1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
|
/*
* 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.search.aggregations.bucket.significant;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.LeafReaderContext;
import org.apache.lucene.util.BytesRef;
import org.elasticsearch.common.lease.Releasables;
import org.elasticsearch.common.util.LongHash;
import org.elasticsearch.search.DocValueFormat;
import org.elasticsearch.search.aggregations.Aggregator;
import org.elasticsearch.search.aggregations.AggregatorFactories;
import org.elasticsearch.search.aggregations.LeafBucketCollector;
import org.elasticsearch.search.aggregations.LeafBucketCollectorBase;
import org.elasticsearch.search.aggregations.bucket.significant.heuristics.SignificanceHeuristic;
import org.elasticsearch.search.aggregations.bucket.terms.GlobalOrdinalsStringTermsAggregator;
import org.elasticsearch.search.aggregations.bucket.terms.support.IncludeExclude;
import org.elasticsearch.search.aggregations.pipeline.PipelineAggregator;
import org.elasticsearch.search.aggregations.support.ValuesSource;
import org.elasticsearch.search.internal.ContextIndexSearcher;
import org.elasticsearch.search.internal.SearchContext;
import java.io.IOException;
import java.util.Arrays;
import java.util.List;
import java.util.Map;
import static java.util.Collections.emptyList;
/**
* An global ordinal based implementation of significant terms, based on {@link SignificantStringTermsAggregator}.
*/
public class GlobalOrdinalsSignificantTermsAggregator extends GlobalOrdinalsStringTermsAggregator {
protected long numCollectedDocs;
protected final SignificantTermsAggregatorFactory termsAggFactory;
private final SignificanceHeuristic significanceHeuristic;
public GlobalOrdinalsSignificantTermsAggregator(String name, AggregatorFactories factories,
ValuesSource.Bytes.WithOrdinals.FieldData valuesSource, DocValueFormat format,
BucketCountThresholds bucketCountThresholds, IncludeExclude.OrdinalsFilter includeExclude,
SearchContext context, Aggregator parent,
SignificanceHeuristic significanceHeuristic, SignificantTermsAggregatorFactory termsAggFactory,
List<PipelineAggregator> pipelineAggregators, Map<String, Object> metaData) throws IOException {
super(name, factories, valuesSource, null, format, bucketCountThresholds, includeExclude, context, parent,
SubAggCollectionMode.DEPTH_FIRST, false, pipelineAggregators, metaData);
this.significanceHeuristic = significanceHeuristic;
this.termsAggFactory = termsAggFactory;
}
@Override
public LeafBucketCollector getLeafCollector(LeafReaderContext ctx,
final LeafBucketCollector sub) throws IOException {
return new LeafBucketCollectorBase(super.getLeafCollector(ctx, sub), null) {
@Override
public void collect(int doc, long bucket) throws IOException {
super.collect(doc, bucket);
numCollectedDocs++;
}
};
}
@Override
public SignificantStringTerms buildAggregation(long owningBucketOrdinal) throws IOException {
assert owningBucketOrdinal == 0;
if (globalOrds == null) { // no context in this reader
return buildEmptyAggregation();
}
final int size;
if (bucketCountThresholds.getMinDocCount() == 0) {
// if minDocCount == 0 then we can end up with more buckets then maxBucketOrd() returns
size = (int) Math.min(globalOrds.getValueCount(), bucketCountThresholds.getShardSize());
} else {
size = (int) Math.min(maxBucketOrd(), bucketCountThresholds.getShardSize());
}
long supersetSize = termsAggFactory.getSupersetNumDocs();
long subsetSize = numCollectedDocs;
BucketSignificancePriorityQueue<SignificantStringTerms.Bucket> ordered = new BucketSignificancePriorityQueue<>(size);
SignificantStringTerms.Bucket spare = null;
for (long globalTermOrd = 0; globalTermOrd < globalOrds.getValueCount(); ++globalTermOrd) {
if (includeExclude != null && !acceptedGlobalOrdinals.get(globalTermOrd)) {
continue;
}
final long bucketOrd = getBucketOrd(globalTermOrd);
final int bucketDocCount = bucketOrd < 0 ? 0 : bucketDocCount(bucketOrd);
if (bucketCountThresholds.getMinDocCount() > 0 && bucketDocCount == 0) {
continue;
}
if (bucketDocCount < bucketCountThresholds.getShardMinDocCount()) {
continue;
}
if (spare == null) {
spare = new SignificantStringTerms.Bucket(new BytesRef(), 0, 0, 0, 0, null, format);
}
spare.bucketOrd = bucketOrd;
copy(globalOrds.lookupOrd(globalTermOrd), spare.termBytes);
spare.subsetDf = bucketDocCount;
spare.subsetSize = subsetSize;
spare.supersetDf = termsAggFactory.getBackgroundFrequency(spare.termBytes);
spare.supersetSize = supersetSize;
// During shard-local down-selection we use subset/superset stats
// that are for this shard only
// Back at the central reducer these properties will be updated with
// global stats
spare.updateScore(significanceHeuristic);
spare = ordered.insertWithOverflow(spare);
}
final SignificantStringTerms.Bucket[] list = new SignificantStringTerms.Bucket[ordered.size()];
for (int i = ordered.size() - 1; i >= 0; i--) {
final SignificantStringTerms.Bucket bucket = ordered.pop();
// the terms are owned by the BytesRefHash, we need to pull a copy since the BytesRef hash data may be recycled at some point
bucket.termBytes = BytesRef.deepCopyOf(bucket.termBytes);
bucket.aggregations = bucketAggregations(bucket.bucketOrd);
list[i] = bucket;
}
return new SignificantStringTerms(name, bucketCountThresholds.getRequiredSize(), bucketCountThresholds.getMinDocCount(),
pipelineAggregators(), metaData(), format, subsetSize, supersetSize, significanceHeuristic, Arrays.asList(list));
}
@Override
public SignificantStringTerms buildEmptyAggregation() {
// We need to account for the significance of a miss in our global stats - provide corpus size as context
ContextIndexSearcher searcher = context.searcher();
IndexReader topReader = searcher.getIndexReader();
int supersetSize = topReader.numDocs();
return new SignificantStringTerms(name, bucketCountThresholds.getRequiredSize(), bucketCountThresholds.getMinDocCount(),
pipelineAggregators(), metaData(), format, 0, supersetSize, significanceHeuristic, emptyList());
}
@Override
protected void doClose() {
Releasables.close(termsAggFactory);
}
public static class WithHash extends GlobalOrdinalsSignificantTermsAggregator {
private final LongHash bucketOrds;
public WithHash(String name, AggregatorFactories factories, ValuesSource.Bytes.WithOrdinals.FieldData valuesSource,
DocValueFormat format, BucketCountThresholds bucketCountThresholds, IncludeExclude.OrdinalsFilter includeExclude,
SearchContext context, Aggregator parent, SignificanceHeuristic significanceHeuristic,
SignificantTermsAggregatorFactory termsAggFactory, List<PipelineAggregator> pipelineAggregators,
Map<String, Object> metaData) throws IOException {
super(name, factories, valuesSource, format, bucketCountThresholds, includeExclude, context, parent, significanceHeuristic,
termsAggFactory, pipelineAggregators, metaData);
bucketOrds = new LongHash(1, context.bigArrays());
}
@Override
public LeafBucketCollector getLeafCollector(LeafReaderContext ctx,
final LeafBucketCollector sub) throws IOException {
return new LeafBucketCollectorBase(super.getLeafCollector(ctx, sub), null) {
@Override
public void collect(int doc, long bucket) throws IOException {
assert bucket == 0;
numCollectedDocs++;
globalOrds.setDocument(doc);
final int numOrds = globalOrds.cardinality();
for (int i = 0; i < numOrds; i++) {
final long globalOrd = globalOrds.ordAt(i);
long bucketOrd = bucketOrds.add(globalOrd);
if (bucketOrd < 0) {
bucketOrd = -1 - bucketOrd;
collectExistingBucket(sub, doc, bucketOrd);
} else {
collectBucket(sub, doc, bucketOrd);
}
}
}
};
}
@Override
protected long getBucketOrd(long termOrd) {
return bucketOrds.find(termOrd);
}
@Override
protected void doClose() {
Releasables.close(termsAggFactory, bucketOrds);
}
}
}
|
