Merge branch 'master' into feature/client_aggs_parsing

This commit is contained in:
Christoph Büscher 2017-04-21 12:18:06 +02:00
commit 4a69b658cd
17 changed files with 647 additions and 138 deletions

View File

@ -57,7 +57,7 @@ import java.util.Collection;
import java.util.HashMap; import java.util.HashMap;
import java.util.List; import java.util.List;
import java.util.Map; import java.util.Map;
import java.util.Collections;
import static java.util.Collections.unmodifiableMap; import static java.util.Collections.unmodifiableMap;
import static org.elasticsearch.common.lucene.search.Queries.fixNegativeQueryIfNeeded; import static org.elasticsearch.common.lucene.search.Queries.fixNegativeQueryIfNeeded;
@ -91,7 +91,8 @@ public class MapperQueryParser extends QueryParser {
public void reset(QueryParserSettings settings) { public void reset(QueryParserSettings settings) {
this.settings = settings; this.settings = settings;
if (settings.fieldsAndWeights().isEmpty()) { if (settings.fieldsAndWeights() == null) {
// this query has no explicit fields to query so we fallback to the default field
this.field = settings.defaultField(); this.field = settings.defaultField();
} else if (settings.fieldsAndWeights().size() == 1) { } else if (settings.fieldsAndWeights().size() == 1) {
this.field = settings.fieldsAndWeights().keySet().iterator().next(); this.field = settings.fieldsAndWeights().keySet().iterator().next();
@ -148,6 +149,11 @@ public class MapperQueryParser extends QueryParser {
if (fields != null) { if (fields != null) {
if (fields.size() == 1) { if (fields.size() == 1) {
return getFieldQuerySingle(fields.iterator().next(), queryText, quoted); return getFieldQuerySingle(fields.iterator().next(), queryText, quoted);
} else if (fields.isEmpty()) {
// the requested fields do not match any field in the mapping
// happens for wildcard fields only since we cannot expand to a valid field name
// if there is no match in the mappings.
return new MatchNoDocsQuery("empty fields");
} }
if (settings.useDisMax()) { if (settings.useDisMax()) {
List<Query> queries = new ArrayList<>(); List<Query> queries = new ArrayList<>();
@ -721,7 +727,7 @@ public class MapperQueryParser extends QueryParser {
} }
private Query applyBoost(String field, Query q) { private Query applyBoost(String field, Query q) {
Float fieldBoost = settings.fieldsAndWeights().get(field); Float fieldBoost = settings.fieldsAndWeights() == null ? null : settings.fieldsAndWeights().get(field);
if (fieldBoost != null && fieldBoost != 1f) { if (fieldBoost != null && fieldBoost != 1f) {
return new BoostQuery(q, fieldBoost); return new BoostQuery(q, fieldBoost);
} }
@ -780,7 +786,8 @@ public class MapperQueryParser extends QueryParser {
if (field != null) { if (field != null) {
fields = context.simpleMatchToIndexNames(field); fields = context.simpleMatchToIndexNames(field);
} else { } else {
fields = settings.fieldsAndWeights().keySet(); Map<String, Float> fieldsAndWeights = settings.fieldsAndWeights();
fields = fieldsAndWeights == null ? Collections.emptyList() : fieldsAndWeights.keySet();
} }
return fields; return fields;
} }

View File

@ -22,6 +22,7 @@ package org.elasticsearch.common.xcontent;
import org.elasticsearch.common.CheckedFunction; import org.elasticsearch.common.CheckedFunction;
import org.elasticsearch.common.ParseField; import org.elasticsearch.common.ParseField;
import org.elasticsearch.common.bytes.BytesReference; import org.elasticsearch.common.bytes.BytesReference;
import org.elasticsearch.common.xcontent.ObjectParser.NamedObjectParser;
import org.elasticsearch.common.xcontent.ObjectParser.ValueType; import org.elasticsearch.common.xcontent.ObjectParser.ValueType;
import org.elasticsearch.common.xcontent.json.JsonXContent; import org.elasticsearch.common.xcontent.json.JsonXContent;
@ -30,6 +31,7 @@ import java.util.ArrayList;
import java.util.List; import java.util.List;
import java.util.function.BiConsumer; import java.util.function.BiConsumer;
import java.util.function.BiFunction; import java.util.function.BiFunction;
import java.util.function.Consumer;
/** /**
* Superclass for {@link ObjectParser} and {@link ConstructingObjectParser}. Defines most of the "declare" methods so they can be shared. * Superclass for {@link ObjectParser} and {@link ConstructingObjectParser}. Defines most of the "declare" methods so they can be shared.
@ -44,6 +46,94 @@ public abstract class AbstractObjectParser<Value, Context>
public abstract <T> void declareField(BiConsumer<Value, T> consumer, ContextParser<Context, T> parser, ParseField parseField, public abstract <T> void declareField(BiConsumer<Value, T> consumer, ContextParser<Context, T> parser, ParseField parseField,
ValueType type); ValueType type);
/**
* Declares named objects in the style of aggregations. These are named
* inside and object like this:
*
* <pre>
* <code>
* {
* "aggregations": {
* "name_1": { "aggregation_type": {} },
* "name_2": { "aggregation_type": {} },
* "name_3": { "aggregation_type": {} }
* }
* }
* }
* </code>
* </pre>
*
* Unlike the other version of this method, "ordered" mode (arrays of
* objects) is not supported.
*
* See NamedObjectHolder in ObjectParserTests for examples of how to invoke
* this.
*
* @param consumer
* sets the values once they have been parsed
* @param namedObjectParser
* parses each named object
* @param parseField
* the field to parse
*/
public abstract <T> void declareNamedObjects(BiConsumer<Value, List<T>> consumer, NamedObjectParser<T, Context> namedObjectParser,
ParseField parseField);
/**
* Declares named objects in the style of highlighting's field element.
* These are usually named inside and object like this:
*
* <pre>
* <code>
* {
* "highlight": {
* "fields": { &lt;------ this one
* "title": {},
* "body": {},
* "category": {}
* }
* }
* }
* </code>
* </pre>
*
* but, when order is important, some may be written this way:
*
* <pre>
* <code>
* {
* "highlight": {
* "fields": [ &lt;------ this one
* {"title": {}},
* {"body": {}},
* {"category": {}}
* ]
* }
* }
* </code>
* </pre>
*
* This is because json doesn't enforce ordering. Elasticsearch reads it in
* the order sent but tools that generate json are free to put object
* members in an unordered Map, jumbling them. Thus, if you care about order
* you can send the object in the second way.
*
* See NamedObjectHolder in ObjectParserTests for examples of how to invoke
* this.
*
* @param consumer
* sets the values once they have been parsed
* @param namedObjectParser
* parses each named object
* @param orderedModeCallback
* called when the named object is parsed using the "ordered"
* mode (the array of objects)
* @param parseField
* the field to parse
*/
public abstract <T> void declareNamedObjects(BiConsumer<Value, List<T>> consumer, NamedObjectParser<T, Context> namedObjectParser,
Consumer<Value> orderedModeCallback, ParseField parseField);
public <T> void declareField(BiConsumer<Value, T> consumer, CheckedFunction<XContentParser, T, IOException> parser, public <T> void declareField(BiConsumer<Value, T> consumer, CheckedFunction<XContentParser, T, IOException> parser,
ParseField parseField, ValueType type) { ParseField parseField, ValueType type) {
if (parser == null) { if (parser == null) {

View File

@ -21,6 +21,7 @@ package org.elasticsearch.common.xcontent;
import org.elasticsearch.common.ParseField; import org.elasticsearch.common.ParseField;
import org.elasticsearch.common.ParsingException; import org.elasticsearch.common.ParsingException;
import org.elasticsearch.common.xcontent.ObjectParser.NamedObjectParser;
import org.elasticsearch.common.xcontent.ObjectParser.ValueType; import org.elasticsearch.common.xcontent.ObjectParser.ValueType;
import java.io.IOException; import java.io.IOException;
@ -77,14 +78,14 @@ public final class ConstructingObjectParser<Value, Context> extends AbstractObje
/** /**
* Consumer that marks a field as a required constructor argument instead of a real object field. * Consumer that marks a field as a required constructor argument instead of a real object field.
*/ */
private static final BiConsumer<Object, Object> REQUIRED_CONSTRUCTOR_ARG_MARKER = (a, b) -> { private static final BiConsumer<?, ?> REQUIRED_CONSTRUCTOR_ARG_MARKER = (a, b) -> {
throw new UnsupportedOperationException("I am just a marker I should never be called."); throw new UnsupportedOperationException("I am just a marker I should never be called.");
}; };
/** /**
* Consumer that marks a field as an optional constructor argument instead of a real object field. * Consumer that marks a field as an optional constructor argument instead of a real object field.
*/ */
private static final BiConsumer<Object, Object> OPTIONAL_CONSTRUCTOR_ARG_MARKER = (a, b) -> { private static final BiConsumer<?, ?> OPTIONAL_CONSTRUCTOR_ARG_MARKER = (a, b) -> {
throw new UnsupportedOperationException("I am just a marker I should never be called."); throw new UnsupportedOperationException("I am just a marker I should never be called.");
}; };
@ -189,7 +190,7 @@ public final class ConstructingObjectParser<Value, Context> extends AbstractObje
if (consumer == REQUIRED_CONSTRUCTOR_ARG_MARKER || consumer == OPTIONAL_CONSTRUCTOR_ARG_MARKER) { if (consumer == REQUIRED_CONSTRUCTOR_ARG_MARKER || consumer == OPTIONAL_CONSTRUCTOR_ARG_MARKER) {
/* /*
* Constructor arguments are detected by this "marker" consumer. It keeps the API looking clean even if it is a bit sleezy. We * Constructor arguments are detected by these "marker" consumers. It keeps the API looking clean even if it is a bit sleezy. We
* then build a new consumer directly against the object parser that triggers the "constructor arg just arrived behavior" of the * then build a new consumer directly against the object parser that triggers the "constructor arg just arrived behavior" of the
* parser. Conveniently, we can close over the position of the constructor in the argument list so we don't need to do any fancy * parser. Conveniently, we can close over the position of the constructor in the argument list so we don't need to do any fancy
* or expensive lookups whenever the constructor args come in. * or expensive lookups whenever the constructor args come in.
@ -204,6 +205,91 @@ public final class ConstructingObjectParser<Value, Context> extends AbstractObje
} }
} }
@Override
public <T> void declareNamedObjects(BiConsumer<Value, List<T>> consumer, NamedObjectParser<T, Context> namedObjectParser,
ParseField parseField) {
if (consumer == null) {
throw new IllegalArgumentException("[consumer] is required");
}
if (namedObjectParser == null) {
throw new IllegalArgumentException("[parser] is required");
}
if (parseField == null) {
throw new IllegalArgumentException("[parseField] is required");
}
if (consumer == REQUIRED_CONSTRUCTOR_ARG_MARKER || consumer == OPTIONAL_CONSTRUCTOR_ARG_MARKER) {
/*
* Constructor arguments are detected by this "marker" consumer. It
* keeps the API looking clean even if it is a bit sleezy. We then
* build a new consumer directly against the object parser that
* triggers the "constructor arg just arrived behavior" of the
* parser. Conveniently, we can close over the position of the
* constructor in the argument list so we don't need to do any fancy
* or expensive lookups whenever the constructor args come in.
*/
int position = constructorArgInfos.size();
boolean required = consumer == REQUIRED_CONSTRUCTOR_ARG_MARKER;
constructorArgInfos.add(new ConstructorArgInfo(parseField, required));
objectParser.declareNamedObjects((target, v) -> target.constructorArg(position, parseField, v), namedObjectParser, parseField);
} else {
numberOfFields += 1;
objectParser.declareNamedObjects(queueingConsumer(consumer, parseField), namedObjectParser, parseField);
}
}
@Override
public <T> void declareNamedObjects(BiConsumer<Value, List<T>> consumer, NamedObjectParser<T, Context> namedObjectParser,
Consumer<Value> orderedModeCallback, ParseField parseField) {
if (consumer == null) {
throw new IllegalArgumentException("[consumer] is required");
}
if (namedObjectParser == null) {
throw new IllegalArgumentException("[parser] is required");
}
if (orderedModeCallback == null) {
throw new IllegalArgumentException("[orderedModeCallback] is required");
}
if (parseField == null) {
throw new IllegalArgumentException("[parseField] is required");
}
if (consumer == REQUIRED_CONSTRUCTOR_ARG_MARKER || consumer == OPTIONAL_CONSTRUCTOR_ARG_MARKER) {
/*
* Constructor arguments are detected by this "marker" consumer. It
* keeps the API looking clean even if it is a bit sleezy. We then
* build a new consumer directly against the object parser that
* triggers the "constructor arg just arrived behavior" of the
* parser. Conveniently, we can close over the position of the
* constructor in the argument list so we don't need to do any fancy
* or expensive lookups whenever the constructor args come in.
*/
int position = constructorArgInfos.size();
boolean required = consumer == REQUIRED_CONSTRUCTOR_ARG_MARKER;
constructorArgInfos.add(new ConstructorArgInfo(parseField, required));
objectParser.declareNamedObjects((target, v) -> target.constructorArg(position, parseField, v), namedObjectParser,
wrapOrderedModeCallBack(orderedModeCallback), parseField);
} else {
numberOfFields += 1;
objectParser.declareNamedObjects(queueingConsumer(consumer, parseField), namedObjectParser,
wrapOrderedModeCallBack(orderedModeCallback), parseField);
}
}
private Consumer<Target> wrapOrderedModeCallBack(Consumer<Value> callback) {
return (target) -> {
if (target.targetObject != null) {
// The target has already been built. Call the callback now.
callback.accept(target.targetObject);
return;
}
/*
* The target hasn't been built. Queue the callback.
*/
target.queuedOrderedModeCallback = callback;
};
}
/** /**
* Creates the consumer that does the "field just arrived" behavior. If the targetObject hasn't been built then it queues the value. * Creates the consumer that does the "field just arrived" behavior. If the targetObject hasn't been built then it queues the value.
* Otherwise it just applies the value just like {@linkplain ObjectParser} does. * Otherwise it just applies the value just like {@linkplain ObjectParser} does.
@ -258,6 +344,11 @@ public final class ConstructingObjectParser<Value, Context> extends AbstractObje
* Fields to be saved to the target object when we can build it. This is only allocated if a field has to be queued. * Fields to be saved to the target object when we can build it. This is only allocated if a field has to be queued.
*/ */
private Consumer<Value>[] queuedFields; private Consumer<Value>[] queuedFields;
/**
* OrderedModeCallback to be called with the target object when we can
* build it. This is only allocated if the callback has to be queued.
*/
private Consumer<Value> queuedOrderedModeCallback;
/** /**
* The count of fields already queued. * The count of fields already queued.
*/ */
@ -343,6 +434,9 @@ public final class ConstructingObjectParser<Value, Context> extends AbstractObje
private void buildTarget() { private void buildTarget() {
try { try {
targetObject = builder.apply(constructorArgs); targetObject = builder.apply(constructorArgs);
if (queuedOrderedModeCallback != null) {
queuedOrderedModeCallback.accept(targetObject);
}
while (queuedFieldsCount > 0) { while (queuedFieldsCount > 0) {
queuedFieldsCount -= 1; queuedFieldsCount -= 1;
queuedFields[queuedFieldsCount].accept(targetObject); queuedFields[queuedFieldsCount].accept(targetObject);

View File

@ -227,41 +227,7 @@ public final class ObjectParser<Value, Context> extends AbstractObjectParser<Val
}, field, ValueType.OBJECT_OR_BOOLEAN); }, field, ValueType.OBJECT_OR_BOOLEAN);
} }
/** @Override
* Declares named objects in the style of highlighting's field element. These are usually named inside and object like this:
* <pre><code>
* {
* "highlight": {
* "fields": { &lt;------ this one
* "title": {},
* "body": {},
* "category": {}
* }
* }
* }
* </code></pre>
* but, when order is important, some may be written this way:
* <pre><code>
* {
* "highlight": {
* "fields": [ &lt;------ this one
* {"title": {}},
* {"body": {}},
* {"category": {}}
* ]
* }
* }
* </code></pre>
* This is because json doesn't enforce ordering. Elasticsearch reads it in the order sent but tools that generate json are free to put
* object members in an unordered Map, jumbling them. Thus, if you care about order you can send the object in the second way.
*
* See NamedObjectHolder in ObjectParserTests for examples of how to invoke this.
*
* @param consumer sets the values once they have been parsed
* @param namedObjectParser parses each named object
* @param orderedModeCallback called when the named object is parsed using the "ordered" mode (the array of objects)
* @param field the field to parse
*/
public <T> void declareNamedObjects(BiConsumer<Value, List<T>> consumer, NamedObjectParser<T, Context> namedObjectParser, public <T> void declareNamedObjects(BiConsumer<Value, List<T>> consumer, NamedObjectParser<T, Context> namedObjectParser,
Consumer<Value> orderedModeCallback, ParseField field) { Consumer<Value> orderedModeCallback, ParseField field) {
// This creates and parses the named object // This creates and parses the named object
@ -311,26 +277,7 @@ public final class ObjectParser<Value, Context> extends AbstractObjectParser<Val
}, field, ValueType.OBJECT_ARRAY); }, field, ValueType.OBJECT_ARRAY);
} }
/** @Override
* Declares named objects in the style of aggregations. These are named inside and object like this:
* <pre><code>
* {
* "aggregations": {
* "name_1": { "aggregation_type": {} },
* "name_2": { "aggregation_type": {} },
* "name_3": { "aggregation_type": {} }
* }
* }
* }
* </code></pre>
* Unlike the other version of this method, "ordered" mode (arrays of objects) is not supported.
*
* See NamedObjectHolder in ObjectParserTests for examples of how to invoke this.
*
* @param consumer sets the values once they have been parsed
* @param namedObjectParser parses each named object
* @param field the field to parse
*/
public <T> void declareNamedObjects(BiConsumer<Value, List<T>> consumer, NamedObjectParser<T, Context> namedObjectParser, public <T> void declareNamedObjects(BiConsumer<Value, List<T>> consumer, NamedObjectParser<T, Context> namedObjectParser,
ParseField field) { ParseField field) {
Consumer<Value> orderedModeCallback = (v) -> { Consumer<Value> orderedModeCallback = (v) -> {

View File

@ -236,6 +236,15 @@ public final class KeywordFieldMapper extends FieldMapper {
@Override @Override
protected BytesRef indexedValueForSearch(Object value) { protected BytesRef indexedValueForSearch(Object value) {
if (searchAnalyzer() == Lucene.KEYWORD_ANALYZER) {
// keyword analyzer with the default attribute source which encodes terms using UTF8
// in that case we skip normalization, which may be slow if there many terms need to
// parse (eg. large terms query) since Analyzer.normalize involves things like creating
// attributes through reflection
// This if statement will be used whenever a normalizer is NOT configured
return super.indexedValueForSearch(value);
}
if (value == null) { if (value == null) {
return null; return null;
} }

View File

@ -22,10 +22,8 @@ package org.elasticsearch.index.mapper;
import org.apache.lucene.analysis.Analyzer; import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.TokenStream; import org.apache.lucene.analysis.TokenStream;
import org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute; import org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute;
import org.apache.lucene.document.Field;
import org.apache.lucene.index.IndexOptions; import org.apache.lucene.index.IndexOptions;
import org.apache.lucene.index.IndexableField; import org.apache.lucene.index.IndexableField;
import org.elasticsearch.Version;
import org.elasticsearch.common.settings.Settings; import org.elasticsearch.common.settings.Settings;
import org.elasticsearch.common.xcontent.XContentBuilder; import org.elasticsearch.common.xcontent.XContentBuilder;
import org.elasticsearch.index.analysis.NamedAnalyzer; import org.elasticsearch.index.analysis.NamedAnalyzer;
@ -36,6 +34,7 @@ import java.util.List;
import java.util.Map; import java.util.Map;
import static org.elasticsearch.common.xcontent.support.XContentMapValues.nodeIntegerValue; import static org.elasticsearch.common.xcontent.support.XContentMapValues.nodeIntegerValue;
import static org.elasticsearch.common.xcontent.support.XContentMapValues.nodeBooleanValue;
import static org.elasticsearch.index.mapper.TypeParsers.parseField; import static org.elasticsearch.index.mapper.TypeParsers.parseField;
/** /**
@ -47,10 +46,12 @@ public class TokenCountFieldMapper extends FieldMapper {
public static class Defaults { public static class Defaults {
public static final MappedFieldType FIELD_TYPE = new NumberFieldMapper.NumberFieldType(NumberFieldMapper.NumberType.INTEGER); public static final MappedFieldType FIELD_TYPE = new NumberFieldMapper.NumberFieldType(NumberFieldMapper.NumberType.INTEGER);
public static final boolean DEFAULT_POSITION_INCREMENTS = true;
} }
public static class Builder extends FieldMapper.Builder<Builder, TokenCountFieldMapper> { public static class Builder extends FieldMapper.Builder<Builder, TokenCountFieldMapper> {
private NamedAnalyzer analyzer; private NamedAnalyzer analyzer;
private boolean enablePositionIncrements = Defaults.DEFAULT_POSITION_INCREMENTS;
public Builder(String name) { public Builder(String name) {
super(name, Defaults.FIELD_TYPE, Defaults.FIELD_TYPE); super(name, Defaults.FIELD_TYPE, Defaults.FIELD_TYPE);
@ -66,18 +67,26 @@ public class TokenCountFieldMapper extends FieldMapper {
return analyzer; return analyzer;
} }
public Builder enablePositionIncrements(boolean enablePositionIncrements) {
this.enablePositionIncrements = enablePositionIncrements;
return this;
}
public boolean enablePositionIncrements() {
return enablePositionIncrements;
}
@Override @Override
public TokenCountFieldMapper build(BuilderContext context) { public TokenCountFieldMapper build(BuilderContext context) {
setupFieldType(context); setupFieldType(context);
return new TokenCountFieldMapper(name, fieldType, defaultFieldType, return new TokenCountFieldMapper(name, fieldType, defaultFieldType,
context.indexSettings(), analyzer, multiFieldsBuilder.build(this, context), copyTo); context.indexSettings(), analyzer, enablePositionIncrements, multiFieldsBuilder.build(this, context), copyTo);
} }
} }
public static class TypeParser implements Mapper.TypeParser { public static class TypeParser implements Mapper.TypeParser {
@Override @Override
@SuppressWarnings("unchecked") public Mapper.Builder<?,?> parse(String name, Map<String, Object> node, ParserContext parserContext) throws MapperParsingException {
public Mapper.Builder parse(String name, Map<String, Object> node, ParserContext parserContext) throws MapperParsingException {
TokenCountFieldMapper.Builder builder = new TokenCountFieldMapper.Builder(name); TokenCountFieldMapper.Builder builder = new TokenCountFieldMapper.Builder(name);
for (Iterator<Map.Entry<String, Object>> iterator = node.entrySet().iterator(); iterator.hasNext();) { for (Iterator<Map.Entry<String, Object>> iterator = node.entrySet().iterator(); iterator.hasNext();) {
Map.Entry<String, Object> entry = iterator.next(); Map.Entry<String, Object> entry = iterator.next();
@ -93,6 +102,9 @@ public class TokenCountFieldMapper extends FieldMapper {
} }
builder.analyzer(analyzer); builder.analyzer(analyzer);
iterator.remove(); iterator.remove();
} else if (propName.equals("enable_position_increments")) {
builder.enablePositionIncrements(nodeBooleanValue(propNode));
iterator.remove();
} }
} }
parseField(builder, name, node, parserContext); parseField(builder, name, node, parserContext);
@ -104,11 +116,13 @@ public class TokenCountFieldMapper extends FieldMapper {
} }
private NamedAnalyzer analyzer; private NamedAnalyzer analyzer;
private boolean enablePositionIncrements;
protected TokenCountFieldMapper(String simpleName, MappedFieldType fieldType, MappedFieldType defaultFieldType, protected TokenCountFieldMapper(String simpleName, MappedFieldType fieldType, MappedFieldType defaultFieldType,
Settings indexSettings, NamedAnalyzer analyzer, MultiFields multiFields, CopyTo copyTo) { Settings indexSettings, NamedAnalyzer analyzer, boolean enablePositionIncrements, MultiFields multiFields, CopyTo copyTo) {
super(simpleName, fieldType, defaultFieldType, indexSettings, multiFields, copyTo); super(simpleName, fieldType, defaultFieldType, indexSettings, multiFields, copyTo);
this.analyzer = analyzer; this.analyzer = analyzer;
this.enablePositionIncrements = enablePositionIncrements;
} }
@Override @Override
@ -124,7 +138,7 @@ public class TokenCountFieldMapper extends FieldMapper {
if (value == null) { if (value == null) {
tokenCount = (Integer) fieldType().nullValue(); tokenCount = (Integer) fieldType().nullValue();
} else { } else {
tokenCount = countPositions(analyzer, name(), value); tokenCount = countPositions(analyzer, name(), value, enablePositionIncrements);
} }
boolean indexed = fieldType().indexOptions() != IndexOptions.NONE; boolean indexed = fieldType().indexOptions() != IndexOptions.NONE;
@ -138,19 +152,26 @@ public class TokenCountFieldMapper extends FieldMapper {
* @param analyzer analyzer to create token stream * @param analyzer analyzer to create token stream
* @param fieldName field name to pass to analyzer * @param fieldName field name to pass to analyzer
* @param fieldValue field value to pass to analyzer * @param fieldValue field value to pass to analyzer
* @param enablePositionIncrements should we count position increments ?
* @return number of position increments in a token stream * @return number of position increments in a token stream
* @throws IOException if tokenStream throws it * @throws IOException if tokenStream throws it
*/ */
static int countPositions(Analyzer analyzer, String fieldName, String fieldValue) throws IOException { static int countPositions(Analyzer analyzer, String fieldName, String fieldValue, boolean enablePositionIncrements) throws IOException {
try (TokenStream tokenStream = analyzer.tokenStream(fieldName, fieldValue)) { try (TokenStream tokenStream = analyzer.tokenStream(fieldName, fieldValue)) {
int count = 0; int count = 0;
PositionIncrementAttribute position = tokenStream.addAttribute(PositionIncrementAttribute.class); PositionIncrementAttribute position = tokenStream.addAttribute(PositionIncrementAttribute.class);
tokenStream.reset(); tokenStream.reset();
while (tokenStream.incrementToken()) { while (tokenStream.incrementToken()) {
if (enablePositionIncrements) {
count += position.getPositionIncrement(); count += position.getPositionIncrement();
} else {
count += Math.min(1, position.getPositionIncrement());
}
} }
tokenStream.end(); tokenStream.end();
if (enablePositionIncrements) {
count += position.getPositionIncrement(); count += position.getPositionIncrement();
}
return count; return count;
} }
} }
@ -163,6 +184,14 @@ public class TokenCountFieldMapper extends FieldMapper {
return analyzer.name(); return analyzer.name();
} }
/**
* Indicates if position increments are counted.
* @return <code>true</code> if position increments are counted
*/
public boolean enablePositionIncrements() {
return enablePositionIncrements;
}
@Override @Override
protected String contentType() { protected String contentType() {
return CONTENT_TYPE; return CONTENT_TYPE;
@ -172,12 +201,16 @@ public class TokenCountFieldMapper extends FieldMapper {
protected void doMerge(Mapper mergeWith, boolean updateAllTypes) { protected void doMerge(Mapper mergeWith, boolean updateAllTypes) {
super.doMerge(mergeWith, updateAllTypes); super.doMerge(mergeWith, updateAllTypes);
this.analyzer = ((TokenCountFieldMapper) mergeWith).analyzer; this.analyzer = ((TokenCountFieldMapper) mergeWith).analyzer;
this.enablePositionIncrements = ((TokenCountFieldMapper) mergeWith).enablePositionIncrements;
} }
@Override @Override
protected void doXContentBody(XContentBuilder builder, boolean includeDefaults, Params params) throws IOException { protected void doXContentBody(XContentBuilder builder, boolean includeDefaults, Params params) throws IOException {
super.doXContentBody(builder, includeDefaults, params); super.doXContentBody(builder, includeDefaults, params);
builder.field("analyzer", analyzer()); builder.field("analyzer", analyzer());
if (includeDefaults || enablePositionIncrements() != Defaults.DEFAULT_POSITION_INCREMENTS) {
builder.field("enable_position_increments", enablePositionIncrements());
}
} }
} }

View File

@ -981,7 +981,11 @@ public class QueryStringQueryBuilder extends AbstractQueryBuilder<QueryStringQue
} }
qpSettings.lenient(lenient == null ? context.queryStringLenient() : lenient); qpSettings.lenient(lenient == null ? context.queryStringLenient() : lenient);
} }
if (fieldsAndWeights.isEmpty() == false || resolvedFields.isEmpty() == false) {
// We set the fields and weight only if we have explicit fields to query
// Otherwise we set it to null and fallback to the default field.
qpSettings.fieldsAndWeights(resolvedFields); qpSettings.fieldsAndWeights(resolvedFields);
}
qpSettings.defaultOperator(defaultOperator.toQueryParserOperator()); qpSettings.defaultOperator(defaultOperator.toQueryParserOperator());
if (analyzer == null) { if (analyzer == null) {

View File

@ -28,6 +28,7 @@ import org.apache.lucene.search.Query;
import org.apache.lucene.search.QueryCache; import org.apache.lucene.search.QueryCache;
import org.apache.lucene.search.QueryCachingPolicy; import org.apache.lucene.search.QueryCachingPolicy;
import org.apache.lucene.search.Scorer; import org.apache.lucene.search.Scorer;
import org.apache.lucene.search.ScorerSupplier;
import org.apache.lucene.search.Weight; import org.apache.lucene.search.Weight;
import org.elasticsearch.common.component.AbstractComponent; import org.elasticsearch.common.component.AbstractComponent;
import org.elasticsearch.common.lucene.ShardCoreKeyMap; import org.elasticsearch.common.lucene.ShardCoreKeyMap;
@ -145,6 +146,12 @@ public class IndicesQueryCache extends AbstractComponent implements QueryCache,
return in.scorer(context); return in.scorer(context);
} }
@Override
public ScorerSupplier scorerSupplier(LeafReaderContext context) throws IOException {
shardKeyMap.add(context.reader());
return in.scorerSupplier(context);
}
@Override @Override
public BulkScorer bulkScorer(LeafReaderContext context) throws IOException { public BulkScorer bulkScorer(LeafReaderContext context) throws IOException {
shardKeyMap.add(context.reader()); shardKeyMap.add(context.reader());

View File

@ -24,6 +24,7 @@ import org.elasticsearch.common.Nullable;
import org.elasticsearch.common.ParseField; import org.elasticsearch.common.ParseField;
import org.elasticsearch.common.ParsingException; import org.elasticsearch.common.ParsingException;
import org.elasticsearch.common.bytes.BytesReference; import org.elasticsearch.common.bytes.BytesReference;
import org.elasticsearch.common.xcontent.ObjectParserTests.NamedObject;
import org.elasticsearch.common.xcontent.json.JsonXContent; import org.elasticsearch.common.xcontent.json.JsonXContent;
import org.elasticsearch.test.ESTestCase; import org.elasticsearch.test.ESTestCase;
import org.hamcrest.Matcher; import org.hamcrest.Matcher;
@ -37,6 +38,7 @@ import static org.elasticsearch.common.xcontent.ConstructingObjectParser.constru
import static org.elasticsearch.common.xcontent.ConstructingObjectParser.optionalConstructorArg; import static org.elasticsearch.common.xcontent.ConstructingObjectParser.optionalConstructorArg;
import static org.hamcrest.Matchers.anyOf; import static org.hamcrest.Matchers.anyOf;
import static org.hamcrest.Matchers.equalTo; import static org.hamcrest.Matchers.equalTo;
import static org.hamcrest.Matchers.hasSize;
import static org.hamcrest.Matchers.instanceOf; import static org.hamcrest.Matchers.instanceOf;
import static org.hamcrest.Matchers.nullValue; import static org.hamcrest.Matchers.nullValue;
@ -397,4 +399,188 @@ public class ConstructingObjectParserTests extends ESTestCase {
return parser; return parser;
} }
} }
public void testParseNamedObject() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": {\n"
+ " \"a\": {}"
+ "},\"named_in_constructor\": {\n"
+ " \"b\": {}"
+ "}}");
NamedObjectHolder h = NamedObjectHolder.PARSER.apply(parser, null);
assertThat(h.named, hasSize(1));
assertEquals("a", h.named.get(0).name);
assertThat(h.namedInConstructor, hasSize(1));
assertEquals("b", h.namedInConstructor.get(0).name);
assertFalse(h.namedSuppliedInOrder);
}
public void testParseNamedObjectInOrder() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": [\n"
+ " {\"a\": {}}"
+ "],\"named_in_constructor\": [\n"
+ " {\"b\": {}}"
+ "]}");
NamedObjectHolder h = NamedObjectHolder.PARSER.apply(parser, null);
assertThat(h.named, hasSize(1));
assertEquals("a", h.named.get(0).name);
assertThat(h.namedInConstructor, hasSize(1));
assertEquals("b", h.namedInConstructor.get(0).name);
assertTrue(h.namedSuppliedInOrder);
}
public void testParseNamedObjectTwoFieldsInArray() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": [\n"
+ " {\"a\": {}, \"b\": {}}"
+ "],\"named_in_constructor\": [\n"
+ " {\"c\": {}}"
+ "]}");
ParsingException e = expectThrows(ParsingException.class, () -> NamedObjectHolder.PARSER.apply(parser, null));
assertEquals("[named_object_holder] failed to parse field [named]", e.getMessage());
assertEquals(
"[named] can be a single object with any number of fields or an array where each entry is an object with a single field",
e.getCause().getMessage());
}
public void testParseNamedObjectTwoFieldsInArrayConstructorArg() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": [\n"
+ " {\"a\": {}}"
+ "],\"named_in_constructor\": [\n"
+ " {\"c\": {}, \"d\": {}}"
+ "]}");
ParsingException e = expectThrows(ParsingException.class, () -> NamedObjectHolder.PARSER.apply(parser, null));
assertEquals("[named_object_holder] failed to parse field [named_in_constructor]", e.getMessage());
assertEquals(
"[named_in_constructor] can be a single object with any number of fields or an array where each entry is an object with a "
+ "single field", e.getCause().getMessage());
}
public void testParseNamedObjectNoFieldsInArray() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": [\n"
+ " {}"
+ "],\"named_in_constructor\": [\n"
+ " {\"a\": {}}"
+ "]}");
ParsingException e = expectThrows(ParsingException.class, () -> NamedObjectHolder.PARSER.apply(parser, null));
assertEquals("[named_object_holder] failed to parse field [named]", e.getMessage());
assertEquals(
"[named] can be a single object with any number of fields or an array where each entry is an object with a single field",
e.getCause().getMessage());
}
public void testParseNamedObjectNoFieldsInArrayConstructorArg() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": [\n"
+ " {\"a\": {}}"
+ "],\"named_in_constructor\": [\n"
+ " {}"
+ "]}");
ParsingException e = expectThrows(ParsingException.class, () -> NamedObjectHolder.PARSER.apply(parser, null));
assertEquals("[named_object_holder] failed to parse field [named_in_constructor]", e.getMessage());
assertEquals(
"[named_in_constructor] can be a single object with any number of fields or an array where each entry is an object with a "
+ "single field", e.getCause().getMessage());
}
public void testParseNamedObjectJunkInArray() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": [\n"
+ " \"junk\""
+ "],\"named_in_constructor\": [\n"
+ " {\"a\": {}}"
+ "]}");
ParsingException e = expectThrows(ParsingException.class, () -> NamedObjectHolder.PARSER.apply(parser, null));
assertEquals("[named_object_holder] failed to parse field [named]", e.getMessage());
assertEquals(
"[named] can be a single object with any number of fields or an array where each entry is an object with a single field",
e.getCause().getMessage());
}
public void testParseNamedObjectJunkInArrayConstructorArg() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": [\n"
+ " {\"a\": {}}"
+ "],\"named_in_constructor\": [\n"
+ " \"junk\""
+ "]}");
ParsingException e = expectThrows(ParsingException.class, () -> NamedObjectHolder.PARSER.apply(parser, null));
assertEquals("[named_object_holder] failed to parse field [named_in_constructor]", e.getMessage());
assertEquals(
"[named_in_constructor] can be a single object with any number of fields or an array where each entry is an object with a "
+ "single field", e.getCause().getMessage());
}
public void testParseNamedObjectInOrderNotSupported() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": [\n"
+ " {\"a\": {}}"
+ "],\"named_in_constructor\": {\"b\": {}}"
+ "}");
// Create our own parser for this test so we can disable support for the "ordered" mode specified by the array above
@SuppressWarnings("unchecked")
ConstructingObjectParser<NamedObjectHolder, Void> objectParser = new ConstructingObjectParser<>("named_object_holder",
a -> new NamedObjectHolder(((List<NamedObject>) a[0])));
objectParser.declareNamedObjects(ConstructingObjectParser.constructorArg(), NamedObject.PARSER,
new ParseField("named_in_constructor"));
objectParser.declareNamedObjects(NamedObjectHolder::setNamed, NamedObject.PARSER, new ParseField("named"));
// Now firing the xml through it fails
ParsingException e = expectThrows(ParsingException.class, () -> objectParser.apply(parser, null));
assertEquals("[named_object_holder] failed to parse field [named]", e.getMessage());
assertEquals("[named] doesn't support arrays. Use a single object with multiple fields.", e.getCause().getMessage());
}
public void testParseNamedObjectInOrderNotSupportedConstructorArg() throws IOException {
XContentParser parser = createParser(JsonXContent.jsonXContent,
"{\"named\": {\"a\": {}}"
+ ",\"named_in_constructor\": [\n"
+ " {\"b\": {}}"
+ "]}");
// Create our own parser for this test so we can disable support for the "ordered" mode specified by the array above
@SuppressWarnings("unchecked")
ConstructingObjectParser<NamedObjectHolder, Void> objectParser = new ConstructingObjectParser<>("named_object_holder",
a -> new NamedObjectHolder(((List<NamedObject>) a[0])));
objectParser.declareNamedObjects(ConstructingObjectParser.constructorArg(), NamedObject.PARSER,
new ParseField("named_in_constructor"));
objectParser.declareNamedObjects(NamedObjectHolder::setNamed, NamedObject.PARSER, new ParseField("named"));
// Now firing the xml through it fails
ParsingException e = expectThrows(ParsingException.class, () -> objectParser.apply(parser, null));
assertEquals("[named_object_holder] failed to parse field [named_in_constructor]", e.getMessage());
assertEquals("[named_in_constructor] doesn't support arrays. Use a single object with multiple fields.", e.getCause().getMessage());
}
static class NamedObjectHolder {
@SuppressWarnings("unchecked")
public static final ConstructingObjectParser<NamedObjectHolder, Void> PARSER = new ConstructingObjectParser<>("named_object_holder",
a -> new NamedObjectHolder(((List<NamedObject>) a[0])));
static {
PARSER.declareNamedObjects(ConstructingObjectParser.constructorArg(), NamedObject.PARSER, NamedObjectHolder::keepNamedInOrder,
new ParseField("named_in_constructor"));
PARSER.declareNamedObjects(NamedObjectHolder::setNamed, NamedObject.PARSER, NamedObjectHolder::keepNamedInOrder,
new ParseField("named"));
}
private List<NamedObject> named;
private List<NamedObject> namedInConstructor;
private boolean namedSuppliedInOrder = false;
NamedObjectHolder(List<NamedObject> namedInConstructor) {
this.namedInConstructor = namedInConstructor;
}
public void setNamed(List<NamedObject> named) {
this.named = named;
}
public void keepNamedInOrder() {
namedSuppliedInOrder = true;
}
}
} }

View File

@ -150,4 +150,13 @@ public class KeywordFieldTypeTests extends FieldTypeTestCase {
() -> ft.fuzzyQuery("foo", Fuzziness.fromEdits(2), 1, 50, true)); () -> ft.fuzzyQuery("foo", Fuzziness.fromEdits(2), 1, 50, true));
assertEquals("Cannot search on field [field] since it is not indexed.", e.getMessage()); assertEquals("Cannot search on field [field] since it is not indexed.", e.getMessage());
} }
public void testNormalizeQueries() {
MappedFieldType ft = createDefaultFieldType();
ft.setName("field");
ft.setSearchAnalyzer(Lucene.KEYWORD_ANALYZER);
assertEquals(new TermQuery(new Term("field", new BytesRef("FOO"))), ft.termQuery("FOO", null));
ft.setSearchAnalyzer(Lucene.STANDARD_ANALYZER);
assertEquals(new TermQuery(new Term("field", new BytesRef("foo"))), ft.termQuery("FOO", null));
}
} }

View File

@ -131,6 +131,12 @@ public class TokenCountFieldMapperIntegrationIT extends ESIntegTestCase {
.field("analyzer", "standard") .field("analyzer", "standard")
.field("doc_values", true) .field("doc_values", true)
.endObject() .endObject()
.startObject("token_count_without_position_increments")
.field("type", "token_count")
.field("analyzer", "english")
.field("enable_position_increments", false)
.field("store", true)
.endObject()
.endObject() .endObject()
.endObject() .endObject()
.endObject() .endObject()
@ -169,6 +175,7 @@ public class TokenCountFieldMapperIntegrationIT extends ESIntegTestCase {
private SearchRequestBuilder prepareSearch() { private SearchRequestBuilder prepareSearch() {
SearchRequestBuilder request = client().prepareSearch("test").setTypes("test"); SearchRequestBuilder request = client().prepareSearch("test").setTypes("test");
request.addStoredField("foo.token_count"); request.addStoredField("foo.token_count");
request.addStoredField("foo.token_count_without_position_increments");
if (loadCountedFields) { if (loadCountedFields) {
request.addStoredField("foo"); request.addStoredField("foo");
} }
@ -186,32 +193,38 @@ public class TokenCountFieldMapperIntegrationIT extends ESIntegTestCase {
for (SearchHit hit : result.getHits()) { for (SearchHit hit : result.getHits()) {
String id = hit.getId(); String id = hit.getId();
if (id.equals("single")) { if (id.equals("single")) {
assertSearchHit(hit, 4); assertSearchHit(hit, new int[]{4}, new int[]{4});
} else if (id.equals("bulk1")) { } else if (id.equals("bulk1")) {
assertSearchHit(hit, 3); assertSearchHit(hit, new int[]{3}, new int[]{3});
} else if (id.equals("bulk2")) { } else if (id.equals("bulk2")) {
assertSearchHit(hit, 5); assertSearchHit(hit, new int[]{5}, new int[]{4});
} else if (id.equals("multi")) { } else if (id.equals("multi")) {
assertSearchHit(hit, 2, 7); assertSearchHit(hit, new int[]{2, 7}, new int[]{2, 7});
} else if (id.equals("multibulk1")) { } else if (id.equals("multibulk1")) {
assertSearchHit(hit, 1, 8); assertSearchHit(hit, new int[]{1, 8}, new int[]{1, 8});
} else if (id.equals("multibulk2")) { } else if (id.equals("multibulk2")) {
assertSearchHit(hit, 6, 10); assertSearchHit(hit, new int[]{6, 10}, new int[]{3, 9});
} else { } else {
throw new ElasticsearchException("Unexpected response!"); throw new ElasticsearchException("Unexpected response!");
} }
} }
} }
private void assertSearchHit(SearchHit hit, int... termCounts) { private void assertSearchHit(SearchHit hit, int[] standardTermCounts, int[] englishTermCounts) {
assertThat(hit.field("foo.token_count"), not(nullValue())); assertThat(hit.field("foo.token_count"), not(nullValue()));
assertThat(hit.field("foo.token_count").getValues().size(), equalTo(termCounts.length)); assertThat(hit.field("foo.token_count").getValues().size(), equalTo(standardTermCounts.length));
for (int i = 0; i < termCounts.length; i++) { for (int i = 0; i < standardTermCounts.length; i++) {
assertThat((Integer) hit.field("foo.token_count").getValues().get(i), equalTo(termCounts[i])); assertThat((Integer) hit.field("foo.token_count").getValues().get(i), equalTo(standardTermCounts[i]));
}
assertThat(hit.field("foo.token_count_without_position_increments"), not(nullValue()));
assertThat(hit.field("foo.token_count_without_position_increments").getValues().size(), equalTo(englishTermCounts.length));
for (int i = 0; i < englishTermCounts.length; i++) {
assertThat((Integer) hit.field("foo.token_count_without_position_increments").getValues().get(i), equalTo(englishTermCounts[i]));
} }
if (loadCountedFields && storeCountedFields) { if (loadCountedFields && storeCountedFields) {
assertThat(hit.field("foo").getValues().size(), equalTo(termCounts.length)); assertThat(hit.field("foo").getValues().size(), equalTo(standardTermCounts.length));
} }
} }
} }

View File

@ -24,23 +24,18 @@ import org.apache.lucene.analysis.CannedTokenStream;
import org.apache.lucene.analysis.MockTokenizer; import org.apache.lucene.analysis.MockTokenizer;
import org.apache.lucene.analysis.Token; import org.apache.lucene.analysis.Token;
import org.apache.lucene.analysis.TokenStream; import org.apache.lucene.analysis.TokenStream;
import org.elasticsearch.Version;
import org.elasticsearch.cluster.metadata.IndexMetaData;
import org.elasticsearch.common.compress.CompressedXContent; import org.elasticsearch.common.compress.CompressedXContent;
import org.elasticsearch.common.settings.Settings;
import org.elasticsearch.common.xcontent.XContentFactory; import org.elasticsearch.common.xcontent.XContentFactory;
import org.elasticsearch.index.IndexService; import org.elasticsearch.index.IndexService;
import org.elasticsearch.plugins.Plugin; import org.elasticsearch.plugins.Plugin;
import org.elasticsearch.test.ESSingleNodeTestCase; import org.elasticsearch.test.ESSingleNodeTestCase;
import org.elasticsearch.test.InternalSettingsPlugin; import org.elasticsearch.test.InternalSettingsPlugin;
import org.elasticsearch.test.VersionUtils;
import java.io.IOException; import java.io.IOException;
import java.util.Arrays; import java.util.Arrays;
import java.util.Collection; import java.util.Collection;
import java.util.Collections; import java.util.Collections;
import static com.carrotsearch.randomizedtesting.RandomizedTest.getRandom;
import static org.hamcrest.Matchers.containsString; import static org.hamcrest.Matchers.containsString;
import static org.hamcrest.Matchers.equalTo; import static org.hamcrest.Matchers.equalTo;
@ -80,15 +75,38 @@ public class TokenCountFieldMapperTests extends ESSingleNodeTestCase {
assertThat(((TokenCountFieldMapper) stage2.mappers().smartNameFieldMapper("tc")).analyzer(), equalTo("standard")); assertThat(((TokenCountFieldMapper) stage2.mappers().smartNameFieldMapper("tc")).analyzer(), equalTo("standard"));
} }
public void testCountPositions() throws IOException { /**
// We're looking to make sure that we: * When position increments are counted, we're looking to make sure that we:
Token t1 = new Token(); // Don't count tokens without an increment - don't count tokens without an increment
- count normal tokens with one increment
- count funny tokens with more than one increment
- count the final token increments on the rare token streams that have them
*/
public void testCountPositionsWithIncrements() throws IOException {
Analyzer analyzer = createMockAnalyzer();
assertThat(TokenCountFieldMapper.countPositions(analyzer, "", "", true), equalTo(7));
}
/**
* When position increments are not counted (only positions are counted), we're looking to make sure that we:
- don't count tokens without an increment
- count normal tokens with one increment
- count funny tokens with more than one increment as only one
- don't count the final token increments on the rare token streams that have them
*/
public void testCountPositionsWithoutIncrements() throws IOException {
Analyzer analyzer = createMockAnalyzer();
assertThat(TokenCountFieldMapper.countPositions(analyzer, "", "", false), equalTo(2));
}
private Analyzer createMockAnalyzer() {
Token t1 = new Token(); // Token without an increment
t1.setPositionIncrement(0); t1.setPositionIncrement(0);
Token t2 = new Token(); Token t2 = new Token();
t2.setPositionIncrement(1); // Count normal tokens with one increment t2.setPositionIncrement(1); // Normal token with one increment
Token t3 = new Token(); Token t3 = new Token();
t2.setPositionIncrement(2); // Count funny tokens with more than one increment t2.setPositionIncrement(2); // Funny token with more than one increment
int finalTokenIncrement = 4; // Count the final token increment on the rare token streams that have them int finalTokenIncrement = 4; // Final token increment
Token[] tokens = new Token[] {t1, t2, t3}; Token[] tokens = new Token[] {t1, t2, t3};
Collections.shuffle(Arrays.asList(tokens), random()); Collections.shuffle(Arrays.asList(tokens), random());
final TokenStream tokenStream = new CannedTokenStream(finalTokenIncrement, 0, tokens); final TokenStream tokenStream = new CannedTokenStream(finalTokenIncrement, 0, tokens);
@ -99,7 +117,7 @@ public class TokenCountFieldMapperTests extends ESSingleNodeTestCase {
return new TokenStreamComponents(new MockTokenizer(), tokenStream); return new TokenStreamComponents(new MockTokenizer(), tokenStream);
} }
}; };
assertThat(TokenCountFieldMapper.countPositions(analyzer, "", ""), equalTo(7)); return analyzer;
} }
@Override @Override

View File

@ -85,8 +85,7 @@ public class QueryStringQueryBuilderTests extends AbstractQueryTestCase<QueryStr
if (randomBoolean()) { if (randomBoolean()) {
queryStringQueryBuilder.defaultField(randomBoolean() ? queryStringQueryBuilder.defaultField(randomBoolean() ?
STRING_FIELD_NAME : randomAlphaOfLengthBetween(1, 10)); STRING_FIELD_NAME : randomAlphaOfLengthBetween(1, 10));
} } else {
if (randomBoolean()) {
int numFields = randomIntBetween(1, 5); int numFields = randomIntBetween(1, 5);
for (int i = 0; i < numFields; i++) { for (int i = 0; i < numFields; i++) {
String fieldName = randomBoolean() ? STRING_FIELD_NAME : randomAlphaOfLengthBetween(1, 10); String fieldName = randomBoolean() ? STRING_FIELD_NAME : randomAlphaOfLengthBetween(1, 10);
@ -324,7 +323,6 @@ public class QueryStringQueryBuilderTests extends AbstractQueryTestCase<QueryStr
MapperQueryParser queryParser = new MapperQueryParser(createShardContext()); MapperQueryParser queryParser = new MapperQueryParser(createShardContext());
QueryParserSettings settings = new QueryParserSettings("first foo-bar-foobar* last"); QueryParserSettings settings = new QueryParserSettings("first foo-bar-foobar* last");
settings.defaultField(STRING_FIELD_NAME); settings.defaultField(STRING_FIELD_NAME);
settings.fieldsAndWeights(Collections.emptyMap());
settings.analyzeWildcard(true); settings.analyzeWildcard(true);
settings.fuzziness(Fuzziness.AUTO); settings.fuzziness(Fuzziness.AUTO);
settings.rewriteMethod(MultiTermQuery.CONSTANT_SCORE_REWRITE); settings.rewriteMethod(MultiTermQuery.CONSTANT_SCORE_REWRITE);
@ -352,7 +350,6 @@ public class QueryStringQueryBuilderTests extends AbstractQueryTestCase<QueryStr
MapperQueryParser queryParser = new MapperQueryParser(createShardContext()); MapperQueryParser queryParser = new MapperQueryParser(createShardContext());
QueryParserSettings settings = new QueryParserSettings("first foo-bar-foobar* last"); QueryParserSettings settings = new QueryParserSettings("first foo-bar-foobar* last");
settings.defaultField(STRING_FIELD_NAME); settings.defaultField(STRING_FIELD_NAME);
settings.fieldsAndWeights(Collections.emptyMap());
settings.analyzeWildcard(true); settings.analyzeWildcard(true);
settings.fuzziness(Fuzziness.AUTO); settings.fuzziness(Fuzziness.AUTO);
settings.rewriteMethod(MultiTermQuery.CONSTANT_SCORE_REWRITE); settings.rewriteMethod(MultiTermQuery.CONSTANT_SCORE_REWRITE);
@ -390,7 +387,6 @@ public class QueryStringQueryBuilderTests extends AbstractQueryTestCase<QueryStr
MapperQueryParser queryParser = new MapperQueryParser(createShardContext()); MapperQueryParser queryParser = new MapperQueryParser(createShardContext());
QueryParserSettings settings = new QueryParserSettings(""); QueryParserSettings settings = new QueryParserSettings("");
settings.defaultField(STRING_FIELD_NAME); settings.defaultField(STRING_FIELD_NAME);
settings.fieldsAndWeights(Collections.emptyMap());
settings.fuzziness(Fuzziness.AUTO); settings.fuzziness(Fuzziness.AUTO);
settings.analyzeWildcard(true); settings.analyzeWildcard(true);
settings.rewriteMethod(MultiTermQuery.CONSTANT_SCORE_REWRITE); settings.rewriteMethod(MultiTermQuery.CONSTANT_SCORE_REWRITE);
@ -689,6 +685,29 @@ public class QueryStringQueryBuilderTests extends AbstractQueryTestCase<QueryStr
assertThat(phraseQuery.getTerms().length, equalTo(2)); assertThat(phraseQuery.getTerms().length, equalTo(2));
} }
public void testToQueryWildcardNonExistingFields() throws IOException {
assumeTrue("test runs only when at least a type is registered", getCurrentTypes().length > 0);
QueryStringQueryBuilder queryStringQueryBuilder =
new QueryStringQueryBuilder("foo bar").field("invalid*");
Query query = queryStringQueryBuilder.toQuery(createShardContext());
Query expectedQuery = new BooleanQuery.Builder()
.add(new MatchNoDocsQuery("empty fields"), Occur.SHOULD)
.add(new MatchNoDocsQuery("empty fields"), Occur.SHOULD)
.build();
assertThat(expectedQuery, equalTo(query));
queryStringQueryBuilder =
new QueryStringQueryBuilder("field:foo bar").field("invalid*");
query = queryStringQueryBuilder.toQuery(createShardContext());
expectedQuery = new BooleanQuery.Builder()
.add(new TermQuery(new Term("field", "foo")), Occur.SHOULD)
.add(new MatchNoDocsQuery("empty fields"), Occur.SHOULD)
.build();
assertThat(expectedQuery, equalTo(query));
}
public void testToQuerySplitOnWhitespace() throws IOException { public void testToQuerySplitOnWhitespace() throws IOException {
assumeTrue("test runs only when at least a type is registered", getCurrentTypes().length > 0); assumeTrue("test runs only when at least a type is registered", getCurrentTypes().length > 0);
// splitOnWhitespace=false // splitOnWhitespace=false

View File

@ -23,25 +23,28 @@ import org.apache.lucene.document.Document;
import org.apache.lucene.index.DirectoryReader; import org.apache.lucene.index.DirectoryReader;
import org.apache.lucene.index.IndexWriter; import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.LeafReaderContext; import org.apache.lucene.index.LeafReaderContext;
import org.apache.lucene.index.Term;
import org.apache.lucene.search.ConstantScoreScorer; import org.apache.lucene.search.ConstantScoreScorer;
import org.apache.lucene.search.ConstantScoreWeight; import org.apache.lucene.search.ConstantScoreWeight;
import org.apache.lucene.search.DocIdSetIterator; import org.apache.lucene.search.DocIdSetIterator;
import org.apache.lucene.search.Explanation;
import org.apache.lucene.search.IndexSearcher; import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.MatchAllDocsQuery;
import org.apache.lucene.search.Query; import org.apache.lucene.search.Query;
import org.apache.lucene.search.QueryCachingPolicy; import org.apache.lucene.search.QueryCachingPolicy;
import org.apache.lucene.search.Scorer; import org.apache.lucene.search.Scorer;
import org.apache.lucene.search.ScorerSupplier;
import org.apache.lucene.search.Weight; import org.apache.lucene.search.Weight;
import org.apache.lucene.store.Directory; import org.apache.lucene.store.Directory;
import org.apache.lucene.util.IOUtils; import org.apache.lucene.util.IOUtils;
import org.elasticsearch.common.lucene.index.ElasticsearchDirectoryReader; import org.elasticsearch.common.lucene.index.ElasticsearchDirectoryReader;
import org.elasticsearch.common.settings.Settings; import org.elasticsearch.common.settings.Settings;
import org.elasticsearch.index.IndexModule;
import org.elasticsearch.index.cache.query.QueryCacheStats; import org.elasticsearch.index.cache.query.QueryCacheStats;
import org.elasticsearch.index.shard.ShardId; import org.elasticsearch.index.shard.ShardId;
import org.elasticsearch.indices.IndicesQueryCache;
import org.elasticsearch.test.ESTestCase; import org.elasticsearch.test.ESTestCase;
import java.io.IOException; import java.io.IOException;
import java.util.Set;
public class IndicesQueryCacheTests extends ESTestCase { public class IndicesQueryCacheTests extends ESTestCase {
@ -328,4 +331,76 @@ public class IndicesQueryCacheTests extends ESTestCase {
cache.close(); // this triggers some assertions cache.close(); // this triggers some assertions
} }
private static class DummyWeight extends Weight {
private final Weight weight;
private boolean scorerCalled;
private boolean scorerSupplierCalled;
DummyWeight(Weight weight) {
super(weight.getQuery());
this.weight = weight;
}
@Override
public void extractTerms(Set<Term> terms) {
weight.extractTerms(terms);
}
@Override
public Explanation explain(LeafReaderContext context, int doc) throws IOException {
return weight.explain(context, doc);
}
@Override
public Scorer scorer(LeafReaderContext context) throws IOException {
scorerCalled = true;
return weight.scorer(context);
}
@Override
public ScorerSupplier scorerSupplier(LeafReaderContext context) throws IOException {
scorerSupplierCalled = true;
return weight.scorerSupplier(context);
}
}
public void testDelegatesScorerSupplier() throws Exception {
Directory dir = newDirectory();
IndexWriter w = new IndexWriter(dir, newIndexWriterConfig());
w.addDocument(new Document());
DirectoryReader r = DirectoryReader.open(w);
w.close();
ShardId shard = new ShardId("index", "_na_", 0);
r = ElasticsearchDirectoryReader.wrap(r, shard);
IndexSearcher s = new IndexSearcher(r);
s.setQueryCachingPolicy(new QueryCachingPolicy() {
@Override
public boolean shouldCache(Query query) throws IOException {
return false; // never cache
}
@Override
public void onUse(Query query) {}
});
Settings settings = Settings.builder()
.put(IndicesQueryCache.INDICES_CACHE_QUERY_COUNT_SETTING.getKey(), 10)
.put(IndicesQueryCache.INDICES_QUERIES_CACHE_ALL_SEGMENTS_SETTING.getKey(), true)
.build();
IndicesQueryCache cache = new IndicesQueryCache(settings);
s.setQueryCache(cache);
Query query = new MatchAllDocsQuery();
final DummyWeight weight = new DummyWeight(s.createNormalizedWeight(query, false));
final Weight cached = cache.doCache(weight, s.getQueryCachingPolicy());
assertNotSame(weight, cached);
assertFalse(weight.scorerCalled);
assertFalse(weight.scorerSupplierCalled);
cached.scorerSupplier(s.getIndexReader().leaves().get(0));
assertFalse(weight.scorerCalled);
assertTrue(weight.scorerSupplierCalled);
IOUtils.close(r, dir);
cache.onClose(shard);
cache.close();
}
} }

View File

@ -32,13 +32,12 @@ public class InternalStatsTests extends InternalAggregationTestCase<InternalStat
@Override @Override
protected InternalStats createTestInstance(String name, List<PipelineAggregator> pipelineAggregators, protected InternalStats createTestInstance(String name, List<PipelineAggregator> pipelineAggregators,
Map<String, Object> metaData) { Map<String, Object> metaData) {
long count = randomIntBetween(1, 50); long count = frequently() ? randomIntBetween(1, Integer.MAX_VALUE) : 0;
double[] minMax = new double[2]; double min = randomDoubleBetween(-1000000, 1000000, true);
minMax[0] = randomDouble(); double max = randomDoubleBetween(-1000000, 1000000, true);
minMax[0] = randomDouble(); double sum = randomDoubleBetween(-1000000, 1000000, true);
double sum = randomDoubleBetween(0, 100, true); DocValueFormat format = randomNumericDocValueFormat();
return new InternalStats(name, count, sum, minMax[0], minMax[1], DocValueFormat.RAW, return new InternalStats(name, count, sum, min, max, format, pipelineAggregators, Collections.emptyMap());
pipelineAggregators, Collections.emptyMap());
} }
@Override @Override
@ -58,7 +57,7 @@ public class InternalStatsTests extends InternalAggregationTestCase<InternalStat
expectedSum += stats.getSum(); expectedSum += stats.getSum();
} }
assertEquals(expectedCount, reduced.getCount()); assertEquals(expectedCount, reduced.getCount());
assertEquals(expectedSum, reduced.getSum(), 1e-10); assertEquals(expectedSum, reduced.getSum(), 1e-7);
assertEquals(expectedMin, reduced.getMin(), 0d); assertEquals(expectedMin, reduced.getMin(), 0d);
assertEquals(expectedMax, reduced.getMax(), 0d); assertEquals(expectedMax, reduced.getMax(), 0d);
} }

View File

@ -48,12 +48,6 @@ GET my_index/_search
<2> The `name.length` field is a `token_count` <<multi-fields,multi-field>> which will index the number of tokens in the `name` field. <2> The `name.length` field is a `token_count` <<multi-fields,multi-field>> which will index the number of tokens in the `name` field.
<3> This query matches only the document containing `Rachel Alice Williams`, as it contains three tokens. <3> This query matches only the document containing `Rachel Alice Williams`, as it contains three tokens.
[NOTE]
===================================================================
Technically the `token_count` type sums position increments rather than
counting tokens. This means that even if the analyzer filters out stop
words they are included in the count.
===================================================================
[[token-count-params]] [[token-count-params]]
==== Parameters for `token_count` fields ==== Parameters for `token_count` fields
@ -68,6 +62,12 @@ The following parameters are accepted by `token_count` fields:
value. Required. For best performance, use an analyzer without token value. Required. For best performance, use an analyzer without token
filters. filters.
`enable_position_increments`::
Indicates if position increments should be counted.
Set to `false` if you don't want to count tokens removed by analyzer filters (like <<analysis-stop-tokenfilter,`stop`>>).
Defaults to `true`.
<<mapping-boost,`boost`>>:: <<mapping-boost,`boost`>>::
Mapping field-level query time boosting. Accepts a floating point number, defaults Mapping field-level query time boosting. Accepts a floating point number, defaults

View File

@ -1338,9 +1338,6 @@ public abstract class ESIntegTestCase extends ESTestCase {
indexRandom(forceRefresh, dummyDocuments, Arrays.asList(builders)); indexRandom(forceRefresh, dummyDocuments, Arrays.asList(builders));
} }
private static final String RANDOM_BOGUS_TYPE = "RANDOM_BOGUS_TYPE______";
/** /**
* Indexes the given {@link IndexRequestBuilder} instances randomly. It shuffles the given builders and either * Indexes the given {@link IndexRequestBuilder} instances randomly. It shuffles the given builders and either
* indexes them in a blocking or async fashion. This is very useful to catch problems that relate to internal document * indexes them in a blocking or async fashion. This is very useful to catch problems that relate to internal document
@ -1388,31 +1385,33 @@ public abstract class ESIntegTestCase extends ESTestCase {
* @param builders the documents to index. * @param builders the documents to index.
*/ */
public void indexRandom(boolean forceRefresh, boolean dummyDocuments, boolean maybeFlush, List<IndexRequestBuilder> builders) throws InterruptedException, ExecutionException { public void indexRandom(boolean forceRefresh, boolean dummyDocuments, boolean maybeFlush, List<IndexRequestBuilder> builders) throws InterruptedException, ExecutionException {
Random random = random(); Random random = random();
Set<String> indicesSet = new HashSet<>(); Map<String, Set<String>> indicesAndTypes = new HashMap<>();
for (IndexRequestBuilder builder : builders) { for (IndexRequestBuilder builder : builders) {
indicesSet.add(builder.request().index()); final Set<String> types = indicesAndTypes.computeIfAbsent(builder.request().index(), index -> new HashSet<>());
types.add(builder.request().type());
} }
Set<Tuple<String, String>> bogusIds = new HashSet<>(); Set<List<String>> bogusIds = new HashSet<>(); // (index, type, id)
if (random.nextBoolean() && !builders.isEmpty() && dummyDocuments) { if (random.nextBoolean() && !builders.isEmpty() && dummyDocuments) {
builders = new ArrayList<>(builders); builders = new ArrayList<>(builders);
final String[] indices = indicesSet.toArray(new String[indicesSet.size()]);
// inject some bogus docs // inject some bogus docs
final int numBogusDocs = scaledRandomIntBetween(1, builders.size() * 2); final int numBogusDocs = scaledRandomIntBetween(1, builders.size() * 2);
final int unicodeLen = between(1, 10); final int unicodeLen = between(1, 10);
for (int i = 0; i < numBogusDocs; i++) { for (int i = 0; i < numBogusDocs; i++) {
String id = randomRealisticUnicodeOfLength(unicodeLen) + Integer.toString(dummmyDocIdGenerator.incrementAndGet()); String id = "bogus_doc_" + randomRealisticUnicodeOfLength(unicodeLen) + Integer.toString(dummmyDocIdGenerator.incrementAndGet());
String index = RandomPicks.randomFrom(random, indices); Map.Entry<String, Set<String>> indexAndTypes = RandomPicks.randomFrom(random, indicesAndTypes.entrySet());
bogusIds.add(new Tuple<>(index, id)); String index = indexAndTypes.getKey();
builders.add(client().prepareIndex(index, RANDOM_BOGUS_TYPE, id).setSource("{}", XContentType.JSON)); String type = RandomPicks.randomFrom(random, indexAndTypes.getValue());
bogusIds.add(Arrays.asList(index, type, id));
// We configure a routing key in case the mapping requires it
builders.add(client().prepareIndex(index, type, id).setSource("{}", XContentType.JSON).setRouting(id));
} }
} }
final String[] indices = indicesSet.toArray(new String[indicesSet.size()]);
Collections.shuffle(builders, random()); Collections.shuffle(builders, random());
final CopyOnWriteArrayList<Tuple<IndexRequestBuilder, Exception>> errors = new CopyOnWriteArrayList<>(); final CopyOnWriteArrayList<Tuple<IndexRequestBuilder, Exception>> errors = new CopyOnWriteArrayList<>();
List<CountDownLatch> inFlightAsyncOperations = new ArrayList<>(); List<CountDownLatch> inFlightAsyncOperations = new ArrayList<>();
// If you are indexing just a few documents then frequently do it one at a time. If many then frequently in bulk. // If you are indexing just a few documents then frequently do it one at a time. If many then frequently in bulk.
final String[] indices = indicesAndTypes.keySet().toArray(new String[0]);
if (builders.size() < FREQUENT_BULK_THRESHOLD ? frequently() : builders.size() < ALWAYS_BULK_THRESHOLD ? rarely() : false) { if (builders.size() < FREQUENT_BULK_THRESHOLD ? frequently() : builders.size() < ALWAYS_BULK_THRESHOLD ? rarely() : false) {
if (frequently()) { if (frequently()) {
logger.info("Index [{}] docs async: [{}] bulk: [{}]", builders.size(), true, false); logger.info("Index [{}] docs async: [{}] bulk: [{}]", builders.size(), true, false);
@ -1454,10 +1453,10 @@ public abstract class ESIntegTestCase extends ESTestCase {
assertThat(actualErrors, emptyIterable()); assertThat(actualErrors, emptyIterable());
if (!bogusIds.isEmpty()) { if (!bogusIds.isEmpty()) {
// delete the bogus types again - it might trigger merges or at least holes in the segments and enforces deleted docs! // delete the bogus types again - it might trigger merges or at least holes in the segments and enforces deleted docs!
for (Tuple<String, String> doc : bogusIds) { for (List<String> doc : bogusIds) {
assertEquals("failed to delete a dummy doc [" + doc.v1() + "][" + doc.v2() + "]", assertEquals("failed to delete a dummy doc [" + doc.get(0) + "][" + doc.get(2) + "]",
DocWriteResponse.Result.DELETED, DocWriteResponse.Result.DELETED,
client().prepareDelete(doc.v1(), RANDOM_BOGUS_TYPE, doc.v2()).get().getResult()); client().prepareDelete(doc.get(0), doc.get(1), doc.get(2)).setRouting(doc.get(2)).get().getResult());
} }
} }
if (forceRefresh) { if (forceRefresh) {