@Test
public void givenAnalyzedQuery_whenMakeAggregationOnTermCount_thenEachTokenCountsSeparately() {
TermsBuilder aggregation = AggregationBuilders.terms("top_tags").field("title");
SearchResponse response =
client
.prepareSearch("blog")
.setTypes("article")
.addAggregation(aggregation)
.execute()
.actionGet();
Map<String, Aggregation> results = response.getAggregations().asMap();
StringTerms topTags = (StringTerms) results.get("top_tags");
List<String> keys = topTags.getBuckets().stream().map(b -> b.getKey()).collect(toList());
Collections.sort(keys);
assertEquals(
asList(
"about",
"article",
"data",
"elasticsearch",
"engines",
"search",
"second",
"spring",
"tutorial"),
keys);
}
/*
* 找到所有职员中最大的共同点(兴趣爱好)是什么
*/
@Test
public void aggregationTest() {
Client client = ClientTemplate.getClient();
SearchResponse response =
client
.prepareSearch("fccs")
.setTypes("employee")
.addAggregation(AggregationBuilders.terms("by_interests").field("interests"))
.get();
StringTerms terms = response.getAggregations().get("by_interests");
List<Bucket> buckets = terms.getBuckets();
for (Bucket bucket : buckets) {
System.out.println("key:" + bucket.getKey() + " doc_count:" + bucket.getDocCount());
}
client.close();
}
@Test
public void
givenNotAnalyzedQuery_whenMakeAggregationOnTermCount_thenEachTermCountsIndividually() {
TermsBuilder aggregation =
AggregationBuilders.terms("top_tags")
.field("tags")
.order(Terms.Order.aggregation("_count", false));
SearchResponse response =
client
.prepareSearch("blog")
.setTypes("article")
.addAggregation(aggregation)
.execute()
.actionGet();
Map<String, Aggregation> results = response.getAggregations().asMap();
StringTerms topTags = (StringTerms) results.get("top_tags");
List<String> keys = topTags.getBuckets().stream().map(b -> b.getKey()).collect(toList());
assertEquals(asList("elasticsearch", "spring data", "search engines", "tutorial"), keys);
}
public void testScoresEqualForPositiveAndNegative(SignificanceHeuristicBuilder heuristic)
throws Exception {
// check that results for both classes are the same with exclude negatives = false and classes
// are routing ids
SearchResponse response =
client()
.prepareSearch("test")
.addAggregation(
new TermsBuilder("class")
.field("class")
.subAggregation(
new SignificantTermsBuilder("mySignificantTerms")
.field("text")
.executionHint(randomExecutionHint())
.significanceHeuristic(heuristic)
.minDocCount(1)
.shardSize(1000)
.size(1000)))
.execute()
.actionGet();
assertSearchResponse(response);
StringTerms classes = (StringTerms) response.getAggregations().get("class");
assertThat(classes.getBuckets().size(), equalTo(2));
Iterator<Terms.Bucket> classBuckets = classes.getBuckets().iterator();
Collection<SignificantTerms.Bucket> classA =
((SignificantTerms) classBuckets.next().getAggregations().get("mySignificantTerms"))
.getBuckets();
Iterator<SignificantTerms.Bucket> classBBucketIterator =
((SignificantTerms) classBuckets.next().getAggregations().get("mySignificantTerms"))
.getBuckets()
.iterator();
assertThat(classA.size(), greaterThan(0));
for (SignificantTerms.Bucket classABucket : classA) {
SignificantTerms.Bucket classBBucket = classBBucketIterator.next();
assertThat(classABucket.getKey(), equalTo(classBBucket.getKey()));
assertThat(
classABucket.getSignificanceScore(), closeTo(classBBucket.getSignificanceScore(), 1.e-5));
}
}
public void testNestNestedAggs() throws Exception {
SearchResponse response =
client()
.prepareSearch("idx_nested_nested_aggs")
.addAggregation(
nested("level1", "nested1")
.subAggregation(
terms("a")
.field("nested1.a")
.collectMode(aggCollectionMode)
.subAggregation(
nested("level2", "nested1.nested2")
.subAggregation(sum("sum").field("nested1.nested2.b")))))
.get();
assertSearchResponse(response);
Nested level1 = response.getAggregations().get("level1");
assertThat(level1, notNullValue());
assertThat(level1.getName(), equalTo("level1"));
assertThat(level1.getDocCount(), equalTo(2L));
StringTerms a = level1.getAggregations().get("a");
Terms.Bucket bBucket = a.getBucketByKey("a");
assertThat(bBucket.getDocCount(), equalTo(1L));
Nested level2 = bBucket.getAggregations().get("level2");
assertThat(level2.getDocCount(), equalTo(1L));
Sum sum = level2.getAggregations().get("sum");
assertThat(sum.getValue(), equalTo(2d));
a = level1.getAggregations().get("a");
bBucket = a.getBucketByKey("b");
assertThat(bBucket.getDocCount(), equalTo(1L));
level2 = bBucket.getAggregations().get("level2");
assertThat(level2.getDocCount(), equalTo(1L));
sum = level2.getAggregations().get("sum");
assertThat(sum.getValue(), equalTo(2d));
}
public void testXContentResponse() throws Exception {
String type = randomBoolean() ? "string" : "long";
String settings = "{\"index.number_of_shards\": 1, \"index.number_of_replicas\": 0}";
SharedSignificantTermsTestMethods.index01Docs(type, settings, this);
SearchResponse response =
client()
.prepareSearch(INDEX_NAME)
.setTypes(DOC_TYPE)
.addAggregation(
new TermsBuilder("class")
.field(CLASS_FIELD)
.subAggregation(new SignificantTermsBuilder("sig_terms").field(TEXT_FIELD)))
.execute()
.actionGet();
assertSearchResponse(response);
StringTerms classes = (StringTerms) response.getAggregations().get("class");
assertThat(classes.getBuckets().size(), equalTo(2));
for (Terms.Bucket classBucket : classes.getBuckets()) {
Map<String, Aggregation> aggs = classBucket.getAggregations().asMap();
assertTrue(aggs.containsKey("sig_terms"));
SignificantTerms agg = (SignificantTerms) aggs.get("sig_terms");
assertThat(agg.getBuckets().size(), equalTo(1));
String term = agg.iterator().next().getKeyAsString();
String classTerm = classBucket.getKeyAsString();
assertTrue(term.equals(classTerm));
}
XContentBuilder responseBuilder = XContentFactory.jsonBuilder();
classes.toXContent(responseBuilder, null);
String result = null;
if (type.equals("long")) {
result =
"\"class\"{\"doc_count_error_upper_bound\":0,\"sum_other_doc_count\":0,\"buckets\":[{\"key\":\"0\",\"doc_count\":4,\"sig_terms\":{\"doc_count\":4,\"buckets\":[{\"key\":0,\"key_as_string\":\"0\",\"doc_count\":4,\"score\":0.39999999999999997,\"bg_count\":5}]}},{\"key\":\"1\",\"doc_count\":3,\"sig_terms\":{\"doc_count\":3,\"buckets\":[{\"key\":1,\"key_as_string\":\"1\",\"doc_count\":3,\"score\":0.75,\"bg_count\":4}]}}]}";
} else {
result =
"\"class\"{\"doc_count_error_upper_bound\":0,\"sum_other_doc_count\":0,\"buckets\":[{\"key\":\"0\",\"doc_count\":4,\"sig_terms\":{\"doc_count\":4,\"buckets\":[{\"key\":\"0\",\"doc_count\":4,\"score\":0.39999999999999997,\"bg_count\":5}]}},{\"key\":\"1\",\"doc_count\":3,\"sig_terms\":{\"doc_count\":3,\"buckets\":[{\"key\":\"1\",\"doc_count\":3,\"score\":0.75,\"bg_count\":4}]}}]}";
}
assertThat(responseBuilder.string(), equalTo(result));
}
@Override
public Map<String, Integer> countTermsByField(SearchCriteria searchCriteria, String field) {
ElasticSearchCriteria criteria = (ElasticSearchCriteria) searchCriteria;
if (criteria == null) return null;
SearchRequestBuilder srb = criteria2builder(criteria);
srb.setFrom(0);
srb.setSize(0);
TermsBuilder tb = AggregationBuilders.terms(field);
tb.field(field);
srb.addAggregation(tb);
try {
SearchResponse response = srb.execute().get();
StringTerms aggr = response.getAggregations().get(field);
Map<String, Integer> result = new LinkedHashMap<>();
for (Terms.Bucket bucket : aggr.getBuckets()) {
result.put(bucket.getKey(), (int) bucket.getDocCount());
}
return result;
} catch (Exception e) {
logger.error(e.getMessage(), e);
}
return Collections.emptyMap();
}
@Override
public StringTerms readResult(StreamInput in) throws IOException {
StringTerms buckets = new StringTerms();
buckets.readFrom(in);
return buckets;
}
public void testPlugin() throws Exception {
String type = randomBoolean() ? "string" : "long";
String settings = "{\"index.number_of_shards\": 1, \"index.number_of_replicas\": 0}";
SharedSignificantTermsTestMethods.index01Docs(type, settings, this);
SearchResponse response =
client()
.prepareSearch(INDEX_NAME)
.setTypes(DOC_TYPE)
.addAggregation(
new TermsBuilder("class")
.field(CLASS_FIELD)
.subAggregation(
(new SignificantTermsBuilder("sig_terms"))
.field(TEXT_FIELD)
.significanceHeuristic(new SimpleHeuristic.SimpleHeuristicBuilder())
.minDocCount(1)))
.execute()
.actionGet();
assertSearchResponse(response);
StringTerms classes = (StringTerms) response.getAggregations().get("class");
assertThat(classes.getBuckets().size(), equalTo(2));
for (Terms.Bucket classBucket : classes.getBuckets()) {
Map<String, Aggregation> aggs = classBucket.getAggregations().asMap();
assertTrue(aggs.containsKey("sig_terms"));
SignificantTerms agg = (SignificantTerms) aggs.get("sig_terms");
assertThat(agg.getBuckets().size(), equalTo(2));
Iterator<SignificantTerms.Bucket> bucketIterator = agg.iterator();
SignificantTerms.Bucket sigBucket = bucketIterator.next();
String term = sigBucket.getKeyAsString();
String classTerm = classBucket.getKeyAsString();
assertTrue(term.equals(classTerm));
assertThat(sigBucket.getSignificanceScore(), closeTo(2.0, 1.e-8));
sigBucket = bucketIterator.next();
assertThat(sigBucket.getSignificanceScore(), closeTo(1.0, 1.e-8));
}
// we run the same test again but this time we do not call assertSearchResponse() before the
// assertions
// the reason is that this would trigger toXContent and we would like to check that this has no
// potential side effects
response =
client()
.prepareSearch(INDEX_NAME)
.setTypes(DOC_TYPE)
.addAggregation(
new TermsBuilder("class")
.field(CLASS_FIELD)
.subAggregation(
(new SignificantTermsBuilder("sig_terms"))
.field(TEXT_FIELD)
.significanceHeuristic(new SimpleHeuristic.SimpleHeuristicBuilder())
.minDocCount(1)))
.execute()
.actionGet();
classes = (StringTerms) response.getAggregations().get("class");
assertThat(classes.getBuckets().size(), equalTo(2));
for (Terms.Bucket classBucket : classes.getBuckets()) {
Map<String, Aggregation> aggs = classBucket.getAggregations().asMap();
assertTrue(aggs.containsKey("sig_terms"));
SignificantTerms agg = (SignificantTerms) aggs.get("sig_terms");
assertThat(agg.getBuckets().size(), equalTo(2));
Iterator<SignificantTerms.Bucket> bucketIterator = agg.iterator();
SignificantTerms.Bucket sigBucket = bucketIterator.next();
String term = sigBucket.getKeyAsString();
String classTerm = classBucket.getKeyAsString();
assertTrue(term.equals(classTerm));
assertThat(sigBucket.getSignificanceScore(), closeTo(2.0, 1.e-8));
sigBucket = bucketIterator.next();
assertThat(sigBucket.getSignificanceScore(), closeTo(1.0, 1.e-8));
}
}