Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Clarify what query components cross_fields supports #68795

Merged
merged 3 commits into from
Feb 18, 2021
Merged

Clarify what query components cross_fields supports #68795

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
3068268
Clarify that cross-fields only supports boolean queries.
jtibshirani Feb 9, 2021
c5545d8
Rename BlendedQueryBuilder -> CrossFieldsQueryBuilder.
jtibshirani Feb 9, 2021
27f5506
Make sure aliases are handled correctly.
jtibshirani Feb 10, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
23 changes: 8 additions & 15 deletions server/src/main/java/org/elasticsearch/index/search/MatchQueryParser.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -241,6 +241,8 @@ public Query parse(Type type, String fieldName, Object value) throws IOException
assert analyzer != null;

MatchQueryBuilder builder = new MatchQueryBuilder(analyzer, fieldType, enablePositionIncrements, autoGenerateSynonymsPhraseQuery);
String resolvedFieldName = fieldType.name();
String stringValue = value.toString();

/*
* If a keyword analyzer is used, we know that further analysis isn't
Expand All @@ -249,7 +251,7 @@ public Query parse(Type type, String fieldName, Object value) throws IOException
* a prefix query instead
*/
if (analyzer == Lucene.KEYWORD_ANALYZER && type != Type.PHRASE_PREFIX) {
final Term term = new Term(fieldType.name(), value.toString());
final Term term = new Term(resolvedFieldName, stringValue);
if (type == Type.BOOLEAN_PREFIX
&& (fieldType instanceof TextFieldMapper.TextFieldType || fieldType instanceof KeywordFieldMapper.KeywordFieldType)) {
return builder.newPrefixQuery(term);
Expand All @@ -258,32 +260,23 @@ public Query parse(Type type, String fieldName, Object value) throws IOException
}
}

return parseInternal(type, fieldType.name(), builder, value);
}

protected final Query parseInternal(Type type, String fieldName, MatchQueryBuilder builder, Object value) throws IOException {
final Query query;
Query query;
switch (type) {
case BOOLEAN:
query = builder.createBooleanQuery(fieldName, value.toString(), occur);
query = builder.createBooleanQuery(resolvedFieldName, stringValue, occur);
break;

case BOOLEAN_PREFIX:
query = builder.createBooleanPrefixQuery(fieldName, value.toString(), occur);
query = builder.createBooleanPrefixQuery(resolvedFieldName, stringValue, occur);
break;

case PHRASE:
query = builder.createPhraseQuery(fieldName, value.toString(), phraseSlop);
query = builder.createPhraseQuery(resolvedFieldName, stringValue, phraseSlop);
break;

case PHRASE_PREFIX:
query = builder.createPhrasePrefixQuery(fieldName, value.toString(), phraseSlop);
query = builder.createPhrasePrefixQuery(resolvedFieldName, stringValue, phraseSlop);
break;

default:
throw new IllegalStateException("No type found for [" + type + "]");
}

return query == null ? zeroTermsQuery() : query;
}

Expand Down
47 changes: 30 additions & 17 deletions server/src/main/java/org/elasticsearch/index/search/MultiMatchQueryParser.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,7 @@
import org.apache.lucene.analysis.TokenStream;
import org.apache.lucene.index.Term;
import org.apache.lucene.queries.BlendedTermQuery;
import org.apache.lucene.search.BooleanClause;
import org.apache.lucene.search.BoostQuery;
import org.apache.lucene.search.DisjunctionMaxQuery;
import org.apache.lucene.search.MatchNoDocsQuery;
Expand Down Expand Up @@ -66,7 +67,7 @@ public Query parse(MultiMatchQueryBuilder.Type type, Map<String, Float> fieldNam
break;

case CROSS_FIELDS:
queries = buildCrossFieldQuery(type, fieldNames, value, minimumShouldMatch, tieBreaker);
queries = buildCrossFieldQuery(fieldNames, value, minimumShouldMatch, tieBreaker);
break;

default:
Expand Down Expand Up @@ -108,15 +109,16 @@ private List<Query> buildFieldQueries(MultiMatchQueryBuilder.Type type, Map<Stri
return queries;
}

private List<Query> buildCrossFieldQuery(MultiMatchQueryBuilder.Type type, Map<String, Float> fieldNames,
Object value, String minimumShouldMatch, float tieBreaker) throws IOException {
private List<Query> buildCrossFieldQuery(Map<String, Float> fieldNames,
Object value, String minimumShouldMatch, float tieBreaker) {

Map<Analyzer, List<FieldAndBoost>> groups = new HashMap<>();
List<Query> queries = new ArrayList<>();
for (Map.Entry<String, Float> entry : fieldNames.entrySet()) {
String name = entry.getKey();
MappedFieldType fieldType = context.getFieldType(name);
if (fieldType != null) {
Analyzer actualAnalyzer = getAnalyzer(fieldType, type == MultiMatchQueryBuilder.Type.PHRASE);
Analyzer actualAnalyzer = getAnalyzer(fieldType, false);
if (groups.containsKey(actualAnalyzer) == false) {
groups.put(actualAnalyzer, new ArrayList<>());
}
Expand All @@ -130,7 +132,7 @@ private List<Query> buildCrossFieldQuery(MultiMatchQueryBuilder.Type type, Map<S
builder = new MatchQueryBuilder(group.getKey(), group.getValue().get(0).fieldType,
enablePositionIncrements, autoGenerateSynonymsPhraseQuery);
} else {
builder = new BlendedQueryBuilder(group.getKey(), group.getValue(), tieBreaker,
builder = new CrossFieldsQueryBuilder(group.getKey(), group.getValue(), tieBreaker,
enablePositionIncrements, autoGenerateSynonymsPhraseQuery);
}

Expand All @@ -140,7 +142,11 @@ private List<Query> buildCrossFieldQuery(MultiMatchQueryBuilder.Type type, Map<S
* fields are already grouped by their analyzers/types.
*/
String representativeField = group.getValue().get(0).fieldType.name();
Query query = parseInternal(type.matchQueryType(), representativeField, builder, value);
Query query = builder.createBooleanQuery(representativeField, value.toString(), occur);
if (query == null) {
query = zeroTermsQuery();
}

query = Queries.maybeApplyMinimumShouldMatch(query, minimumShouldMatch);
if (query != null) {
if (group.getValue().size() == 1) {
Expand All @@ -157,17 +163,32 @@ private List<Query> buildCrossFieldQuery(MultiMatchQueryBuilder.Type type, Map<S
return queries;
}

private class BlendedQueryBuilder extends MatchQueryBuilder {
private class CrossFieldsQueryBuilder extends MatchQueryBuilder {
private final List<FieldAndBoost> blendedFields;
private final float tieBreaker;

BlendedQueryBuilder(Analyzer analyzer, List<FieldAndBoost> blendedFields, float tieBreaker,
CrossFieldsQueryBuilder(Analyzer analyzer, List<FieldAndBoost> blendedFields, float tieBreaker,
boolean enablePositionIncrements, boolean autoGenerateSynonymsPhraseQuery) {
super(analyzer, blendedFields.get(0).fieldType, enablePositionIncrements, autoGenerateSynonymsPhraseQuery);
this.blendedFields = blendedFields;
this.tieBreaker = tieBreaker;
}

@Override
public Query createPhraseQuery(String field, String queryText, int phraseSlop) {
throw new IllegalArgumentException("[multi_match] queries in [cross_fields] mode don't support phrases");
}

@Override
protected Query createPhrasePrefixQuery(String field, String queryText, int slop) {
throw new IllegalArgumentException("[multi_match] queries in [cross_fields] mode don't support phrase prefix");
}

@Override
protected Query createBooleanPrefixQuery(String field, String queryText, BooleanClause.Occur occur) {
throw new IllegalArgumentException("[multi_match] queries in [cross_fields] mode don't support boolean prefix");
}

@Override
protected Query newSynonymQuery(TermAndBoost[] terms) {
BytesRef[] values = new BytesRef[terms.length];
Expand All @@ -184,15 +205,7 @@ protected Query newTermQuery(Term term, float boost) {

@Override
protected Query newPrefixQuery(Term term) {
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

A couple notes:

  • Although the cross_fields query doesn't allow phrases, the analyzePhrase method must be implemented to support multi-word synonyms with auto_generate_synonyms_phrase_query: true.
  • I think we can remove analyzeMultiPhrase. This method supports phrase queries that contain a term with synonyms, which should never occur with cross_fields. I opted not to remove it since the logic is tricky and I didn't want to risk a regression when we aim to deprecate cross_fields anyways.

List<Query> disjunctions = new ArrayList<>();
for (FieldAndBoost fieldType : blendedFields) {
Query query = fieldType.fieldType.prefixQuery(term.text(), null, context);
if (fieldType.boost != 1f) {
query = new BoostQuery(query, fieldType.boost);
}
disjunctions.add(query);
}
return new DisjunctionMaxQuery(disjunctions, tieBreaker);
throw new IllegalArgumentException("[multi_match] queries in [cross_fields] mode don't support prefix");
}

@Override
Expand Down