene.search.Query;
@@ -31,6 +30,7 @@ import org.apache.lucene.search.similarities.TFIDFSimilarity;
 import org.elasticsearch.common.io.FastStringReader;
 
 import java.io.IOException;
+import java.io.Reader;
 import java.util.Arrays;
 import java.util.Set;
 
@@ -43,18 +43,18 @@ public class MoreLikeThisQuery extends Query {
 
     private TFIDFSimilarity similarity;
 
-    private String likeText;
+    private String[] likeText;
     private String[] moreLikeFields;
     private Analyzer analyzer;
     private float percentTermsToMatch = DEFAULT_PERCENT_TERMS_TO_MATCH;
-    private int minTermFrequency = MoreLikeThis.DEFAULT_MIN_TERM_FREQ;
-    private int maxQueryTerms = MoreLikeThis.DEFAULT_MAX_QUERY_TERMS;
-    private Set<?> stopWords = MoreLikeThis.DEFAULT_STOP_WORDS;
-    private int minDocFreq = MoreLikeThis.DEFAULT_MIN_DOC_FREQ;
-    private int maxDocFreq = MoreLikeThis.DEFAULT_MAX_DOC_FREQ;
-    private int minWordLen = MoreLikeThis.DEFAULT_MIN_WORD_LENGTH;
-    private int maxWordLen = MoreLikeThis.DEFAULT_MAX_WORD_LENGTH;
-    private boolean boostTerms = MoreLikeThis.DEFAULT_BOOST;
+    private int minTermFrequency = XMoreLikeThis.DEFAULT_MIN_TERM_FREQ;
+    private int maxQueryTerms = XMoreLikeThis.DEFAULT_MAX_QUERY_TERMS;
+    private Set<?> stopWords = XMoreLikeThis.DEFAULT_STOP_WORDS;
+    private int minDocFreq = XMoreLikeThis.DEFAULT_MIN_DOC_FREQ;
+    private int maxDocFreq = XMoreLikeThis.DEFAULT_MAX_DOC_FREQ;
+    private int minWordLen = XMoreLikeThis.DEFAULT_MIN_WORD_LENGTH;
+    private int maxWordLen = XMoreLikeThis.DEFAULT_MAX_WORD_LENGTH;
+    private boolean boostTerms = XMoreLikeThis.DEFAULT_BOOST;
     private float boostTermsFactor = 1;
 
 
@@ -63,7 +63,7 @@ public class MoreLikeThisQuery extends Query {
     }
 
     public MoreLikeThisQuery(String likeText, String[] moreLikeFields, Analyzer analyzer) {
-        this.likeText = likeText;
+        this.likeText = new String[]{likeText};
         this.moreLikeFields = moreLikeFields;
         this.analyzer = analyzer;
     }
@@ -72,7 +72,7 @@ public class MoreLikeThisQuery extends Query {
     public int hashCode() {
         int result = boostTerms ? 1 : 0;
         result = 31 * result + Float.floatToIntBits(boostTermsFactor);
-        result = 31 * result + likeText.hashCode();
+        result = 31 * result + Arrays.hashCode(likeText);
         result = 31 * result + maxDocFreq;
         result = 31 * result + maxQueryTerms;
         result = 31 * result + maxWordLen;
@@ -99,7 +99,7 @@ public class MoreLikeThisQuery extends Query {
             return false;
         if (boostTermsFactor != other.boostTermsFactor)
             return false;
-        if (!likeText.equals(other.likeText))
+        if (!(Arrays.equals(likeText, other.likeText)))
             return false;
         if (maxDocFreq != other.maxDocFreq)
             return false;
@@ -132,7 +132,7 @@ public class MoreLikeThisQuery extends Query {
 
     @Override
     public Query rewrite(IndexReader reader) throws IOException {
-        MoreLikeThis mlt = new MoreLikeThis(reader, similarity == null ? new DefaultSimilarity() : similarity);
+        XMoreLikeThis mlt = new XMoreLikeThis(reader, similarity == null ? new DefaultSimilarity() : similarity);
 
         mlt.setFieldNames(moreLikeFields);
         mlt.setAnalyzer(analyzer);
@@ -145,10 +145,15 @@ public class MoreLikeThisQuery extends Query {
         mlt.setStopWords(stopWords);
         mlt.setBoost(boostTerms);
         mlt.setBoostFactor(boostTermsFactor);
+
+        Reader[] readers = new Reader[likeText.length];
+        for (int i = 0; i < readers.length; i++) {
+            readers[i] = new FastStringReader(likeText[i]);
+        }
         //LUCENE 4 UPGRADE this mapps the 3.6 behavior (only use the first field)
-        BooleanQuery bq = (BooleanQuery) mlt.like(new FastStringReader(likeText), moreLikeFields[0]);
-        BooleanClause[] clauses = bq.getClauses();
+        BooleanQuery bq = (BooleanQuery) mlt.like(moreLikeFields[0], readers);
 
+        BooleanClause[] clauses = bq.getClauses();
         bq.setMinimumNumberShouldMatch((int) (clauses.length * percentTermsToMatch));
 
         bq.setBoost(getBoost());
@@ -157,14 +162,22 @@ public class MoreLikeThisQuery extends Query {
 
     @Override
     public String toString(String field) {
-        return "like:" + likeText;
+        return "like:" + Arrays.toString(likeText);
     }
 
     public String getLikeText() {
+        return (likeText == null ? null : likeText[0]);
+    }
+
+    public String[] getLikeTexts() {
         return likeText;
     }
 
     public void setLikeText(String likeText) {
+        this.likeText = new String[]{likeText};
+    }
+
+    public void setLikeText(String... likeText) {
         this.likeText = likeText;
     }
 

src/main/java/org/elasticsearch/common/lucene/search/XMoreLikeThis.java

@@ -0,0 +1,964 @@
+/*
+ * Licensed to Elasticsearch under one or more contributor
+ * license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright
+ * ownership. Elasticsearch 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.
+ */
+
+/**
+ * Copyright 2004-2005 The Apache Software Foundation.
+ *
+ * Licensed 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.
+ */
+
+package org.elasticsearch.common.lucene.search;
+
+import java.io.*;
+import java.util.*;
+
+import org.apache.lucene.analysis.Analyzer;
+import org.apache.lucene.analysis.TokenStream;
+import org.apache.lucene.analysis.tokenattributes.CharTermAttribute;
+import org.apache.lucene.document.Document;
+import org.apache.lucene.index.Fields;
+import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.index.IndexableField;
+import org.apache.lucene.index.MultiFields;
+import org.apache.lucene.index.Term;
+import org.apache.lucene.index.Terms;
+import org.apache.lucene.index.TermsEnum;
+import org.apache.lucene.search.*;
+import org.apache.lucene.search.similarities.DefaultSimilarity;
+import org.apache.lucene.search.similarities.TFIDFSimilarity;
+import org.apache.lucene.util.BytesRef;
+import org.apache.lucene.util.CharsRef;
+import org.apache.lucene.util.IOUtils;
+import org.apache.lucene.util.PriorityQueue;
+import org.apache.lucene.util.UnicodeUtil;
+import org.elasticsearch.Version;
+import org.elasticsearch.common.io.FastStringReader;
+
+/**
+ * Generate "more like this" similarity queries.
+ * Based on this mail:
+ * <code><pre>
+ * Lucene does let you access the document frequency of terms, with IndexReader.docFreq().
+ * Term frequencies can be computed by re-tokenizing the text, which, for a single document,
+ * is usually fast enough.  But looking up the docFreq() of every term in the document is
+ * probably too slow.
+ * <p/>
+ * You can use some heuristics to prune the set of terms, to avoid calling docFreq() too much,
+ * or at all.  Since you're trying to maximize a tf*idf score, you're probably most interested
+ * in terms with a high tf. Choosing a tf threshold even as low as two or three will radically
+ * reduce the number of terms under consideration.  Another heuristic is that terms with a
+ * high idf (i.e., a low df) tend to be longer.  So you could threshold the terms by the
+ * number of characters, not selecting anything less than, e.g., six or seven characters.
+ * With these sorts of heuristics you can usually find small set of, e.g., ten or fewer terms
+ * that do a pretty good job of characterizing a document.
+ * <p/>
+ * It all depends on what you're trying to do.  If you're trying to eek out that last percent
+ * of precision and recall regardless of computational difficulty so that you can win a TREC
+ * competition, then the techniques I mention above are useless.  But if you're trying to
+ * provide a "more like this" button on a search results page that does a decent job and has
+ * good performance, such techniques might be useful.
+ * <p/>
+ * An efficient, effective "more-like-this" query generator would be a great contribution, if
+ * anyone's interested.  I'd imagine that it would take a Reader or a String (the document's
+ * text), analyzer Analyzer, and return a set of representative terms using heuristics like those
+ * above.  The frequency and length thresholds could be parameters, etc.
+ * <p/>
+ * Doug
+ * </pre></code>
+ * <p/>
+ * <p/>
+ * <p/>
+ * <h3>Initial Usage</h3>
+ * <p/>
+ * This class has lots of options to try to make it efficient and flexible.
+ * The simplest possible usage is as follows. The bold
+ * fragment is specific to this class.
+ * <p/>
+ * <pre class="prettyprint">
+ * <p/>
+ * IndexReader ir = ...
+ * IndexSearcher is = ...
+ * <p/>
+ * MoreLikeThis mlt = new MoreLikeThis(ir);
+ * Reader target = ... // orig source of doc you want to find similarities to
+ * Query query = mlt.like( target);
+ * <p/>
+ * Hits hits = is.search(query);
+ * // now the usual iteration thru 'hits' - the only thing to watch for is to make sure
+ * //you ignore the doc if it matches your 'target' document, as it should be similar to itself
+ * <p/>
+ * </pre>
+ * <p/>
+ * Thus you:
+ * <ol>
+ * <li> do your normal, Lucene setup for searching,
+ * <li> create a MoreLikeThis,
+ * <li> get the text of the doc you want to find similarities to
+ * <li> then call one of the like() calls to generate a similarity query
+ * <li> call the searcher to find the similar docs
+ * </ol>
+ * <p/>
+ * <h3>More Advanced Usage</h3>
+ * <p/>
+ * You may want to use {@link #setFieldNames setFieldNames(...)} so you can examine
+ * multiple fields (e.g. body and title) for similarity.
+ * <p/>
+ * <p/>
+ * Depending on the size of your index and the size and makeup of your documents you
+ * may want to call the other set methods to control how the similarity queries are
+ * generated:
+ * <ul>
+ * <li> {@link #setMinTermFreq setMinTermFreq(...)}
+ * <li> {@link #setMinDocFreq setMinDocFreq(...)}
+ * <li> {@link #setMaxDocFreq setMaxDocFreq(...)}
+ * <li> {@link #setMaxDocFreqPct setMaxDocFreqPct(...)}
+ * <li> {@link #setMinWordLen setMinWordLen(...)}
+ * <li> {@link #setMaxWordLen setMaxWordLen(...)}
+ * <li> {@link #setMaxQueryTerms setMaxQueryTerms(...)}
+ * <li> {@link #setMaxNumTokensParsed setMaxNumTokensParsed(...)}
+ * <li> {@link #setStopWords setStopWord(...)}
+ * </ul>
+ * <p/>
+ * <hr>
+ * <pre>
+ * Changes: Mark Harwood 29/02/04
+ * Some bugfixing, some refactoring, some optimisation.
+ * - bugfix: retrieveTerms(int docNum) was not working for indexes without a termvector -added missing code
+ * - bugfix: No significant terms being created for fields with a termvector - because
+ * was only counting one occurrence per term/field pair in calculations(ie not including frequency info from TermVector)
+ * - refactor: moved common code into isNoiseWord()
+ * - optimise: when no termvector support available - used maxNumTermsParsed to limit amount of tokenization
+ * </pre>
+ */
+
+public final class XMoreLikeThis {
+
+    static {
+        assert Version.CURRENT.luceneVersion == org.apache.lucene.util.Version.LUCENE_48: "Remove this class once we upgrade to Lucene 4.9";
+    }
+
+    /**
+     * Default maximum number of tokens to parse in each example doc field that is not stored with TermVector support.
+     *
+     * @see #getMaxNumTokensParsed
+     */
+    public static final int DEFAULT_MAX_NUM_TOKENS_PARSED = 5000;
+
+    /**
+     * Ignore terms with less than this frequency in the source doc.
+     *
+     * @see #getMinTermFreq
+     * @see #setMinTermFreq
+     */
+    public static final int DEFAULT_MIN_TERM_FREQ = 2;
+
+    /**
+     * Ignore words which do not occur in at least this many docs.
+     *
+     * @see #getMinDocFreq
+     * @see #setMinDocFreq
+     */
+    public static final int DEFAULT_MIN_DOC_FREQ = 5;
+
+    /**
+     * Ignore words which occur in more than this many docs.
+     *
+     * @see #getMaxDocFreq
+     * @see #setMaxDocFreq
+     * @see #setMaxDocFreqPct
+     */
+    public static final int DEFAULT_MAX_DOC_FREQ = Integer.MAX_VALUE;
+
+    /**
+     * Boost terms in query based on score.
+     *
+     * @see #isBoost
+     * @see #setBoost
+     */
+    public static final boolean DEFAULT_BOOST = false;
+
+    /**
+     * Default field names. Null is used to specify that the field names should be looked
+     * up at runtime from the provided reader.
+     */
+    public static final String[] DEFAULT_FIELD_NAMES = new String[]{"contents"};
+
+    /**
+     * Ignore words less than this length or if 0 then this has no effect.
+     *
+     * @see #getMinWordLen
+     * @see #setMinWordLen
+     */
+    public static final int DEFAULT_MIN_WORD_LENGTH = 0;
+
+    /**
+     * Ignore words greater than this length or if 0 then this has no effect.
+     *
+     * @see #getMaxWordLen
+     * @see #setMaxWordLen
+     */
+    public static final int DEFAULT_MAX_WORD_LENGTH = 0;
+
+    /**
+     * Default set of stopwords.
+     * If null means to allow stop words.
+     *
+     * @see #setStopWords
+     * @see #getStopWords
+     */
+    public static final Set<?> DEFAULT_STOP_WORDS = null;
+
+    /**
+     * Current set of stop words.
+     */
+    private Set<?> stopWords = DEFAULT_STOP_WORDS;
+
+    /**
+     * Return a Query with no more than this many terms.
+     *
+     * @see BooleanQuery#getMaxClauseCount
+     * @see #getMaxQueryTerms
+     * @see #setMaxQueryTerms
+     */
+    public static final int DEFAULT_MAX_QUERY_TERMS = 25;
+
+    /**
+     * Analyzer that will be used to parse the doc.
+     */
+    private Analyzer analyzer = null;
+
+    /**
+     * Ignore words less frequent that this.
+     */
+    private int minTermFreq = DEFAULT_MIN_TERM_FREQ;
+
+    /**
+     * Ignore words which do not occur in at least this many docs.
+     */
+    private int minDocFreq = DEFAULT_MIN_DOC_FREQ;
+
+    /**
+     * Ignore words which occur in more than this many docs.
+     */
+    private int maxDocFreq = DEFAULT_MAX_DOC_FREQ;
+
+    /**
+     * Should we apply a boost to the Query based on the scores?
+     */
+    private boolean boost = DEFAULT_BOOST;
+
+    /**
+     * Field name we'll analyze.
+     */
+    private String[] fieldNames = DEFAULT_FIELD_NAMES;
+
+    /**
+     * The maximum number of tokens to parse in each example doc field that is not stored with TermVector support
+     */
+    private int maxNumTokensParsed = DEFAULT_MAX_NUM_TOKENS_PARSED;
+
+    /**
+     * Ignore words if less than this len.
+     */
+    private int minWordLen = DEFAULT_MIN_WORD_LENGTH;
+
+    /**
+     * Ignore words if greater than this len.
+     */
+    private int maxWordLen = DEFAULT_MAX_WORD_LENGTH;
+
+    /**
+     * Don't return a query longer than this.
+     */
+    private int maxQueryTerms = DEFAULT_MAX_QUERY_TERMS;
+
+    /**
+     * For idf() calculations.
+     */
+    private TFIDFSimilarity similarity;// = new DefaultSimilarity();
+
+    /**
+     * IndexReader to use
+     */
+    private final IndexReader ir;
+
+    /**
+     * Boost factor to use when boosting the terms
+     */
+    private float boostFactor = 1;
+
+    /**
+     * Returns the boost factor used when boosting terms
+     *
+     * @return the boost factor used when boosting terms
+     * @see #setBoostFactor(float)
+     */
+    public float getBoostFactor() {
+        return boostFactor;
+    }
+
+    /**
+     * Sets the boost factor to use when boosting terms
+     *
+     * @see #getBoostFactor()
+     */
+    public void setBoostFactor(float boostFactor) {
+        this.boostFactor = boostFactor;
+    }
+
+    /**
+     * Constructor requiring an IndexReader.
+     */
+    public XMoreLikeThis(IndexReader ir) {
+        this(ir, new DefaultSimilarity());
+    }
+
+    public XMoreLikeThis(IndexReader ir, TFIDFSimilarity sim) {
+        this.ir = ir;
+        this.similarity = sim;
+    }
+
+
+    public TFIDFSimilarity getSimilarity() {
+        return similarity;
+    }
+
+    public void setSimilarity(TFIDFSimilarity similarity) {
+        this.similarity = similarity;
+    }
+
+    /**
+     * Returns an analyzer that will be used to parse source doc with. The default analyzer
+     * is not set.
+     *
+     * @return the analyzer that will be used to parse source doc with.
+     */
+    public Analyzer getAnalyzer() {
+        return analyzer;
+    }
+
+    /**
+     * Sets the analyzer to use. An analyzer is not required for generating a query with the
+     * {@link #like(int)} method, all other 'like' methods require an analyzer.
+     *
+     * @param analyzer the analyzer to use to tokenize text.
+     */
+    public void setAnalyzer(Analyzer analyzer) {
+        this.analyzer = analyzer;
+    }
+
+    /**
+     * Returns the frequency below which terms will be ignored in the source doc. The default
+     * frequency is the {@link #DEFAULT_MIN_TERM_FREQ}.
+     *
+     * @return the frequency below which terms will be ignored in the source doc.
+     */
+    public int getMinTermFreq() {
+        return minTermFreq;
+    }
+
+    /**
+     * Sets the frequency below which terms will be ignored in the source doc.
+     *
+     * @param minTermFreq the frequency below which terms will be ignored in the source doc.
+     */
+    public void setMinTermFreq(int minTermFreq) {
+        this.minTermFreq = minTermFreq;
+    }
+
+    /**
+     * Returns the frequency at which words will be ignored which do not occur in at least this
+     * many docs. The default frequency is {@link #DEFAULT_MIN_DOC_FREQ}.
+     *
+     * @return the frequency at which words will be ignored which do not occur in at least this
+     *         many docs.
+     */
+    public int getMinDocFreq() {
+        return minDocFreq;
+    }
+
+    /**
+     * Sets the frequency at which words will be ignored which do not occur in at least this
+     * many docs.
+     *
+     * @param minDocFreq the frequency at which words will be ignored which do not occur in at
+     * least this many docs.
+     */
+    public void setMinDocFreq(int minDocFreq) {
+        this.minDocFreq = minDocFreq;
+    }
+
+    /**
+     * Returns the maximum frequency in which words may still appear.
+     * Words that appear in more than this many docs will be ignored. The default frequency is
+     * {@link #DEFAULT_MAX_DOC_FREQ}.
+     *
+     * @return get the maximum frequency at which words are still allowed,
+     *         words which occur in more docs than this are ignored.
+     */
+    public int getMaxDocFreq() {
+        return maxDocFreq;
+    }
+
+    /**
+     * Set the maximum frequency in which words may still appear. Words that appear
+     * in more than this many docs will be ignored.
+     *
+     * @param maxFreq the maximum count of documents that a term may appear
+     * in to be still considered relevant
+     */
+    public void setMaxDocFreq(int maxFreq) {
+        this.maxDocFreq = maxFreq;
+    }
+
+    /**
+     * Set the maximum percentage in which words may still appear. Words that appear
+     * in more than this many percent of all docs will be ignored.
+     *
+     * @param maxPercentage the maximum percentage of documents (0-100) that a term may appear
+     * in to be still considered relevant
+     */
+    public void setMaxDocFreqPct(int maxPercentage) {
+        this.maxDocFreq = maxPercentage * ir.numDocs() / 100;
+    }
+
+
+    /**
+     * Returns whether to boost terms in query based on "score" or not. The default is
+     * {@link #DEFAULT_BOOST}.
+     *
+     * @return whether to boost terms in query based on "score" or not.
+     * @see #setBoost
+     */
+    public boolean isBoost() {
+        return boost;
+    }
+
+    /**
+     * Sets whether to boost terms in query based on "score" or not.
+     *
+     * @param boost true to boost terms in query based on "score", false otherwise.
+     * @see #isBoost
+     */
+    public void setBoost(boolean boost) {
+        this.boost = boost;
+    }
+
+    /**
+     * Returns the field names that will be used when generating the 'More Like This' query.
+     * The default field names that will be used is {@link #DEFAULT_FIELD_NAMES}.
+     *
+     * @return the field names that will be used when generating the 'More Like This' query.
+     */
+    public String[] getFieldNames() {
+        return fieldNames;
+    }
+
+    /**
+     * Sets the field names that will be used when generating the 'More Like This' query.
+     * Set this to null for the field names to be determined at runtime from the IndexReader
+     * provided in the constructor.
+     *
+     * @param fieldNames the field names that will be used when generating the 'More Like This'
+     * query.
+     */
+    public void setFieldNames(String[] fieldNames) {
+        this.fieldNames = fieldNames;
+    }
+
+    /**
+     * Returns the minimum word length below which words will be ignored. Set this to 0 for no
+     * minimum word length. The default is {@link #DEFAULT_MIN_WORD_LENGTH}.
+     *
+     * @return the minimum word length below which words will be ignored.
+     */
+    public int getMinWordLen() {
+        return minWordLen;
+    }
+
+    /**
+     * Sets the minimum word length below which words will be ignored.
+     *
+     * @param minWordLen the minimum word length below which words will be ignored.
+     */
+    public void setMinWordLen(int minWordLen) {
+        this.minWordLen = minWordLen;
+    }
+
+    /**
+     * Returns the maximum word length above which words will be ignored. Set this to 0 for no
+     * maximum word length. The default is {@link #DEFAULT_MAX_WORD_LENGTH}.
+     *
+     * @return the maximum word length above which words will be ignored.
+     */
+    public int getMaxWordLen() {
+        return maxWordLen;
+    }
+
+    /**
+     * Sets the maximum word length above which words will be ignored.
+     *
+     * @param maxWordLen the maximum word length above which words will be ignored.
+     */
+    public void setMaxWordLen(int maxWordLen) {
+        this.maxWordLen = maxWordLen;
+    }
+
+    /**
+     * Set the set of stopwords.
+     * Any word in this set is considered "uninteresting" and ignored.
+     * Even if your Analyzer allows stopwords, you might want to tell the MoreLikeThis code to ignore them, as
+     * for the purposes of document similarity it seems reasonable to assume that "a stop word is never interesting".
+     *
+     * @param stopWords set of stopwords, if null it means to allow stop words
+     * @see #getStopWords
+     */
+    public void setStopWords(Set<?> stopWords) {
+        this.stopWords = stopWords;
+    }
+
+    /**
+     * Get the current stop words being used.
+     *
+     * @see #setStopWords
+     */
+    public Set<?> getStopWords() {
+        return stopWords;
+    }
+
+
+    /**
+     * Returns the maximum number of query terms that will be included in any generated query.
+     * The default is {@link #DEFAULT_MAX_QUERY_TERMS}.
+     *
+     * @return the maximum number of query terms that will be included in any generated query.
+     */
+    public int getMaxQueryTerms() {
+        return maxQueryTerms;
+    }
+
+    /**
+     * Sets the maximum number of query terms that will be included in any generated query.
+     *
+     * @param maxQueryTerms the maximum number of query terms that will be included in any
+     * generated query.
+     */
+    public void setMaxQueryTerms(int maxQueryTerms) {
+        this.maxQueryTerms = maxQueryTerms;
+    }
+
+    /**
+     * @return The maximum number of tokens to parse in each example doc field that is not stored with TermVector support
+     * @see #DEFAULT_MAX_NUM_TOKENS_PARSED
+     */
+    public int getMaxNumTokensParsed() {
+        return maxNumTokensParsed;
+    }
+
+    /**
+     * @param i The maximum number of tokens to parse in each example doc field that is not stored with TermVector support
+     */
+    public void setMaxNumTokensParsed(int i) {
+        maxNumTokensParsed = i;
+    }
+
+
+    /**
+     * Return a query that will return docs like the passed lucene document ID.
+     *
+     * @param docNum the documentID of the lucene doc to generate the 'More Like This" query for.
+     * @return a query that will return docs like the passed lucene document ID.
+     */
+    public Query like(int docNum) throws IOException {
+        if (fieldNames == null) {
+            // gather list of valid fields from lucene
+            Collection<String> fields = MultiFields.getIndexedFields(ir);
+            fieldNames = fields.toArray(new String[fields.size()]);
+        }
+
+        return createQuery(retrieveTerms(docNum));
+    }
+
+    /**
+     * Return a query that will return docs like the passed Reader.
+     *
+     * @return a query that will return docs like the passed Reader.
+     */
+    @Deprecated
+    public Query like(Reader r, String fieldName) throws IOException {
+        return like(fieldName, r);
+    }
+
+    /**
+     * Return a query that will return docs like the passed Readers.
+     * This was added in order to treat multi-value fields.
+     *
+     * @return a query that will return docs like the passed Readers.
+     */
+    public Query like(String fieldName, Reader... readers) throws IOException {
+        Map<String, Int> words = new HashMap<>();
+        for (Reader r : readers) {
+            addTermFrequencies(r, words, fieldName);
+        }
+        return createQuery(createQueue(words));
+    }
+
+    /**
+     * Create the More like query from a PriorityQueue
+     */
+    private Query createQuery(PriorityQueue<Object[]> q) {
+        BooleanQuery query = new BooleanQuery();
+        Object cur;
+        int qterms = 0;
+        float bestScore = 0;
+
+        while ((cur = q.pop()) != null) {
+            Object[] ar = (Object[]) cur;
+            TermQuery tq = new TermQuery(new Term((String) ar[1], (String) ar[0]));
+
+            if (boost) {
+                if (qterms == 0) {
+                    bestScore = ((Float) ar[2]);
+                }
+                float myScore = ((Float) ar[2]);
+
+                tq.setBoost(boostFactor * myScore / bestScore);
+            }
+
+            try {
+                query.add(tq, BooleanClause.Occur.SHOULD);
+            }
+            catch (BooleanQuery.TooManyClauses ignore) {
+                break;
+            }
+
+            qterms++;
+            if (maxQueryTerms > 0 && qterms >= maxQueryTerms) {
+                break;
+            }
+        }
+
+        return query;
+    }
+
+    /**
+     * Create a PriorityQueue from a word->tf map.
+     *
+     * @param words a map of words keyed on the word(String) with Int objects as the values.
+     */
+    private PriorityQueue<Object[]> createQueue(Map<String, Int> words) throws IOException {
+        // have collected all words in doc and their freqs
+        int numDocs = ir.numDocs();
+        FreqQ res = new FreqQ(words.size()); // will order words by score
+
+        for (String word : words.keySet()) { // for every word
+            int tf = words.get(word).x; // term freq in the source doc
+            if (minTermFreq > 0 && tf < minTermFreq) {
+                continue; // filter out words that don't occur enough times in the source
+            }
+
+            // go through all the fields and find the largest document frequency
+            String topField = fieldNames[0];
+            int docFreq = 0;
+            for (String fieldName : fieldNames) {
+                int freq = ir.docFreq(new Term(fieldName, word));
+                topField = (freq > docFreq) ? fieldName : topField;
+                docFreq = (freq > docFreq) ? freq : docFreq;
+            }
+
+            if (minDocFreq > 0 && docFreq < minDocFreq) {
+                continue; // filter out words that don't occur in enough docs
+            }
+
+            if (docFreq > maxDocFreq) {
+                continue; // filter out words that occur in too many docs
+            }
+
+            if (docFreq == 0) {
+                continue; // index update problem?
+            }
+
+            float idf = similarity.idf(docFreq, numDocs);
+            float score = tf * idf;
+
+            // only really need 1st 3 entries, other ones are for troubleshooting
+            res.insertWithOverflow(new Object[]{word,                   // the word
+                    topField,               // the top field
+                    score,       // overall score
+                    idf,         // idf
+                    docFreq,   // freq in all docs
+                    tf
+            });
+        }
+        return res;
+    }
+
+    /**
+     * Describe the parameters that control how the "more like this" query is formed.
+     */
+    public String describeParams() {
+        StringBuilder sb = new StringBuilder();
+        sb.append("\t").append("maxQueryTerms  : ").append(maxQueryTerms).append("\n");
+        sb.append("\t").append("minWordLen     : ").append(minWordLen).append("\n");
+        sb.append("\t").append("maxWordLen     : ").append(maxWordLen).append("\n");
+        sb.append("\t").append("fieldNames     : ");
+        String delim = "";
+        for (String fieldName : fieldNames) {
+            sb.append(delim).append(fieldName);
+            delim = ", ";
+        }
+        sb.append("\n");
+        sb.append("\t").append("boost          : ").append(boost).append("\n");
+        sb.append("\t").append("minTermFreq    : ").append(minTermFreq).append("\n");
+        sb.append("\t").append("minDocFreq     : ").append(minDocFreq).append("\n");
+        return sb.toString();
+    }
+
+    /**
+     * Find words for a more-like-this query former.
+     *
+     * @param docNum the id of the lucene document from which to find terms
+     */
+    public PriorityQueue<Object[]> retrieveTerms(int docNum) throws IOException {
+        Map<String, Int> termFreqMap = new HashMap<>();
+        for (String fieldName : fieldNames) {
+            final Fields vectors = ir.getTermVectors(docNum);
+            final Terms vector;
+            if (vectors != null) {
+                vector = vectors.terms(fieldName);
+            } else {
+                vector = null;
+            }
+
+            // field does not store term vector info
+            if (vector == null) {
+                Document d = ir.document(docNum);
+                IndexableField fields[] = d.getFields(fieldName);
+                for (IndexableField field : fields) {
+                    final String stringValue = field.stringValue();
+                    if (stringValue != null) {
+                        addTermFrequencies(new FastStringReader(stringValue), termFreqMap, fieldName);
+                    }
+                }
+            } else {
+                addTermFrequencies(termFreqMap, vector);
+            }
+        }
+
+        return createQueue(termFreqMap);
+    }
+
+    /**
+     * Adds terms and frequencies found in vector into the Map termFreqMap
+     *
+     * @param termFreqMap a Map of terms and their frequencies
+     * @param vector List of terms and their frequencies for a doc/field
+     */
+    private void addTermFrequencies(Map<String, Int> termFreqMap, Terms vector) throws IOException {
+        final TermsEnum termsEnum = vector.iterator(null);
+        final CharsRef spare = new CharsRef();
+        BytesRef text;
+        while((text = termsEnum.next()) != null) {
+            UnicodeUtil.UTF8toUTF16(text, spare);
+            final String term = spare.toString();
+            if (isNoiseWord(term)) {
+                continue;
+            }
+            final int freq = (int) termsEnum.totalTermFreq();
+
+            // increment frequency
+            Int cnt = termFreqMap.get(term);
+            if (cnt == null) {
+                cnt = new Int();
+                termFreqMap.put(term, cnt);
+                cnt.x = freq;
+            } else {
+                cnt.x += freq;
+            }
+        }
+    }
+
+    /**
+     * Adds term frequencies found by tokenizing text from reader into the Map words
+     *
+     * @param r a source of text to be tokenized
+     * @param termFreqMap a Map of terms and their frequencies
+     * @param fieldName Used by analyzer for any special per-field analysis
+     */
+    private void addTermFrequencies(Reader r, Map<String, Int> termFreqMap, String fieldName)
+            throws IOException {
+        if (analyzer == null) {
+            throw new UnsupportedOperationException("To use MoreLikeThis without " +
+                    "term vectors, you must provide an Analyzer");
+        }
+        TokenStream ts = analyzer.tokenStream(fieldName, r);
+        try {
+            int tokenCount = 0;
+            // for every token
+            CharTermAttribute termAtt = ts.addAttribute(CharTermAttribute.class);
+            ts.reset();
+            while (ts.incrementToken()) {
+                String word = termAtt.toString();
+                tokenCount++;
+                if (tokenCount > maxNumTokensParsed) {
+                    break;
+                }
+                if (isNoiseWord(word)) {
+                    continue;
+                }
+
+                // increment frequency
+                Int cnt = termFreqMap.get(word);
+                if (cnt == null) {
+                    termFreqMap.put(word, new Int());
+                } else {
+                    cnt.x++;
+                }
+            }
+            ts.end();
+        } finally {
+            IOUtils.closeWhileHandlingException(ts);
+        }
+    }
+
+
+    /**
+     * determines if the passed term is likely to be of interest in "more like" comparisons
+     *
+     * @param term The word being considered
+     * @return true if should be ignored, false if should be used in further analysis
+     */
+    private boolean isNoiseWord(String term) {
+        int len = term.length();
+        if (minWordLen > 0 && len < minWordLen) {
+            return true;
+        }
+        if (maxWordLen > 0 && len > maxWordLen) {
+            return true;
+        }
+        return stopWords != null && stopWords.contains(term);
+    }
+
+
+    /**
+     * Find words for a more-like-this query former.
+     * The result is a priority queue of arrays with one entry for <b>every word</b> in the document.
+     * Each array has 6 elements.
+     * The elements are:
+     * <ol>
+     * <li> The word (String)
+     * <li> The top field that this word comes from (String)
+     * <li> The score for this word (Float)
+     * <li> The IDF value (Float)
+     * <li> The frequency of this word in the index (Integer)
+     * <li> The frequency of this word in the source document (Integer)
+     * </ol>
+     * This is a somewhat "advanced" routine, and in general only the 1st entry in the array is of interest.
+     * This method is exposed so that you can identify the "interesting words" in a document.
+     * For an easier method to call see {@link #retrieveInterestingTerms retrieveInterestingTerms()}.
+     *
+     * @param r the reader that has the content of the document
+     * @param fieldName field passed to the analyzer to use when analyzing the content
+     * @return the most interesting words in the document ordered by score, with the highest scoring, or best entry, first
+     * @see #retrieveInterestingTerms
+     */
+    public PriorityQueue<Object[]> retrieveTerms(Reader r, String fieldName) throws IOException {
+        Map<String, Int> words = new HashMap<>();
+        addTermFrequencies(r, words, fieldName);
+        return createQueue(words);
+    }
+
+    /**
+     * @see #retrieveInterestingTerms(java.io.Reader, String)
+     */
+    public String[] retrieveInterestingTerms(int docNum) throws IOException {
+        ArrayList<Object> al = new ArrayList<>(maxQueryTerms);
+        PriorityQueue<Object[]> pq = retrieveTerms(docNum);
+        Object cur;
+        int lim = maxQueryTerms; // have to be careful, retrieveTerms returns all words but that's probably not useful to our caller...
+        // we just want to return the top words
+        while (((cur = pq.pop()) != null) && lim-- > 0) {
+            Object[] ar = (Object[]) cur;
+            al.add(ar[0]); // the 1st entry is the interesting word
+        }
+        String[] res = new String[al.size()];
+        return al.toArray(res);
+    }
+
+    /**
+     * Convenience routine to make it easy to return the most interesting words in a document.
+     * More advanced users will call {@link #retrieveTerms(Reader, String) retrieveTerms()} directly.
+     *
+     * @param r the source document
+     * @param fieldName field passed to analyzer to use when analyzing the content
+     * @return the most interesting words in the document
+     * @see #retrieveTerms(java.io.Reader, String)
+     * @see #setMaxQueryTerms
+     */
+    public String[] retrieveInterestingTerms(Reader r, String fieldName) throws IOException {
+        ArrayList<Object> al = new ArrayList<>(maxQueryTerms);
+        PriorityQueue<Object[]> pq = retrieveTerms(r, fieldName);
+        Object cur;
+        int lim = maxQueryTerms; // have to be careful, retrieveTerms returns all words but that's probably not useful to our caller...
+        // we just want to return the top words
+        while (((cur = pq.pop()) != null) && lim-- > 0) {
+            Object[] ar = (Object[]) cur;
+            al.add(ar[0]); // the 1st entry is the interesting word
+        }
+        String[] res = new String[al.size()];
+        return al.toArray(res);
+    }
+
+    /**
+     * PriorityQueue that orders words by score.
+     */
+    private static class FreqQ extends PriorityQueue<Object[]> {
+        FreqQ(int s) {
+            super(s);
+        }
+
+        @Override
+        protected boolean lessThan(Object[] aa, Object[] bb) {
+            Float fa = (Float) aa[2];
+            Float fb = (Float) bb[2];
+            return fa > fb;
+        }
+    }
+
+    /**
+     * Use for frequencies and to avoid renewing Integers.
+     */
+    private static class Int {
+        int x;
+
+        Int() {
+            x = 1;
+        }
+    }
+}

src/main/java/org/elasticsearch/index/query/MoreLikeThisQueryParser.java

@@ -39,6 +39,7 @@ import org.elasticsearch.index.analysis.Analysis;
 import org.elasticsearch.index.mapper.Uid;
 import org.elasticsearch.index.mapper.internal.UidFieldMapper;
 import org.elasticsearch.index.search.morelikethis.MoreLikeThisFetchService;
+import org.elasticsearch.index.search.morelikethis.MoreLikeThisFetchService.LikeText;
 
 import java.io.IOException;
 import java.util.*;
@@ -205,11 +206,11 @@ public class MoreLikeThisQueryParser implements QueryParser {
                 }
             }
             // fetching the items with multi-get
-            List<MoreLikeThisFetchService.LikeText> likeTexts = fetchService.fetch(items);
+            List<LikeText> likeTexts = fetchService.fetch(items);
             // right now we are just building a boolean query
             BooleanQuery boolQuery = new BooleanQuery();
-            for (MoreLikeThisFetchService.LikeText likeText : likeTexts) {
-                addMoreLikeThis(boolQuery, mltQuery, likeText.field, likeText.text);
+            for (LikeText likeText : likeTexts) {
+                addMoreLikeThis(boolQuery, mltQuery, likeText);
             }
             // exclude the items from the search
             if (!include) {
@@ -227,10 +228,10 @@ public class MoreLikeThisQueryParser implements QueryParser {
         return mltQuery;
     }
 
-    private void addMoreLikeThis(BooleanQuery boolQuery, MoreLikeThisQuery mltQuery, String fieldName, String likeText) {
+    private void addMoreLikeThis(BooleanQuery boolQuery, MoreLikeThisQuery mltQuery, LikeText likeText) {
         MoreLikeThisQuery mlt = new MoreLikeThisQuery();
-        mlt.setMoreLikeFields(new String[] {fieldName});
-        mlt.setLikeText(likeText);
+        mlt.setMoreLikeFields(new String[] {likeText.field});
+        mlt.setLikeText(likeText.text);
         mlt.setAnalyzer(mltQuery.getAnalyzer());
         mlt.setPercentTermsToMatch(mltQuery.getPercentTermsToMatch());
         mlt.setBoostTerms(mltQuery.isBoostTerms());

src/main/java/org/elasticsearch/index/search/morelikethis/MoreLikeThisFetchService.java

@@ -40,9 +40,14 @@ public class MoreLikeThisFetchService extends AbstractComponent {
 
     public static final class LikeText {
         public final String field;
-        public final String text;
+        public final String[] text;
 
         public LikeText(String field, String text) {
+            this.field = field;
+            this.text = new String[]{text};
+        }
+
+        public LikeText(String field, String... text) {
             this.field = field;
             this.text = text;
         }
@@ -73,9 +78,11 @@ public class MoreLikeThisFetchService extends AbstractComponent {
             }
 
             for (GetField getField : getResponse.getFields().values()) {
-                for (Object value : getField.getValues()) {
-                    likeTexts.add(new LikeText(getField.getName(), value.toString()));
+                String[] text = new String[getField.getValues().size()];
+                for (int i = 0; i < text.length; i++) {
+                    text[i] = getField.getValues().get(i).toString();
                 }
+                likeTexts.add(new LikeText(getField.getName(), text));
             }
         }
         return likeTexts;

src/test/java/org/elasticsearch/index/query/ItemSerializationTests.java

@@ -41,22 +41,12 @@ import static org.hamcrest.Matchers.is;
 
 public class ItemSerializationTests extends ElasticsearchTestCase {
 
-    private String[] generateRandomStringArray(int arraySize, int stringSize) {
-        String[] array = randomBoolean() ? new String[randomInt(arraySize)] : null; // allow empty arrays
-        if (array != null) {
-            for (int i = 0; i < array.length; i++) {
-                array[i] = randomAsciiOfLength(stringSize);
-            }
-        }
-        return array;
-    }
-
     private Item generateRandomItem(int arraySize, int stringSize) {
         String index = randomAsciiOfLength(stringSize);
         String type = randomAsciiOfLength(stringSize);
         String id = String.valueOf(Math.abs(randomInt()));
         String routing = randomBoolean() ? randomAsciiOfLength(stringSize) : null;
-        String[] fields = generateRandomStringArray(arraySize, stringSize);
+        String[] fields = generateRandomStringArray(arraySize, stringSize, true);
 
         long version = Math.abs(randomLong());
         VersionType versionType = RandomPicks.randomFrom(new Random(), VersionType.values());
@@ -67,11 +57,11 @@ public class ItemSerializationTests extends ElasticsearchTestCase {
                 fetchSourceContext = new FetchSourceContext(randomBoolean());
                 break;
             case 1 :
-                fetchSourceContext = new FetchSourceContext(generateRandomStringArray(arraySize, stringSize));
+                fetchSourceContext = new FetchSourceContext(generateRandomStringArray(arraySize, stringSize, true));
                 break;
             case 2 :
-                fetchSourceContext = new FetchSourceContext(generateRandomStringArray(arraySize, stringSize),
-                        generateRandomStringArray(arraySize, stringSize));
+                fetchSourceContext = new FetchSourceContext(generateRandomStringArray(arraySize, stringSize, true),
+                        generateRandomStringArray(arraySize, stringSize, true));
                 break;
             default:
                 fetchSourceContext = null;

src/test/java/org/elasticsearch/index/query/SimpleIndexQueryParserTests.java

@@ -1701,10 +1701,11 @@ public class SimpleIndexQueryParserTests extends ElasticsearchTestCase {
         // check each clause is for each item
         BooleanClause[] boolClauses = booleanQuery.getClauses();
         for (int i=0; i<likeTexts.size(); i++) {
-            assertThat(boolClauses[i].getOccur(), is(BooleanClause.Occur.SHOULD));
-            assertThat(boolClauses[i].getQuery(), instanceOf(MoreLikeThisQuery.class));
-            MoreLikeThisQuery mltQuery = (MoreLikeThisQuery) boolClauses[i].getQuery();
-            assertThat(mltQuery.getLikeText(), is(likeTexts.get(i).text));
+            BooleanClause booleanClause = booleanQuery.getClauses()[i];
+            assertThat(booleanClause.getOccur(), is(BooleanClause.Occur.SHOULD));
+            assertThat(booleanClause.getQuery(), instanceOf(MoreLikeThisQuery.class));
+            MoreLikeThisQuery mltQuery = (MoreLikeThisQuery) booleanClause.getQuery();
+            assertThat(mltQuery.getLikeTexts(), is(likeTexts.get(i).text));
             assertThat(mltQuery.getMoreLikeFields()[0], equalTo(likeTexts.get(i).field));
         }
 

src/test/java/org/elasticsearch/mlt/MoreLikeThisActionTests.java

@@ -483,4 +483,44 @@ public class MoreLikeThisActionTests extends ElasticsearchIntegrationTest {
         assertHitCount(mltResponse, numOfTypes);
     }
 
+    @Test
+    public void testMoreLikeThisMultiValueFields() throws Exception {
+        logger.info("Creating the index ...");
+        assertAcked(prepareCreate("test")
+                .addMapping("type1", "text", "type=string,analyzer=keyword")
+                .setSettings(SETTING_NUMBER_OF_SHARDS, 1));
+        ensureGreen();
+
+        logger.info("Indexing ...");
+        String[] values = {"aaaa", "bbbb", "cccc", "dddd", "eeee", "ffff", "gggg", "hhhh", "iiii", "jjjj"};
+        List<IndexRequestBuilder> builders = new ArrayList<>(values.length + 1);
+        // index one document with all the values
+        builders.add(client().prepareIndex("test", "type1", "0").setSource("text", values));
+        // index each document with only one of the values
+        for (int i = 0; i < values.length; i++) {
+            builders.add(client().prepareIndex("test", "type1", String.valueOf(i + 1)).setSource("text", values[i]));
+        }
+        indexRandom(true, builders);
+
+        int maxIters = randomIntBetween(10, 20);
+        for (int i = 0; i < maxIters; i++)
+        {
+            int max_query_terms = randomIntBetween(1, values.length);
+            logger.info("Running More Like This with max_query_terms = %s", max_query_terms);
+            MoreLikeThisQueryBuilder mltQuery = moreLikeThisQuery("text").ids("0").minTermFreq(1).minDocFreq(1)
+                    .maxQueryTerms(max_query_terms).percentTermsToMatch(0);
+            SearchResponse response = client().prepareSearch("test").setTypes("type1")
+                    .setQuery(mltQuery).execute().actionGet();
+            assertSearchResponse(response);
+            assertHitCount(response, max_query_terms);
+
+            logger.info("Running More Like This API with with max_query_terms = %s returns all docs!", max_query_terms);
+            response = client().moreLikeThis(moreLikeThisRequest("test").type("type1")
+                    .id("0").fields("text").minTermFreq(1).minDocFreq(1)
+                    .maxQueryTerms(max_query_terms).percentTermsToMatch(0))
+                    .actionGet();
+            assertSearchResponse(response);
+            assertHitCount(response, values.length);
+        }
+    }
 }

src/test/java/org/elasticsearch/test/ElasticsearchTestCase.java

@@ -294,4 +294,19 @@ public abstract class ElasticsearchTestCase extends AbstractRandomizedTest {
     public static <T> T randomFrom(T... values) {
         return RandomizedTest.randomFrom(values);
     }
+
+    public static String[] generateRandomStringArray(int maxArraySize, int maxStringSize, boolean allowNull) {
+        if (allowNull && randomBoolean()) {
+            return null;
+        }
+        String[] array = new String[randomInt(maxArraySize)]; // allow empty arrays
+        for (int i = 0; i < array.length; i++) {
+            array[i] = randomAsciiOfLength(maxStringSize);
+        }
+        return array;
+    }
+
+    public static String[] generateRandomStringArray(int maxArraySize, int maxStringSize) {
+        return generateRandomStringArray(maxArraySize, maxStringSize, false);
+    }
 }