Allow field types to optimize phrase prefix queries (#37575)

This change adds a way to customize how phrase prefix queries should be created
on field types. The match phrase prefix query is exposed in field types in order
to allow optimizations based on the options set on the field.
For instance the text field uses the configured prefix field (if available) to
build a span near that mixes the original field and the prefix field on the last
position.
This change also contains a small refactoring of the match/multi_match query that
simplifies the interactions between the builders.

Closes #31921

Author: jimczi

plugins/mapper-annotated-text/src/main/java/org/elasticsearch/index/mapper/annotatedtext/AnnotatedTextFieldMapper.java

@@ -30,17 +30,17 @@
 import org.apache.lucene.analysis.tokenattributes.OffsetAttribute;
 import org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute;
 import org.apache.lucene.analysis.tokenattributes.PositionLengthAttribute;
-import org.apache.lucene.analysis.tokenattributes.TermToBytesRefAttribute;
 import org.apache.lucene.analysis.tokenattributes.TypeAttribute;
 import org.apache.lucene.document.Field;
 import org.apache.lucene.index.IndexOptions;
 import org.apache.lucene.index.IndexableField;
 import org.apache.lucene.index.Term;
-import org.apache.lucene.search.MultiPhraseQuery;
 import org.apache.lucene.search.NormsFieldExistsQuery;
-import org.apache.lucene.search.PhraseQuery;
+import org.apache.lucene.search.PrefixQuery;
 import org.apache.lucene.search.Query;
 import org.apache.lucene.search.TermQuery;
+import org.apache.lucene.search.spans.SpanMultiTermQueryWrapper;
+import org.apache.lucene.search.spans.SpanQuery;
 import org.elasticsearch.ElasticsearchParseException;
 import org.elasticsearch.common.settings.Settings;
 import org.elasticsearch.common.xcontent.XContentBuilder;
@@ -657,63 +657,12 @@ public Query existsQuery(QueryShardContext context) {
         }
 
         @Override
-        public Query phraseQuery(String field, TokenStream stream, int slop, boolean enablePosIncrements) throws IOException {
-            PhraseQuery.Builder builder = new PhraseQuery.Builder();
-            builder.setSlop(slop);
-
-            TermToBytesRefAttribute termAtt = stream.getAttribute(TermToBytesRefAttribute.class);
-            PositionIncrementAttribute posIncrAtt = stream.getAttribute(PositionIncrementAttribute.class);
-            int position = -1;
-
-            stream.reset();
-            while (stream.incrementToken()) {
-                if (enablePosIncrements) {
-                    position += posIncrAtt.getPositionIncrement();
-                }
-                else {
-                    position += 1;
-                }
-                builder.add(new Term(field, termAtt.getBytesRef()), position);
-            }
-
-            return builder.build();
+        public SpanQuery spanPrefixQuery(String value, SpanMultiTermQueryWrapper.SpanRewriteMethod method, QueryShardContext context) {
+            SpanMultiTermQueryWrapper<?> spanMulti =
+                new SpanMultiTermQueryWrapper<>(new PrefixQuery(new Term(name(), indexedValueForSearch(value))));
+            spanMulti.setRewriteMethod(method);
+            return spanMulti;
         }
-        
-        @Override
-        public Query multiPhraseQuery(String field, TokenStream stream, int slop, boolean enablePositionIncrements) throws IOException {
-
-            MultiPhraseQuery.Builder mpqb = new MultiPhraseQuery.Builder();
-            mpqb.setSlop(slop);
-
-            TermToBytesRefAttribute termAtt = stream.getAttribute(TermToBytesRefAttribute.class);
-
-            PositionIncrementAttribute posIncrAtt = stream.getAttribute(PositionIncrementAttribute.class);
-            int position = -1;
-
-            List<Term> multiTerms = new ArrayList<>();
-            stream.reset();
-            while (stream.incrementToken()) {
-                int positionIncrement = posIncrAtt.getPositionIncrement();
-
-                if (positionIncrement > 0 && multiTerms.size() > 0) {
-                    if (enablePositionIncrements) {
-                        mpqb.add(multiTerms.toArray(new Term[0]), position);
-                    } else {
-                        mpqb.add(multiTerms.toArray(new Term[0]));
-                    }
-                    multiTerms.clear();
-                }
-                position += positionIncrement;
-                multiTerms.add(new Term(field, termAtt.getBytesRef()));
-            }
-
-            if (enablePositionIncrements) {
-                mpqb.add(multiTerms.toArray(new Term[0]), position);
-            } else {
-                mpqb.add(multiTerms.toArray(new Term[0]));
-            }
-            return mpqb.build();
-        }        
     }
 
     private int positionIncrementGap;

server/src/main/java/org/apache/lucene/queries/ExtendedCommonTermsQuery.java

@@ -19,12 +19,8 @@
 
 package org.apache.lucene.queries;
 
-import org.apache.lucene.index.Term;
-import org.apache.lucene.index.TermContext;
 import org.apache.lucene.search.BooleanClause.Occur;
-import org.apache.lucene.search.Query;
 import org.elasticsearch.common.lucene.search.Queries;
-import org.elasticsearch.index.mapper.MappedFieldType;
 
 /**
  * Extended version of {@link CommonTermsQuery} that allows to pass in a
@@ -33,11 +29,8 @@
  */
 public class ExtendedCommonTermsQuery extends CommonTermsQuery {
 
-    private final MappedFieldType fieldType;
-
-    public ExtendedCommonTermsQuery(Occur highFreqOccur, Occur lowFreqOccur, float maxTermFrequency, MappedFieldType fieldType) {
+    public ExtendedCommonTermsQuery(Occur highFreqOccur, Occur lowFreqOccur, float maxTermFrequency) {
         super(highFreqOccur, lowFreqOccur, maxTermFrequency);
-        this.fieldType = fieldType;
     }
 
     private String lowFreqMinNumShouldMatchSpec;
@@ -80,16 +73,4 @@ public float getMaxTermFrequency() {
         return this.maxTermFrequency;
     }
 
-    @Override
-    protected Query newTermQuery(Term term, TermContext context) {
-        if (fieldType == null) {
-            return super.newTermQuery(term, context);
-        }
-        final Query query = fieldType.queryStringTermQuery(term);
-        if (query == null) {
-            return super.newTermQuery(term, context);
-        } else {
-            return query;
-        }
-    }
 }

server/src/main/java/org/elasticsearch/common/lucene/search/MultiPhrasePrefixQuery.java

@@ -39,16 +39,21 @@
 import java.util.Iterator;
 import java.util.List;
 import java.util.ListIterator;
+import java.util.Objects;
 
 public class MultiPhrasePrefixQuery extends Query {
 
-    private String field;
+    private final String field;
     private ArrayList<Term[]> termArrays = new ArrayList<>();
     private ArrayList<Integer> positions = new ArrayList<>();
     private int maxExpansions = Integer.MAX_VALUE;
 
     private int slop = 0;
 
+    public MultiPhrasePrefixQuery(String field) {
+        this.field = Objects.requireNonNull(field);
+    }
+
     /**
      * Sets the phrase slop for this query.
      *
@@ -102,9 +107,6 @@ public void add(Term[] terms) {
      * @see org.apache.lucene.search.PhraseQuery.Builder#add(Term, int)
      */
     public void add(Term[] terms, int position) {
-        if (termArrays.size() == 0)
-            field = terms[0].field();
-
         for (int i = 0; i < terms.length; i++) {
             if (terms[i].field() != field) {
                 throw new IllegalArgumentException(
@@ -212,7 +214,7 @@ private void getPrefixTerms(ObjectHashSet<Term> terms, final Term prefix, final
     @Override
     public final String toString(String f) {
         StringBuilder buffer = new StringBuilder();
-        if (field == null || !field.equals(f)) {
+        if (field.equals(f) == false) {
             buffer.append(field);
             buffer.append(":");
         }

server/src/main/java/org/elasticsearch/common/lucene/search/SpanBooleanQueryRewriteWithMaxClause.java

@@ -0,0 +1,119 @@
+/*
+ * 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.
+ */
+
+package org.elasticsearch.common.lucene.search;
+
+import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.index.IndexReaderContext;
+import org.apache.lucene.index.LeafReaderContext;
+import org.apache.lucene.index.Term;
+import org.apache.lucene.index.Terms;
+import org.apache.lucene.index.TermsEnum;
+import org.apache.lucene.queries.SpanMatchNoDocsQuery;
+import org.apache.lucene.search.BooleanQuery;
+import org.apache.lucene.search.MultiTermQuery;
+import org.apache.lucene.search.Query;
+import org.apache.lucene.search.spans.SpanMultiTermQueryWrapper;
+import org.apache.lucene.search.spans.SpanOrQuery;
+import org.apache.lucene.search.spans.SpanQuery;
+import org.apache.lucene.search.spans.SpanTermQuery;
+import org.apache.lucene.util.BytesRef;
+
+import java.io.IOException;
+import java.util.Collection;
+import java.util.HashSet;
+import java.util.Set;
+
+/**
+ * A span rewrite method that extracts the first <code>maxExpansions</code> terms
+ * that match the {@link MultiTermQuery} in the terms dictionary.
+ * The rewrite throws an error if more than <code>maxExpansions</code> terms are found and <code>hardLimit</code>
+ * is set.
+ */
+public class SpanBooleanQueryRewriteWithMaxClause extends SpanMultiTermQueryWrapper.SpanRewriteMethod {
+    private final int maxExpansions;
+    private final boolean hardLimit;
+
+    public SpanBooleanQueryRewriteWithMaxClause() {
+        this(BooleanQuery.getMaxClauseCount(), true);
+    }
+
+    public SpanBooleanQueryRewriteWithMaxClause(int maxExpansions, boolean hardLimit) {
+        this.maxExpansions = maxExpansions;
+        this.hardLimit = hardLimit;
+    }
+
+    public int getMaxExpansions() {
+        return maxExpansions;
+    }
+
+    public boolean isHardLimit() {
+        return hardLimit;
+    }
+
+    @Override
+    public SpanQuery rewrite(IndexReader reader, MultiTermQuery query) throws IOException {
+        final MultiTermQuery.RewriteMethod delegate = new MultiTermQuery.RewriteMethod() {
+            @Override
+            public Query rewrite(IndexReader reader, MultiTermQuery query) throws IOException {
+                Collection<SpanQuery> queries = collectTerms(reader, query);
+                if (queries.size() == 0) {
+                    return new SpanMatchNoDocsQuery(query.getField(), "no expansion found for " + query.toString());
+                } else if (queries.size() == 1) {
+                    return queries.iterator().next();
+                } else {
+                    return new SpanOrQuery(queries.toArray(new SpanQuery[0]));
+                }
+            }
+
+            private Collection<SpanQuery> collectTerms(IndexReader reader, MultiTermQuery query) throws IOException {
+                Set<SpanQuery> queries = new HashSet<>();
+                IndexReaderContext topReaderContext = reader.getContext();
+                for (LeafReaderContext context : topReaderContext.leaves()) {
+                    final Terms terms = context.reader().terms(query.getField());
+                    if (terms == null) {
+                        // field does not exist
+                        continue;
+                    }
+
+                    final TermsEnum termsEnum = getTermsEnum(query, terms, null);
+                    assert termsEnum != null;
+
+                    if (termsEnum == TermsEnum.EMPTY)
+                        continue;
+
+                    BytesRef bytes;
+                    while ((bytes = termsEnum.next()) != null) {
+                        if (queries.size() >= maxExpansions) {
+                            if (hardLimit) {
+                                throw new RuntimeException("[" + query.toString() + " ] " +
+                                    "exceeds maxClauseCount [ Boolean maxClauseCount is set to " + BooleanQuery.getMaxClauseCount() + "]");
+                            } else {
+                                return queries;
+                            }
+                        }
+                        queries.add(new SpanTermQuery(new Term(query.getField(), bytes)));
+                    }
+                }
+                return queries;
+            }
+        };
+        return (SpanQuery) delegate.rewrite(reader, query);
+    }
+}

server/src/main/java/org/elasticsearch/index/mapper/MappedFieldType.java

@@ -34,6 +34,8 @@
 import org.apache.lucene.search.Query;
 import org.apache.lucene.search.TermInSetQuery;
 import org.apache.lucene.search.TermQuery;
+import org.apache.lucene.search.spans.SpanMultiTermQueryWrapper;
+import org.apache.lucene.search.spans.SpanQuery;
 import org.apache.lucene.util.BytesRef;
 import org.elasticsearch.ElasticsearchParseException;
 import org.elasticsearch.common.Nullable;
@@ -399,14 +401,24 @@ public Query nullValueQuery() {
 
     public abstract Query existsQuery(QueryShardContext context);
 
-    public Query phraseQuery(String field, TokenStream stream, int slop, boolean enablePositionIncrements) throws IOException {
+    public Query phraseQuery(TokenStream stream, int slop, boolean enablePositionIncrements) throws IOException {
         throw new IllegalArgumentException("Attempted to build a phrase query with multiple terms against non-text field [" + name + "]");
     }
 
-    public Query multiPhraseQuery(String field, TokenStream stream, int slop, boolean enablePositionIncrements) throws IOException {
+    public Query multiPhraseQuery(TokenStream stream, int slop, boolean enablePositionIncrements) throws IOException {
         throw new IllegalArgumentException("Attempted to build a phrase query with multiple terms against non-text field [" + name + "]");
     }
 
+    public Query phrasePrefixQuery(TokenStream stream, int slop, int maxExpansions) throws IOException {
+        throw new IllegalArgumentException("Attempted to build a phrase prefix query with multiple terms against non-text field [" + name +
+            "]");
+    }
+
+    public SpanQuery spanPrefixQuery(String value, SpanMultiTermQueryWrapper.SpanRewriteMethod method, QueryShardContext context) {
+        throw new IllegalArgumentException("Can only use span prefix queries on text fields - not on [" + name
+            + "] which is of type [" + typeName() + "]");
+    }
+
     /**
      * An enum used to describe the relation between the range of terms in a
      * shard when compared with a query range