Apache
/
lucene
mirror of https://github.com/apache/lucene.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

SOLR-1132: Added support for poly fields 

git-svn-id: https://svn.apache.org/repos/asf/lucene/solr/trunk@893746 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Grant Ingersoll 2009-12-24 13:03:22 +00:00
parent 9a5a4ed7b4
commit 05237cf80f
36 changed files with 1817 additions and 341 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
CHANGES.txt
View File

@ -74,6 +74,9 @@ New Features
* SOLR-1653: Add PatternReplaceCharFilter (koji)
* SOLR-1131: FieldTypes can now output multiple Fields per Type and still be searched. This can be handy for hiding the details of a particular
implementation such as in the spatial case. (Chris Mattmann, shalin, noble, gsingers, yonik)
Optimizations
----------------------

3
example/exampledocs/hd.xml
View File

@ -28,6 +28,7 @@
<field name="popularity">6</field>
<field name="inStock">true</field>
<field name="manufacturedate_dt">2006-02-13T15:26:37Z</field>
<field name="store">45.17614,-93.87341</field>
</doc>
<doc>
@ -42,6 +43,8 @@
<field name="price">350</field>
<field name="popularity">6</field>
<field name="inStock">true</field>
<!-- Buffalo store -->
<field name="store">45.17614,-93.87341</field>
<field name="manufacturedate_dt">2006-02-13T15:26:37Z</field>
</doc>
</add>

4
example/exampledocs/ipod_other.xml
View File

@ -28,6 +28,8 @@
<field name="price">19.95</field>
<field name="popularity">1</field>
<field name="inStock">false</field>
<!-- Buffalo store -->
<field name="store">45.17614,-93.87341</field>
<field name="manufacturedate_dt">2005-08-01T16:30:25Z</field>
</doc>
@ -42,6 +44,8 @@
<field name="price">11.50</field>
<field name="popularity">1</field>
<field name="inStock">false</field>
<!-- San Francisco store -->
<field name="store">37.7752,-122.4232</field>
<field name="manufacturedate_dt">2006-02-14T23:55:59Z</field>
</doc>

2
example/exampledocs/ipod_video.xml
View File

@ -32,5 +32,7 @@
<field name="price">399.00</field>
<field name="popularity">10</field>
<field name="inStock">true</field>
<!-- San Francisco store -->
<field name="store">37.7752,-122.4232</field>
<field name="manufacturedate_dt">2005-10-12T08:00:00Z</field>
</doc></add>

6
example/exampledocs/mem.xml
View File

@ -26,6 +26,8 @@
<field name="price">185</field>
<field name="popularity">5</field>
<field name="inStock">true</field>
<!-- San Francisco store -->
<field name="store">37.7752,-122.4232</field>
<field name="manufacturedate_dt">2006-02-13T15:26:37Z</field>
</doc>
@ -38,6 +40,8 @@
<field name="price">74.99</field>
<field name="popularity">7</field>
<field name="inStock">true</field>
<!-- San Francisco store -->
<field name="store">37.7752,-122.4232</field>
<field name="manufacturedate_dt">2006-02-13T15:26:37Z</field>
</doc>
@ -51,6 +55,8 @@
<!-- note: price & popularity is missing on this one -->
<field name="popularity">0</field>
<field name="inStock">true</field>
<!-- Buffalo store -->
<field name="store">45.17614,-93.87341</field>
<field name="manufacturedate_dt">2006-02-13T15:26:37Z</field>
</doc>

2
example/exampledocs/monitor.xml
View File

@ -27,5 +27,7 @@
<field name="price">2199</field>
<field name="popularity">6</field>
<field name="inStock">true</field>
<!-- Buffalo store -->
<field name="store">45.17614,-93.87341</field>
</doc></add>

2
example/exampledocs/monitor2.xml
View File

@ -26,5 +26,7 @@
<field name="price">279.95</field>
<field name="popularity">6</field>
<field name="inStock">true</field>
<!-- Buffalo store -->
<field name="store">45.17614,-93.87341</field>
</doc></add>

2
example/exampledocs/mp500.xml
View File

@ -35,5 +35,7 @@
<field name="price">179.99</field>
<field name="popularity">6</field>
<field name="inStock">true</field>
<!-- Buffalo store -->
<field name="store">45.17614,-93.87341</field>
</doc></add>

2
example/exampledocs/sd500.xml
View File

@ -31,4 +31,6 @@
<field name="popularity">7</field>
<field name="inStock">true</field>
<field name="manufacturedate_dt">2006-02-13T15:26:37Z</field>
<!-- Buffalo store -->
<field name="store">45.17614,-93.87341</field>
</doc></add>

3
example/exampledocs/vidcard.xml
View File

@ -30,6 +30,7 @@
<field name="weight">16</field>
<field name="price">479.95</field>
<field name="popularity">7</field>
<field name="store">40.7143,-74.006</field>
<field name="inStock">false</field>
<field name="manufacturedate_dt">2006-02-13T15:26:37Z/DAY</field>
</doc>
@ -50,5 +51,7 @@
<field name="popularity">7</field>
<field name="inStock">false</field>
<field name="manufacturedate_dt">2006-02-13T15:26:37Z/DAY</field>
<!-- NYC store -->
<field name="store">40.7143,-74.006</field>
</doc>
</add>

9
example/solr/conf/schema.xml
View File

@ -394,7 +394,12 @@
<!-- since fields of this type are by default not stored or indexed,
any data added to them will be ignored outright. -->
<fieldtype name="ignored" stored="false" indexed="false" multiValued="true" class="solr.StrField" />
<fieldtype name="ignored" stored="false" indexed="false" multiValued="true" class="solr.StrField" />
<!--
A PointType is a Poly Field. It can either declare a subFieldType or a subFieldSuffix
-->
<fieldType name="location" class="solr.PointType" dimension="2" subFieldType="double"/>
</types>
@ -440,6 +445,8 @@
<field name="popularity" type="int" indexed="true" stored="true" />
<field name="inStock" type="boolean" indexed="true" stored="true" />
<field name="store" type="location" indexed="true" stored="true"/>
<!-- Common metadata fields, named specifically to match up with
SolrCell metadata when parsing rich documents such as Word, PDF.

2
src/common/org/apache/solr/common/SolrException.java
View File

@ -38,7 +38,7 @@ public class SolrException extends RuntimeException {
SERVER_ERROR( 500 ),
SERVICE_UNAVAILABLE( 503 ),
UNKNOWN(0);
final int code;
public final int code;
private ErrorCode( int c )
{

138
src/java/org/apache/solr/schema/CoordinateFieldType.java Normal file
View File

@ -0,0 +1,138 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF 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.apache.solr.schema;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.TermQuery;
import org.apache.lucene.index.Term;
import org.apache.solr.search.QParser;
import org.apache.solr.search.function.ValueSource;
import org.apache.solr.common.SolrException;
import org.apache.solr.common.params.SolrParams;
import org.apache.solr.common.params.MapSolrParams;
import java.util.Map;
import java.util.List;
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashMap;
/**
* A CoordinateFieldType is the base class for {@link org.apache.solr.schema.FieldType}s that have semantics
* related to items in a coordinate system.
* <br/>
* Implementations depend on a delegating work to a sub {@link org.apache.solr.schema.FieldType}, specified by
* either the {@link #SUB_FIELD_SUFFIX} or the {@link #SUB_FIELD_TYPE} (the latter is used if both are defined.
* <br/>
* Example:
* <pre>&lt;fieldType name="xy" class="solr.PointType" dimension="2" subFieldType="double"/&gt;
* </pre>
* In theory, classes deriving from this should be able to do things like represent a point, a polygon, a line, etc.
* <br/>
* NOTE: There can only be one sub Field Type.
*
*/
public abstract class CoordinateFieldType extends FieldType implements SchemaAware {
/**
* The dimension of the coordinate system
*/
protected int dimension;
protected FieldType subType;
public static final String SUB_FIELD_SUFFIX = "subFieldSuffix";
public static final String SUB_FIELD_TYPE = "subFieldType";
private String suffix;//need to keep this around between init and inform, since dynamic fields aren't created until before inform
protected int dynFieldProps;
public int getDimension() {
return dimension;
}
public FieldType getSubType() {
return subType;
}
@Override
protected void init(IndexSchema schema, Map<String, String> args) {
//it's not a first class citizen for the IndexSchema
SolrParams p = new MapSolrParams(args);
String subFT = p.get(SUB_FIELD_TYPE);
String subSuffix = p.get(SUB_FIELD_SUFFIX);
if (subFT != null) {
args.remove(SUB_FIELD_TYPE);
subType = schema.getFieldTypeByName(subFT.trim());
} else if (subSuffix != null) {
args.remove(SUB_FIELD_SUFFIX);
suffix = subSuffix;
}else {
throw new SolrException(SolrException.ErrorCode.SERVER_ERROR, "The field type: " + typeName
+ " must specify the " +
SUB_FIELD_TYPE + " attribute or the " + SUB_FIELD_SUFFIX + " attribute.");
}
super.init(schema, args);
}
public void inform(IndexSchema schema) {
//Can't do this until here b/c the Dynamic Fields are not initialized until here.
if (suffix != null){
SchemaField sf = schema.getField(suffix);
subType = sf.getType();//this means it is already registered
dynFieldProps = sf.getProperties();
}
else if (subType != null) {
SchemaField proto = registerPolyFieldDynamicPrototype(schema, subType);
dynFieldProps = proto.getProperties();
} else {
throw new SolrException(SolrException.ErrorCode.SERVER_ERROR, "The field type: " + typeName
+ " must specify the " +
SUB_FIELD_TYPE + " attribute or the " + SUB_FIELD_SUFFIX + " attribute.");
}
}
/**
* Helper method for creating a dynamic field SchemaField prototype. Returns a {@link org.apache.solr.schema.SchemaField} with
* the {@link org.apache.solr.schema.FieldType} given and a name of "*" + {@link org.apache.solr.schema.FieldType#POLY_FIELD_SEPARATOR} + {@link org.apache.solr.schema.FieldType#typeName}
* and props of indexed=true, stored=false.
* @param schema the IndexSchema
* @param type The {@link org.apache.solr.schema.FieldType} of the prototype.
* @return The {@link org.apache.solr.schema.SchemaField}
*/
static SchemaField registerPolyFieldDynamicPrototype(IndexSchema schema, FieldType type){
String name = "*" + FieldType.POLY_FIELD_SEPARATOR + type.typeName;
Map<String, String> props = new HashMap<String, String>();
//Just set these, delegate everything else to the field type
props.put("indexed", "true");
props.put("stored", "false");
int p = SchemaField.calcProps(name, type, props);
SchemaField proto = SchemaField.create(name,
type, p, null);
schema.registerDynamicField(proto);
return proto;
}
/**
* Throws UnsupportedOperationException()
*/
public Query getFieldQuery(QParser parser, SchemaField field, String externalVal) {
throw new UnsupportedOperationException();
}
}

209
src/java/org/apache/solr/schema/FieldType.java
View File

@ -20,14 +20,14 @@ package org.apache.solr.schema;
import org.apache.lucene.document.Field;
import org.apache.lucene.document.Fieldable;
import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.TokenStream;
import org.apache.lucene.analysis.Tokenizer;
import org.apache.lucene.analysis.Token;
import org.apache.lucene.analysis.tokenattributes.TermAttribute;
import org.apache.lucene.analysis.tokenattributes.OffsetAttribute;
import org.apache.lucene.search.SortField;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.TermRangeQuery;
import org.apache.lucene.search.TermQuery;
import org.apache.lucene.index.Term;
import org.apache.solr.search.function.ValueSource;
import org.apache.solr.search.function.OrdFieldSource;
import org.apache.solr.search.Sorting;
@ -36,11 +36,17 @@ import org.apache.solr.request.XMLWriter;
import org.apache.solr.request.TextResponseWriter;
import org.apache.solr.analysis.SolrAnalyzer;
import org.apache.solr.common.SolrException;
import org.apache.solr.common.params.SolrParams;
import org.apache.solr.common.params.MapSolrParams;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.Map;
import java.util.HashMap;
import java.util.Collection;
import java.util.Collections;
import java.util.List;
import java.util.ArrayList;
import java.io.Reader;
import java.io.IOException;
@ -52,6 +58,14 @@ import java.io.IOException;
public abstract class FieldType extends FieldProperties {
public static final Logger log = LoggerFactory.getLogger(FieldType.class);
/**
* The default poly field separator.
*
* @see #createFields(SchemaField, String, float)
* @see #isPolyField()
*/
public static final String POLY_FIELD_SEPARATOR = "___";
/** The name of the type (not the name of the field) */
protected String typeName;
/** additional arguments specified in the field type declaration */
@ -59,9 +73,10 @@ public abstract class FieldType extends FieldProperties {
/** properties explicitly set to true */
protected int trueProperties;
/** properties explicitly set to false */
protected int falseProperties;
protected int falseProperties;
int properties;
/** Returns true if fields of this type should be tokenized */
public boolean isTokenized() {
return (properties & TOKENIZED) != 0;
@ -72,6 +87,18 @@ public abstract class FieldType extends FieldProperties {
return (properties & MULTIVALUED) != 0;
}
/**
* A "polyField" is a FieldType that can produce more than one Field per FieldType, via the {@link #createFields(org.apache.solr.schema.SchemaField, String, float)} method. This is useful
* when hiding the implementation details of a field from the Solr end user. For instance, a spatial point may be represented by three different field types, all of which may produce 1 or more
* fields.
* @return true if the {@link #createFields(org.apache.solr.schema.SchemaField, String, float)} method may return more than one field
*/
public boolean isPolyField(){
return false;
}
/** Returns true if a single field value of this type has multiple logical values
* for the purposes of faceting, sorting, etc. Text fields normally return
* true since each token/word is a logical value.
@ -85,7 +112,8 @@ public abstract class FieldType extends FieldProperties {
* Common boolean properties have already been handled.
*
*/
protected void init(IndexSchema schema, Map<String,String> args) {
protected void init(IndexSchema schema, Map<String, String> args) {
}
protected String getArg(String n, Map<String,String> args) {
@ -191,8 +219,15 @@ public abstract class FieldType extends FieldProperties {
* :TODO: clean up and clarify this explanation.
*
* @see #toInternal
*
*
*/
public Field createField(SchemaField field, String externalVal, float boost) {
if (!field.indexed() && !field.stored()) {
if (log.isTraceEnabled())
log.trace("Ignoring unindexed/unstored field: " + field);
return null;
}
String val;
try {
val = toInternal(externalVal);
@ -200,23 +235,123 @@ public abstract class FieldType extends FieldProperties {
throw new SolrException( SolrException.ErrorCode.SERVER_ERROR, "Error while creating field '" + field + "' from value '" + externalVal + "'", e, false);
}
if (val==null) return null;
if (!field.indexed() && !field.stored()) {
if (log.isTraceEnabled())
log.trace("Ignoring unindexed/unstored field: " + field);
return null;
return createField(field.getName(), val, getFieldStore(field, val),
getFieldIndex(field, val), getFieldTermVec(field, val), field.omitNorms(),
field.omitTf(), boost);
}
/**
* Create multiple fields from a single field and multiple values. Fields are named as SchemaField.getName() + {@link #POLY_FIELD_SEPARATOR} + i, where
* i starts at 0.
* <p/>
* If the field is stored, then an extra field gets created that contains the storageVal. It is this field that also
*
* @param field The {@link org.apache.solr.schema.SchemaField}
* @param props The properties to use
* @param delegatedType An optional type to use. If null, then field.getType() is used. Useful for poly fields.
* @param storageVal If the field stores, then this value will be used for the stored field
* @param boost The boost to apply to all fields
* @param externalVals The values to use
* @return The fields
*/
protected Fieldable[] createFields(SchemaField field, int props,
FieldType delegatedType, String storageVal,
float boost, String ... externalVals) {
int n = field.indexed() ? externalVals.length : 0;
n += field.stored() ? 1 : 0;
if (delegatedType == null) { //if the type isn't being overriden, then just use the base one
delegatedType = field.getType();
}
Field[] results = new Field[n];
//Field.Store.NO,Field.Index.NOT_ANALYZED_NO_NORMS, Field.TermVector.NO, true, true
if (externalVals.length > 0) {
if (field.indexed()) {
String name = field.getName() + "_";
String suffix = POLY_FIELD_SEPARATOR + delegatedType.typeName;
Field f = new Field(field.getName(),
int len = name.length();
StringBuilder bldr = new StringBuilder(len + 3 + suffix.length());//should be enough buffer to handle most values of j.
bldr.append(name);
for (int j = 0; j < externalVals.length; j++) {
//SchemaField is final, as is name, so we need to recreate each time
//put the counter before the separator, b/c dynamic fields can't be asterisks on both the front and the end of the String
bldr.append(j).append(suffix);
SchemaField sf = SchemaField.create(bldr.toString(),
delegatedType, props, null);
//schema.getDynamicField(name + "_" + j + POLY_FIELD_SEPARATOR + delegatedType.typeName);
/**/
//new SchemaField(name, ft, p, defaultValue )
//QUESTION: should we allow for vectors, etc? Not sure that it makes sense
results[j] = delegatedType.createField(sf, externalVals[j], boost);
bldr.setLength(len);//cut the builder back to just the length of the prefix, but keep the capacity
}
}
Field.TermVector fieldTermVec = getFieldTermVec(field, storageVal);
if (field.stored() || fieldTermVec.equals(Field.TermVector.YES)
|| fieldTermVec.equals(Field.TermVector.WITH_OFFSETS)
|| fieldTermVec.equals(Field.TermVector.WITH_POSITIONS)
|| fieldTermVec.equals(Field.TermVector.WITH_POSITIONS_OFFSETS)
) {
//QUESTION: should we allow for vectors, etc? Not sure that it makes sense
results[results.length - 1] = createField(field.getName(), storageVal, getFieldStore(field, storageVal),
Field.Index.NO,
fieldTermVec, field.omitNorms(), field.omitTf(), boost);
}
}
return results;
}
/**
* Create the field from native Lucene parts. Mostly intended for use by FieldTypes outputing multiple
* Fields per SchemaField
* @param name The name of the field
* @param val The _internal_ value to index
* @param storage {@link org.apache.lucene.document.Field.Store}
* @param index {@link org.apache.lucene.document.Field.Index}
* @param vec {@link org.apache.lucene.document.Field.TermVector}
* @param omitNorms true if norms should be omitted
* @param omitTFPos true if term freq and position should be omitted.
* @param boost The boost value
* @return the {@link org.apache.lucene.document.Field}.
*/
protected Field createField(String name, String val, Field.Store storage, Field.Index index,
Field.TermVector vec, boolean omitNorms, boolean omitTFPos, float boost){
Field f = new Field(name,
val,
getFieldStore(field, val),
getFieldIndex(field, val),
getFieldTermVec(field, val));
f.setOmitNorms(field.omitNorms());
f.setOmitTermFreqAndPositions(field.omitTf());
storage,
index,
vec);
f.setOmitNorms(omitNorms);
f.setOmitTermFreqAndPositions(omitTFPos);
f.setBoost(boost);
return f;
}
/**
* Given a {@link org.apache.solr.schema.SchemaField}, create one or more {@link org.apache.lucene.document.Field} instances
* @param field the {@link org.apache.solr.schema.SchemaField}
* @param externalVal The value to add to the field
* @param boost The boost to apply
* @return The {@link org.apache.lucene.document.Field} instances
*
* @see #createField(SchemaField, String, float)
* @see #isPolyField()
*/
public Fieldable[] createFields(SchemaField field, String externalVal, float boost) {
Field f = createField( field, externalVal, boost);
if( f != null ) {
return new Field[] { f };
}
return null;
}
/* Helpers for field construction */
protected Field.TermVector getFieldTermVec(SchemaField field,
String internalVal) {
@ -226,7 +361,7 @@ public abstract class FieldType extends FieldProperties {
else if (field.storeTermPositions())
ftv = Field.TermVector.WITH_POSITIONS;
else if (field.storeTermOffsets())
ftv = Field.TermVector.WITH_OFFSETS;
ftv = Field.TermVector.WITH_OFFSETS;
else if (field.storeTermVector())
ftv = Field.TermVector.YES;
return ftv;
@ -237,7 +372,7 @@ public abstract class FieldType extends FieldProperties {
}
protected Field.Index getFieldIndex(SchemaField field,
String internalVal) {
return field.indexed() ? (isTokenized() ? Field.Index.TOKENIZED :
return field.indexed() ? (isTokenized() ? Field.Index.TOKENIZED :
Field.Index.UN_TOKENIZED) : Field.Index.NO;
}
@ -265,7 +400,7 @@ public abstract class FieldType extends FieldProperties {
}
/**
* Convert the stored-field format to an external object.
* Convert the stored-field format to an external object.
* @see #toInternal
* @since solr 1.3
*/
@ -290,7 +425,7 @@ public abstract class FieldType extends FieldProperties {
// that the indexed form is the same as the stored field form.
return f.stringValue();
}
/** Given the readable value, return the term value that will match it. */
public String readableToIndexed(String val) {
return toInternal(val);
@ -321,7 +456,7 @@ public abstract class FieldType extends FieldProperties {
termAtt.setTermBuffer(s);
offsetAtt.setOffset(correctOffset(0),correctOffset(n));
return true;
}
}
};
return new TokenStreamInfo(ts, ts);
@ -335,7 +470,7 @@ public abstract class FieldType extends FieldProperties {
* @see #getAnalyzer
*/
protected Analyzer analyzer=new DefaultAnalyzer(256);
/**
* Analyzer set by schema for text types to use when searching fields
* of this type, subclasses can set analyzer themselves or override
@ -394,7 +529,7 @@ public abstract class FieldType extends FieldProperties {
*/
public abstract void write(TextResponseWriter writer, String name, Fieldable f) throws IOException;
/**
* Returns the SortField instance that should be used to sort fields
* of this type.
@ -452,4 +587,36 @@ public abstract class FieldType extends FieldProperties {
minInclusive, maxInclusive);
}
/**
* Returns a Query instance for doing searches against a field.
* @param parser The {@link org.apache.solr.search.QParser} calling the method
* @param field The {@link org.apache.solr.schema.SchemaField} of the field to search
* @param externalVal The String representation of the value to search
* @return The {@link org.apache.lucene.search.Query} instance. This implementation returns a {@link org.apache.lucene.search.TermQuery} but overriding queries may not
*
*/
public Query getFieldQuery(QParser parser, SchemaField field, String externalVal) {
return new TermQuery(new Term(field.getName(), toInternal(externalVal)));
}
/**
* Return a collection of all the Fields in the index where the {@link org.apache.solr.schema.SchemaField}
* @param polyField The instance of the {@link org.apache.solr.schema.SchemaField} to find the actual field names from
* @return The {@link java.util.Collection} of names of the actual fields that are a poly field.
*
*
*/
/*protected Collection<String> getPolyFieldNames(SchemaField polyField){
if (polyField.isPolyField()) {
if (polyField != null) {
//we need the names of all the fields. Do this lazily and then cache?
}
} //TODO: Should we throw an exception here in an else clause?
return Collections.emptyList();
}*/
}

173
src/java/org/apache/solr/schema/IndexSchema.java
View File

@ -65,6 +65,30 @@ public final class IndexSchema {
private float version;
private final SolrResourceLoader loader;
private final HashMap<String, SchemaField> fields = new HashMap<String,SchemaField>();
private final HashMap<String, FieldType> fieldTypes = new HashMap<String,FieldType>();
private final List<SchemaField> fieldsWithDefaultValue = new ArrayList<SchemaField>();
private final Collection<SchemaField> requiredFields = new HashSet<SchemaField>();
private DynamicField[] dynamicFields;
private Analyzer analyzer;
private Analyzer queryAnalyzer;
private String defaultSearchFieldName=null;
private String queryParserDefaultOperator = "OR";
private final Map<String, List<CopyField>> copyFieldsMap = new HashMap<String, List<CopyField>>();
private DynamicCopy[] dynamicCopyFields;
/**
* keys are all fields copied to, count is num of copyField
* directives that target them.
*/
private Map<SchemaField, Integer> copyFieldTargetCounts
= new HashMap<SchemaField, Integer>();
/**
* Constructs a schema using the specified file name using the normal
* Config path directory searching rules.
@ -156,10 +180,7 @@ public final class IndexSchema {
@Deprecated
public String getName() { return name; }
private final HashMap<String, SchemaField> fields = new HashMap<String,SchemaField>();
private final HashMap<String, FieldType> fieldTypes = new HashMap<String,FieldType>();
private final List<SchemaField> fieldsWithDefaultValue = new ArrayList<SchemaField>();
private final Collection<SchemaField> requiredFields = new HashSet<SchemaField>();
;
/**
* Provides direct access to the Map containing all explicit
@ -218,7 +239,7 @@ public final class IndexSchema {
*/
public SimilarityFactory getSimilarityFactory() { return similarityFactory; }
private Analyzer analyzer;
/**
* Returns the Analyzer used when indexing documents for this index
@ -230,7 +251,7 @@ public final class IndexSchema {
*/
public Analyzer getAnalyzer() { return analyzer; }
private Analyzer queryAnalyzer;
/**
* Returns the Analyzer used when searching this index
@ -242,8 +263,7 @@ public final class IndexSchema {
*/
public Analyzer getQueryAnalyzer() { return queryAnalyzer; }
private String defaultSearchFieldName=null;
private String queryParserDefaultOperator = "OR";
/**
* A SolrQueryParser linked to this IndexSchema for field datatype
@ -399,7 +419,7 @@ public final class IndexSchema {
Config schemaConf = new Config(loader, "schema", is, "/schema/");
Document document = schemaConf.getDocument();
final XPath xpath = schemaConf.getXPath();
final List<SchemaAware> schemaAware = new ArrayList<SchemaAware>();
Node nd = (Node) xpath.evaluate("/schema/@name", document, XPathConstants.NODE);
if (nd==null) {
log.warn("schema has no name!");
@ -434,6 +454,9 @@ public final class IndexSchema {
ft.setAnalyzer(analyzer);
ft.setQueryAnalyzer(queryAnalyzer);
}
if (ft instanceof SchemaAware){
schemaAware.add((SchemaAware) ft);
}
return ft;
}
@ -494,7 +517,6 @@ public final class IndexSchema {
SolrException.logOnce(log,null,t);
SolrConfig.severeErrors.add( t );
}
log.debug("field defined: " + f);
if( f.getDefaultValue() != null ) {
log.debug(name+" contains default value: " + f.getDefaultValue());
@ -506,23 +528,7 @@ public final class IndexSchema {
}
} else if (node.getNodeName().equals("dynamicField")) {
// make sure nothing else has the same path
boolean dup = false;
for( DynamicField df : dFields ) {
if( df.regex.equals( f.name ) ) {
String msg = "[schema.xml] Duplicate DynamicField definition for '"
+ f.getName() + "' ignoring: "+f.toString();
Throwable t = new SolrException( SolrException.ErrorCode.SERVER_ERROR, msg );
SolrException.logOnce(log,null,t);
SolrConfig.severeErrors.add( t );
dup = true;
break;
}
}
if( !dup ) {
dFields.add(new DynamicField(f));
log.debug("dynamic field defined: " + f);
}
addDynamicField(dFields, f);
} else {
// we should never get here
throw new RuntimeException("Unknown field type");
@ -534,6 +540,7 @@ public final class IndexSchema {
// in DocumentBuilder.getDoc()
requiredFields.addAll(getFieldsWithDefaultValue());
// OK, now sort the dynamic fields largest to smallest size so we don't get
// any false matches. We want to act like a compiler tool and try and match
// the largest string possible.
@ -568,6 +575,9 @@ public final class IndexSchema {
}
};
}
if (similarityFactory instanceof SchemaAware){
schemaAware.add((SchemaAware) similarityFactory);
}
log.debug("using similarity factory" + similarityFactory.getClass().getName());
}
@ -652,7 +662,10 @@ public final class IndexSchema {
entry.getValue()+")");
}
}
//Run the callbacks on SchemaAware now that everything else is done
for (SchemaAware aware : schemaAware) {
aware.inform(this);
}
} catch (SolrException e) {
SolrConfig.severeErrors.add( e );
throw e;
@ -664,6 +677,56 @@ public final class IndexSchema {
// create the field analyzers
refreshAnalyzers();
}
private void addDynamicField(List<DynamicField> dFields, SchemaField f) {
boolean dup = isDuplicateDynField(dFields, f);
if( !dup ) {
addDynamicFieldNoDupCheck(dFields, f);
} else {
String msg = "[schema.xml] Duplicate DynamicField definition for '"
+ f.getName() + "' ignoring: " + f.toString();
Throwable t = new SolrException(SolrException.ErrorCode.SERVER_ERROR, msg);
SolrException.logOnce(log, null, t);
SolrConfig.severeErrors.add(t);
}
}
/**
* Register one or more new Dynamic Field with the Schema.
* @param f The {@link org.apache.solr.schema.SchemaField}
*/
public void registerDynamicField(SchemaField ... f) {
List<DynamicField> dynFields = new ArrayList<DynamicField>(Arrays.asList(dynamicFields));
for (SchemaField field : f) {
if (isDuplicateDynField(dynFields, field) == false) {
log.debug("dynamic field creation for schema field: " + field.getName());
addDynamicFieldNoDupCheck(dynFields, field);
} else {
log.debug("dynamic field creation avoided: dynamic field: [" + field.getName() + "] " +
"already defined in the schema!");
}
}
Collections.sort(dynFields);
dynamicFields = (DynamicField[]) dynFields.toArray(new DynamicField[dynFields.size()]);
}
private void addDynamicFieldNoDupCheck(List<DynamicField> dFields, SchemaField f) {
dFields.add(new DynamicField(f));
log.debug("dynamic field defined: " + f);
}
private boolean isDuplicateDynField(List<DynamicField> dFields, SchemaField f) {
boolean dup = false;
for( DynamicField df : dFields ) {
if( df.regex.equals( f.name ) ) {
dup = true;
break;
}
}
return dup;
}
public void registerCopyField( String source, String dest )
@ -987,7 +1050,7 @@ public final class IndexSchema {
}
}
private DynamicField[] dynamicFields;
public SchemaField[] getDynamicFieldPrototypes() {
SchemaField[] df = new SchemaField[dynamicFields.length];
for (int i=0;i<dynamicFields.length;i++) {
@ -1038,42 +1101,43 @@ public final class IndexSchema {
}
return false;
}
}
/**
* Returns the SchemaField that should be used for the specified field name, or
* null if none exists.
*
* @param fieldName may be an explicitly defined field, or a name that
* @param fieldName may be an explicitly defined field, a PolyField, or a name that
* matches a dynamic field.
* @see #getFieldType
* @see #getField(String)
* @return The {@link org.apache.solr.schema.SchemaField}
*/
public SchemaField getFieldOrNull(String fieldName) {
SchemaField f = fields.get(fieldName);
SchemaField f = fields.get(fieldName);
if (f != null) return f;
for (DynamicField df : dynamicFields) {
if (df.matches(fieldName)) return df.makeSchemaField(fieldName);
}
return f;
}
/**
* Returns the SchemaField that should be used for the specified field name
*
* @param fieldName may be an explicitly defined field, or a name that
* @param fieldName may be an explicitly defined field, a PolyField type, or a name that
* matches a dynamic field.
* @throws SolrException if no such field exists
* @see #getFieldType
* @see #getFieldOrNull(String)
* @return The {@link SchemaField}
*/
public SchemaField getField(String fieldName) {
SchemaField f = fields.get(fieldName);
SchemaField f = getFieldOrNull(fieldName);
if (f != null) return f;
for (DynamicField df : dynamicFields) {
if (df.matches(fieldName)) return df.makeSchemaField(fieldName);
}
// Hmmm, default field could also be implemented with a dynamic field of "*".
// It would have to be special-cased and only used if nothing else matched.
@ -1104,6 +1168,16 @@ public final class IndexSchema {
return getDynamicFieldType(fieldName);
}
/**
* Given the name of a {@link org.apache.solr.schema.FieldType} (not to be confused with {@link #getFieldType(String)} which
* takes in the name of a field), return the {@link org.apache.solr.schema.FieldType}.
* @param fieldTypeName The name of the {@link org.apache.solr.schema.FieldType}
* @return The {@link org.apache.solr.schema.FieldType} or null.
*/
public FieldType getFieldTypeByName(String fieldTypeName){
return fieldTypes.get(fieldTypeName);
}
/**
* Returns the FieldType for the specified field name.
*
@ -1149,15 +1223,22 @@ public final class IndexSchema {
return null;
};
private final Map<String, List<CopyField>> copyFieldsMap = new HashMap<String, List<CopyField>>();
private DynamicCopy[] dynamicCopyFields;
/**
* keys are all fields copied to, count is num of copyField
* directives that target them.
*
* @param fieldName The name of the field
* @return the {@link FieldType} or a {@link org.apache.solr.common.SolrException} if the field is not a poly field.
*/
private Map<SchemaField, Integer> copyFieldTargetCounts
= new HashMap<SchemaField, Integer>();
public FieldType getPolyFieldType(String fieldName){
SchemaField f = fields.get(fieldName);
if (f != null && f.isPolyField()) return f.getType();
throw new SolrException( SolrException.ErrorCode.BAD_REQUEST,"undefined field or not a poly field "+fieldName);
}
public FieldType getPolyFieldTypeNoEx(String fieldName){
SchemaField f = fields.get(fieldName);
if (f != null && f.isPolyField()) return f.getType();
return null;
}
/**
* Get all copy fields, both the static and the dynamic ones.

276
src/java/org/apache/solr/schema/PointType.java Normal file
View File

@ -0,0 +1,276 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF 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.apache.solr.schema;
import org.apache.lucene.document.Field;
import org.apache.lucene.document.Fieldable;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.search.BooleanClause;
import org.apache.lucene.search.BooleanQuery;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.Searcher;
import org.apache.lucene.search.SortField;
import org.apache.solr.common.SolrException;
import org.apache.solr.common.params.MapSolrParams;
import org.apache.solr.common.params.SolrParams;
import org.apache.solr.request.TextResponseWriter;
import org.apache.solr.request.XMLWriter;
import org.apache.solr.search.MultiValueSource;
import org.apache.solr.search.QParser;
import org.apache.solr.search.function.DocValues;
import org.apache.solr.search.function.ValueSource;
import org.apache.solr.search.function.distance.DistanceUtils;
import java.io.IOException;
import java.util.Map;
/**
* A point type that indexes a point in an n-dimensional space as separate fields and uses
* range queries for bounding box calculations.
* <p/>
* <p/>
* NOTE: There can only be one sub type
*/
public class PointType extends CoordinateFieldType {
/**
* 2 dimensional by default
*/
public static final int DEFAULT_DIMENSION = 2;
public static final String DIMENSION = "dimension";
protected IndexSchema schema; // needed for retrieving SchemaFields
@Override
protected void init(IndexSchema schema, Map<String, String> args) {
SolrParams p = new MapSolrParams(args);
dimension = p.getInt(DIMENSION, DEFAULT_DIMENSION);
if (dimension < 1) {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST,
"The dimension must be > 0: " + dimension);
}
args.remove(DIMENSION);
this.schema = schema;
super.init(schema, args);
}
@Override
public boolean isPolyField() {
return true; // really only true if the field is indexed
}
@Override
public Fieldable[] createFields(SchemaField field, String externalVal, float boost) {
String[] point = DistanceUtils.parsePoint(null, externalVal, dimension);
return createFields(field, dynFieldProps, subType, externalVal, boost, point);
}
@Override
public ValueSource getValueSource(SchemaField field, QParser parser) {
return new PointTypeValueSource(field, dimension, subType, parser);
}
//It never makes sense to create a single field, so make it impossible to happen
@Override
public Field createField(SchemaField field, String externalVal, float boost) {
throw new UnsupportedOperationException("PointType uses multiple fields. field=" + field.getName());
}
@Override
public void write(XMLWriter xmlWriter, String name, Fieldable f) throws IOException {
xmlWriter.writeStr(name, f.stringValue());
}
@Override
public void write(TextResponseWriter writer, String name, Fieldable f) throws IOException {
writer.writeStr(name, f.stringValue(), false);
}
@Override
public SortField getSortField(SchemaField field, boolean top) {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, "Sorting not suported on DualPointType " + field.getName());
}
@Override
/**
* Care should be taken in calling this with higher order dimensions for performance reasons.
*/
public Query getRangeQuery(QParser parser, SchemaField field, String part1, String part2, boolean minInclusive, boolean maxInclusive) {
//Query could look like: [x1,y1 TO x2,y2] for 2 dimension, but could look like: [x1,y1,z1 TO x2,y2,z2], and can be extrapolated to n-dimensions
//thus, this query essentially creates a box, cube, etc.
String[] p1 = DistanceUtils.parsePoint(null, part1, dimension);
String[] p2 = DistanceUtils.parsePoint(null, part2, dimension);
BooleanQuery result = new BooleanQuery(true);
String name = field.getName() + "_";
String suffix = POLY_FIELD_SEPARATOR + subType.typeName;
int len = name.length();
StringBuilder bldr = new StringBuilder(len + 3 + suffix.length());//should be enough buffer to handle most values of j.
bldr.append(name);
for (int i = 0; i < dimension; i++) {
bldr.append(i).append(suffix);
SchemaField subSF = schema.getField(bldr.toString());
// points must currently be ordered... should we support specifying any two opposite corner points?
/*new TermRangeQuery(
field.getName() + i + POLY_FIELD_SEPARATOR + subType.typeName,
subType.toInternal(p1[i]),
subType.toInternal(p2[i]),
minInclusive, maxInclusive);*/
result.add(subType.getRangeQuery(parser, subSF, p1[i], p2[i], minInclusive, maxInclusive), BooleanClause.Occur.MUST);
bldr.setLength(len);
}
return result;
}
@Override
public Query getFieldQuery(QParser parser, SchemaField field, String externalVal) {
Query result = null;
String[] p1 = DistanceUtils.parsePoint(null, externalVal, dimension);
//TODO: should we assert that p1.length == dimension?
BooleanQuery bq = new BooleanQuery(true);
String name = field.getName() + "_";
String suffix = POLY_FIELD_SEPARATOR + subType.typeName;
int len = name.length();
StringBuilder bldr = new StringBuilder(len + 3 + suffix.length());//should be enough buffer to handle most values of j.
bldr.append(name);
for (int i = 0; i < dimension; i++) {
bldr.append(i).append(suffix);
SchemaField sf1 = schema.getField(bldr.toString());
Query tq = subType.getFieldQuery(parser, sf1, p1[i]);
//new TermQuery(new Term(bldr.toString(), subType.toInternal(p1[i])));
bq.add(tq, BooleanClause.Occur.MUST);
bldr.setLength(len);
}
result = bq;
return result;
}
class PointTypeValueSource extends MultiValueSource {
protected SchemaField field;
protected FieldType subType;
protected int dimension;
private QParser parser;
public PointTypeValueSource(SchemaField field, int dimension, FieldType subType, QParser parser) {
this.field = field;
this.dimension = dimension;
this.subType = subType;
this.parser = parser;
}
@Override
public void createWeight(Map context, Searcher searcher) throws IOException {
String name = field.getName();
String suffix = FieldType.POLY_FIELD_SEPARATOR + subType.typeName;
int len = name.length();
StringBuilder bldr = new StringBuilder(len + 3 + suffix.length());//should be enough buffer to handle most values of j.
bldr.append(name);
for (int i = 0; i < dimension; i++) {
bldr.append(i).append(suffix);
SchemaField sf = schema.getField(bldr.toString());
subType.getValueSource(sf, parser).createWeight(context, searcher);
bldr.setLength(len);
}
}
public int dimension() {
return dimension;
}
@Override
public DocValues getValues(Map context, IndexReader reader) throws IOException {
final DocValues[] valsArr1 = new DocValues[dimension];
String name = field.getName();
String suffix = FieldType.POLY_FIELD_SEPARATOR + subType.typeName;
int len = name.length();
StringBuilder bldr = new StringBuilder(len + 3 + suffix.length());//should be enough buffer to handle most values of j.
bldr.append(name);
for (int i = 0; i < dimension; i++) {
bldr.append(i).append(suffix);
SchemaField sf = schema.getField(bldr.toString());
valsArr1[i] = subType.getValueSource(sf, parser).getValues(context, reader);
bldr.setLength(len);
}
return new DocValues() {
//TODO: not sure how to handle the other types at this moment
@Override
public void doubleVal(int doc, double[] vals) {
//TODO: check whether vals.length == dimension or assume its handled elsewhere?
for (int i = 0; i < dimension; i++) {
vals[i] = valsArr1[i].doubleVal(doc);
}
}
@Override
public String toString(int doc) {
StringBuilder sb = new StringBuilder("point(");
boolean firstTime = true;
for (DocValues docValues : valsArr1) {
if (firstTime == false) {
sb.append(",");
} else {
firstTime = true;
}
sb.append(docValues.toString(doc));
}
sb.append(")");
return sb.toString();
}
};
}
public String description() {
StringBuilder sb = new StringBuilder();
sb.append("point(");
sb.append("fld=").append(field.name).append(", subType=").append(subType.typeName)
.append(", dimension=").append(dimension).append(')');
return sb.toString();
}
@Override
public boolean equals(Object o) {
if (this == o) return true;
if (!(o instanceof PointTypeValueSource)) return false;
PointTypeValueSource that = (PointTypeValueSource) o;
if (dimension != that.dimension) return false;
if (!field.equals(that.field)) return false;
if (!subType.equals(that.subType)) return false;
return true;
}
@Override
public int hashCode() {
int result = field.hashCode();
result = 31 * result + subType.hashCode();
result = 31 * result + dimension;
return result;
}
}
}

39
src/java/org/apache/solr/schema/SchemaAware.java Normal file
View File

@ -0,0 +1,39 @@
package org.apache.solr.schema;
/**
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF 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.
*/
/**
* An interface that can be extended to provide a callback mechanism for
* informing an {@link IndexSchema} instance of changes to it, dynamically
* performed at runtime.
*
* @since SOLR-1131
*
**/
public interface SchemaAware {
/**
* Informs the {@link IndexSchema} provided by the <code>schema</code>
* parameter of an event (e.g., a new {@link FieldType} was added, etc.
*
* @param schema
* The {@link IndexSchema} instance that inform of the update to.
*
* @since SOLR-1131
*/
public void inform(IndexSchema schema);
}

44
src/java/org/apache/solr/schema/SchemaField.java
View File

@ -90,9 +90,23 @@ public final class SchemaField extends FieldProperties {
boolean isTokenized() { return (properties & TOKENIZED)!=0; }
boolean isBinary() { return (properties & BINARY)!=0; }
public Field createField(String val, float boost) {
return type.createField(this,val,boost);
}
public Fieldable[] createFields(String val, float boost) {
return type.createFields(this,val,boost);
}
/**
* If true, then use {@link #createFields(String, float)}, else use {@link #createField} to save an extra allocation
* @return true if this field is a poly field
*/
public boolean isPolyField(){
return type.isPolyField();
}
@Override
public String toString() {
@ -119,6 +133,29 @@ public final class SchemaField extends FieldProperties {
static SchemaField create(String name, FieldType ft, Map<String,String> props) {
String defaultValue = null;
if( props.containsKey( "default" ) ) {
defaultValue = (String)props.get( "default" );
}
return new SchemaField(name, ft, calcProps(name, ft, props), defaultValue );
}
/**
* Create a SchemaField w/ the props specified. Does not support a default value.
* @param name The name of the SchemaField
* @param ft The {@link org.apache.solr.schema.FieldType} of the field
* @param props The props. See {@link #calcProps(String, org.apache.solr.schema.FieldType, java.util.Map)}
* @param defValue The default Value for the field
* @return The SchemaField
*
* @see #create(String, FieldType, java.util.Map)
*/
static SchemaField create(String name, FieldType ft, int props, String defValue){
return new SchemaField(name, ft, props, defValue);
}
static int calcProps(String name, FieldType ft, Map<String, String> props) {
int trueProps = parseProperties(props,true);
int falseProps = parseProperties(props,false);
@ -166,12 +203,7 @@ public final class SchemaField extends FieldProperties {
p &= ~falseProps;
p |= trueProps;
String defaultValue = null;
if( props.containsKey( "default" ) ) {
defaultValue = (String)props.get( "default" );
}
return new SchemaField(name, ft, p, defaultValue );
return p;
}
public String getDefaultValue() {

202
src/java/org/apache/solr/schema/TextField.java
View File

@ -18,12 +18,28 @@
package org.apache.solr.schema;
import org.apache.lucene.search.SortField;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.PhraseQuery;
import org.apache.lucene.search.TermQuery;
import org.apache.lucene.search.BooleanQuery;
import org.apache.lucene.search.BooleanClause;
import org.apache.lucene.search.MultiPhraseQuery;
import org.apache.lucene.document.Fieldable;
import org.apache.lucene.index.Term;
import org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute;
import org.apache.lucene.analysis.tokenattributes.TermAttribute;
import org.apache.lucene.analysis.CachingTokenFilter;
import org.apache.lucene.analysis.TokenStream;
import org.apache.lucene.analysis.Analyzer;
import org.apache.solr.request.XMLWriter;
import org.apache.solr.request.TextResponseWriter;
import org.apache.solr.search.QParser;
import java.util.Map;
import java.util.List;
import java.util.ArrayList;
import java.io.IOException;
import java.io.StringReader;
/** <code>TextField</code> is the basic type for configurable text analysis.
* Analyzers for field types using this implementation should be defined in the schema.
@ -48,4 +64,190 @@ public class TextField extends CompressableField {
public void write(TextResponseWriter writer, String name, Fieldable f) throws IOException {
writer.writeStr(name, f.stringValue(), true);
}
@Override
public Query getFieldQuery(QParser parser, SchemaField field, String externalVal) {
return parseFieldQuery(parser, getQueryAnalyzer(), field.getName(), externalVal);
}
static Query parseFieldQuery(QParser parser, Analyzer analyzer, String field, String queryText) {
int phraseSlop = 0;
boolean enablePositionIncrements = true;
// most of the following code is taken from the Lucene QueryParser
// Use the analyzer to get all the tokens, and then build a TermQuery,
// PhraseQuery, or nothing based on the term count
TokenStream source;
try {
source = analyzer.reusableTokenStream(field, new StringReader(queryText));
source.reset();
} catch (IOException e) {
source = analyzer.tokenStream(field, new StringReader(queryText));
}
CachingTokenFilter buffer = new CachingTokenFilter(source);
TermAttribute termAtt = null;
PositionIncrementAttribute posIncrAtt = null;
int numTokens = 0;
boolean success = false;
try {
buffer.reset();
success = true;
} catch (IOException e) {
// success==false if we hit an exception
}
if (success) {
if (buffer.hasAttribute(TermAttribute.class)) {
termAtt = (TermAttribute) buffer.getAttribute(TermAttribute.class);
}
if (buffer.hasAttribute(PositionIncrementAttribute.class)) {
posIncrAtt = (PositionIncrementAttribute) buffer.getAttribute(PositionIncrementAttribute.class);
}
}
int positionCount = 0;
boolean severalTokensAtSamePosition = false;
boolean hasMoreTokens = false;
if (termAtt != null) {
try {
hasMoreTokens = buffer.incrementToken();
while (hasMoreTokens) {
numTokens++;
int positionIncrement = (posIncrAtt != null) ? posIncrAtt.getPositionIncrement() : 1;
if (positionIncrement != 0) {
positionCount += positionIncrement;
} else {
severalTokensAtSamePosition = true;
}
hasMoreTokens = buffer.incrementToken();
}
} catch (IOException e) {
// ignore
}
}
try {
// rewind the buffer stream
buffer.reset();
// close original stream - all tokens buffered
source.close();
}
catch (IOException e) {
// ignore
}
if (numTokens == 0)
return null;
else if (numTokens == 1) {
String term = null;
try {
boolean hasNext = buffer.incrementToken();
assert hasNext == true;
term = termAtt.term();
} catch (IOException e) {
// safe to ignore, because we know the number of tokens
}
// return newTermQuery(new Term(field, term));
return new TermQuery(new Term(field, term));
} else {
if (severalTokensAtSamePosition) {
if (positionCount == 1) {
// no phrase query:
// BooleanQuery q = newBooleanQuery(true);
BooleanQuery q = new BooleanQuery(true);
for (int i = 0; i < numTokens; i++) {
String term = null;
try {
boolean hasNext = buffer.incrementToken();
assert hasNext == true;
term = termAtt.term();
} catch (IOException e) {
// safe to ignore, because we know the number of tokens
}
// Query currentQuery = newTermQuery(new Term(field, term));
Query currentQuery = new TermQuery(new Term(field, term));
q.add(currentQuery, BooleanClause.Occur.SHOULD);
}
return q;
}
else {
// phrase query:
// MultiPhraseQuery mpq = newMultiPhraseQuery();
MultiPhraseQuery mpq = new MultiPhraseQuery();
mpq.setSlop(phraseSlop);
List multiTerms = new ArrayList();
int position = -1;
for (int i = 0; i < numTokens; i++) {
String term = null;
int positionIncrement = 1;
try {
boolean hasNext = buffer.incrementToken();
assert hasNext == true;
term = termAtt.term();
if (posIncrAtt != null) {
positionIncrement = posIncrAtt.getPositionIncrement();
}
} catch (IOException e) {
// safe to ignore, because we know the number of tokens
}
if (positionIncrement > 0 && multiTerms.size() > 0) {
if (enablePositionIncrements) {
mpq.add((Term[])multiTerms.toArray(new Term[0]),position);
} else {
mpq.add((Term[])multiTerms.toArray(new Term[0]));
}
multiTerms.clear();
}
position += positionIncrement;
multiTerms.add(new Term(field, term));
}
if (enablePositionIncrements) {
mpq.add((Term[])multiTerms.toArray(new Term[0]),position);
} else {
mpq.add((Term[])multiTerms.toArray(new Term[0]));
}
return mpq;
}
}
else {
// PhraseQuery pq = newPhraseQuery();
PhraseQuery pq = new PhraseQuery();
pq.setSlop(phraseSlop);
int position = -1;
for (int i = 0; i < numTokens; i++) {
String term = null;
int positionIncrement = 1;
try {
boolean hasNext = buffer.incrementToken();
assert hasNext == true;
term = termAtt.term();
if (posIncrAtt != null) {
positionIncrement = posIncrAtt.getPositionIncrement();
}
} catch (IOException e) {
// safe to ignore, because we know the number of tokens
}
if (enablePositionIncrements) {
position += positionIncrement;
pq.add(new Term(field, term),position);
} else {
pq.add(new Term(field, term));
}
}
return pq;
}
}
}
}

97
src/java/org/apache/solr/search/FieldQParserPlugin.java
View File

@ -28,6 +28,7 @@ import org.apache.solr.common.SolrException;
import org.apache.solr.request.SolrQueryRequest;
import org.apache.solr.schema.FieldType;
import org.apache.solr.schema.TextField;
import org.apache.solr.schema.SchemaField;
import java.io.IOException;
import java.io.StringReader;
@ -52,99 +53,9 @@ public class FieldQParserPlugin extends QParserPlugin {
public Query parse() throws ParseException {
String field = localParams.get(QueryParsing.F);
String queryText = localParams.get(QueryParsing.V);
FieldType ft = req.getSchema().getFieldType(field);
if (!(ft instanceof TextField)) {
String internal = ft.toInternal(queryText);
return new TermQuery(new Term(field, internal));
}
int phraseSlop = 0;
Analyzer analyzer = req.getSchema().getQueryAnalyzer();
// most of the following code is taken from the Lucene QueryParser
// Use the analyzer to get all the tokens, and then build a TermQuery,
// PhraseQuery, or nothing based on the term count
TokenStream source = null;
try {
source = analyzer.reusableTokenStream(field, new StringReader(queryText));
source.reset();
} catch (IOException e) {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, e);
}
ArrayList<Token> lst = new ArrayList<Token>();
Token t;
int positionCount = 0;
boolean severalTokensAtSamePosition = false;
while (true) {
try {
t = source.next();
}
catch (IOException e) {
t = null;
}
if (t == null)
break;
lst.add(t);
if (t.getPositionIncrement() != 0)
positionCount += t.getPositionIncrement();
else
severalTokensAtSamePosition = true;
}
try {
source.close();
}
catch (IOException e) {
// ignore
}
if (lst.size() == 0)
return null;
else if (lst.size() == 1) {
t = lst.get(0);
return new TermQuery(new Term(field, new String(t.termBuffer(), 0, t.termLength())));
} else {
if (severalTokensAtSamePosition) {
if (positionCount == 1) {
// no phrase query:
BooleanQuery q = new BooleanQuery(true);
for (int i = 0; i < lst.size(); i++) {
t = (org.apache.lucene.analysis.Token) lst.get(i);
TermQuery currentQuery = new TermQuery(
new Term(field, new String(t.termBuffer(), 0, t.termLength())));
q.add(currentQuery, BooleanClause.Occur.SHOULD);
}
return q;
}
else {
// phrase query:
MultiPhraseQuery mpq = new MultiPhraseQuery();
mpq.setSlop(phraseSlop);
ArrayList multiTerms = new ArrayList();
for (int i = 0; i < lst.size(); i++) {
t = (org.apache.lucene.analysis.Token) lst.get(i);
if (t.getPositionIncrement() == 1 && multiTerms.size() > 0) {
mpq.add((Term[])multiTerms.toArray(new Term[0]));
multiTerms.clear();
}
multiTerms.add(new Term(field, new String(t.termBuffer(), 0, t.termLength())));
}
mpq.add((Term[])multiTerms.toArray(new Term[0]));
return mpq;
}
}
else {
PhraseQuery q = new PhraseQuery();
q.setSlop(phraseSlop);
for (int i = 0; i < lst.size(); i++) {
Token token = lst.get(i);
q.add(new Term(field, new String(token.termBuffer(), 0, token.termLength())));
}
return q;
}
}
SchemaField sf = req.getSchema().getField(field);
FieldType ft = sf.getType();
return ft.getFieldQuery(this, sf, queryText);
}
};
}

29
src/java/org/apache/solr/search/MultiValueSource.java Normal file
View File

@ -0,0 +1,29 @@
package org.apache.solr.search;
/**
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF 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.
*/
import org.apache.solr.search.function.ValueSource;
/**
* A {@link ValueSource} that abstractly represents {@link ValueSource}s for
* poly fields, and other things.
**/
public abstract class MultiValueSource extends ValueSource {
public abstract int dimension();
}

7
src/java/org/apache/solr/search/SolrQueryParser.java
View File

@ -34,6 +34,7 @@ import org.apache.solr.schema.IndexSchema;
import org.apache.solr.schema.SchemaField;
import org.apache.solr.schema.TrieField;
import org.apache.solr.schema.SchemaField;
import org.apache.solr.schema.TextField;
// TODO: implement the analysis of simple fields with
// FieldType.toInternal() instead of going through the
@ -145,6 +146,12 @@ public class SolrQueryParser extends QueryParser {
return parser.subQuery(queryText, null).getQuery();
}
}
//Intercept poly fields, as they get expanded by default to an OR clause of
SchemaField sf = schema.getField(field);
//TODO: is there anyway to avoid this instance of check?
if (sf != null&& !(sf.getType() instanceof TextField)){//we have a poly field, deal with it specially by delegating to the FieldType
return sf.getType().getFieldQuery(parser, sf, queryText);
}
// default to a normal field query
return super.getFieldQuery(field, queryText);

164
src/java/org/apache/solr/search/ToMultiValueSource.java Normal file
View File

@ -0,0 +1,164 @@
package org.apache.solr.search;
/**
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF 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.
*/
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.search.Searcher;
import org.apache.solr.search.function.DocValues;
import org.apache.solr.search.function.ValueSource;
import java.io.IOException;
import java.util.List;
import java.util.Map;
/**
* Converts individual ValueSource instances to leverage the DocValues *Val functions that work with multiple values,
* i.e. {@link org.apache.solr.search.function.DocValues#doubleVal(int, double[])}
*/
//Not crazy about the name, but...
public class ToMultiValueSource extends MultiValueSource {
protected List<ValueSource> sources;
public ToMultiValueSource(List<ValueSource> sources) {
this.sources = sources;
}
public List<ValueSource> getSources() {
return sources;
}
public int dimension() {
return sources.size();
}
@Override
public DocValues getValues(Map context, IndexReader reader) throws IOException {
int size = sources.size();
final DocValues[] valsArr = new DocValues[size];
for (int i = 0; i < size; i++) {
valsArr[i] = sources.get(i).getValues(context, reader);
}
return new DocValues() {
@Override
public void byteVal(int doc, byte[] vals) {
for (int i = 0; i < valsArr.length; i++) {
vals[i] = valsArr[i].byteVal(doc);
}
}
@Override
public void shortVal(int doc, short[] vals) {
for (int i = 0; i < valsArr.length; i++) {
vals[i] = valsArr[i].shortVal(doc);
}
}
@Override
public void floatVal(int doc, float[] vals) {
for (int i = 0; i < valsArr.length; i++) {
vals[i] = valsArr[i].floatVal(doc);
}
}
@Override
public void intVal(int doc, int[] vals) {
for (int i = 0; i < valsArr.length; i++) {
vals[i] = valsArr[i].intVal(doc);
}
}
@Override
public void longVal(int doc, long[] vals) {
for (int i = 0; i < valsArr.length; i++) {
vals[i] = valsArr[i].longVal(doc);
}
}
@Override
public void doubleVal(int doc, double[] vals) {
for (int i = 0; i < valsArr.length; i++) {
vals[i] = valsArr[i].doubleVal(doc);
}
}
@Override
public void strVal(int doc, String[] vals) {
for (int i = 0; i < valsArr.length; i++) {
vals[i] = valsArr[i].strVal(doc);
}
}
@Override
public String toString(int doc) {
StringBuilder sb = new StringBuilder();
sb.append("toMultiVS(");
boolean firstTime = true;
for (DocValues vals : valsArr) {
if (firstTime) {
firstTime = false;
} else {
sb.append(',');
}
sb.append(vals.toString(doc));
}
sb.append(')');
return sb.toString();
}
};
}
public void createWeight(Map context, Searcher searcher) throws IOException {
for (ValueSource source : sources)
source.createWeight(context, searcher);
}
public String description() {
StringBuilder sb = new StringBuilder();
sb.append("toMultiVS(");
boolean firstTime = true;
for (ValueSource source : sources) {
if (firstTime) {
firstTime = false;
} else {
sb.append(',');
}
sb.append(source);
}
sb.append(")");
return sb.toString();
}
@Override
public boolean equals(Object o) {
if (this == o) return true;
if (!(o instanceof ToMultiValueSource)) return false;
ToMultiValueSource that = (ToMultiValueSource) o;
if (!sources.equals(that.sources)) return false;
return true;
}
@Override
public int hashCode() {
return sources.hashCode();
}
}

103
src/java/org/apache/solr/search/ValueSourceParser.java
View File

@ -41,6 +41,7 @@ import java.util.Date;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Collections;
/**
* A factory that parses user queries to generate ValueSource instances.
@ -202,6 +203,11 @@ public abstract class ValueSourceParser implements NamedListInitializedPlugin {
};
}
});
addParser("toMultiVS", new ValueSourceParser(){
public ValueSource parse(FunctionQParser fp) throws ParseException{
return new ToMultiValueSource(fp.parseValueSourceList());
}
});
addParser("query", new ValueSourceParser() {
// boost(query($q),rating)
public ValueSource parse(FunctionQParser fp) throws ParseException {
@ -224,22 +230,47 @@ public abstract class ValueSourceParser implements NamedListInitializedPlugin {
addParser("hsin", new ValueSourceParser() {
public ValueSource parse(FunctionQParser fp) throws ParseException {
ValueSource x1 = fp.parseValueSource();
ValueSource y1 = fp.parseValueSource();
ValueSource x2 = fp.parseValueSource();
ValueSource y2 = fp.parseValueSource();
double radius = fp.parseDouble();
MultiValueSource pv1;
MultiValueSource pv2;
return new HaversineFunction(x1, y1, x2, y2, radius);
ValueSource one = fp.parseValueSource();
ValueSource two = fp.parseValueSource();
if (fp.hasMoreArguments()) {
List<ValueSource> s1 = new ArrayList<ValueSource>();
s1.add(one);
s1.add(two);
pv1 = new ToMultiValueSource(s1);
ValueSource x2 = fp.parseValueSource();
ValueSource y2 = fp.parseValueSource();
List<ValueSource> s2 = new ArrayList<ValueSource>();
s2.add(x2);
s2.add(y2);
pv2 = new ToMultiValueSource(s2);
} else {
//check to see if we have multiValue source
if (one instanceof MultiValueSource && two instanceof MultiValueSource){
pv1 = (MultiValueSource) one;
pv2 = (MultiValueSource) two;
} else {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST,
"Input must either be 2 MultiValueSources, or there must be 4 ValueSources");
}
}
boolean convert = false;
if (fp.hasMoreArguments()){
convert = Boolean.parseBoolean(fp.parseArg());
}
return new HaversineFunction(pv1, pv2, radius, convert);
}
});
addParser("ghhsin", new ValueSourceParser() {
public ValueSource parse(FunctionQParser fp) throws ParseException {
double radius = fp.parseDouble();
ValueSource gh1 = fp.parseValueSource();
ValueSource gh2 = fp.parseValueSource();
double radius = fp.parseDouble();
return new GeohashHaversineFunction(gh1, gh2, radius);
}
@ -393,15 +424,9 @@ public abstract class ValueSourceParser implements NamedListInitializedPlugin {
addParser("sqedist", new ValueSourceParser() {
public ValueSource parse(FunctionQParser fp) throws ParseException {
List<ValueSource> sources = fp.parseValueSourceList();
if (sources.size() % 2 != 0) {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, "Illegal number of sources. There must be an even number of sources");
}
int dim = sources.size() / 2;
List<ValueSource> sources1 = new ArrayList<ValueSource>(dim);
List<ValueSource> sources2 = new ArrayList<ValueSource>(dim);
//Get dim value sources for the first vector
splitSources(dim, sources, sources1, sources2);
return new SquaredEuclideanFunction(sources1, sources2);
MVResult mvr = getMultiValueSources(sources);
return new SquaredEuclideanFunction(mvr.mv1, mvr.mv2);
}
});
@ -409,14 +434,8 @@ public abstract class ValueSourceParser implements NamedListInitializedPlugin {
public ValueSource parse(FunctionQParser fp) throws ParseException {
float power = fp.parseFloat();
List<ValueSource> sources = fp.parseValueSourceList();
if (sources.size() % 2 != 0) {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, "Illegal number of sources. There must be an even number of sources");
}
int dim = sources.size() / 2;
List<ValueSource> sources1 = new ArrayList<ValueSource>(dim);
List<ValueSource> sources2 = new ArrayList<ValueSource>(dim);
splitSources(dim, sources, sources1, sources2);
return new VectorDistanceFunction(power, sources1, sources2);
MVResult mvr = getMultiValueSources(sources);
return new VectorDistanceFunction(power, mvr.mv1, mvr.mv2);
}
});
addParser("ms", new DateValueSourceParser());
@ -445,6 +464,44 @@ public abstract class ValueSourceParser implements NamedListInitializedPlugin {
}
}
private static MVResult getMultiValueSources(List<ValueSource> sources) {
MVResult mvr = new MVResult();
if (sources.size() % 2 != 0) {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, "Illegal number of sources. There must be an even number of sources");
}
if (sources.size() == 2) {
//check to see if these are MultiValueSource
boolean s1MV = sources.get(0) instanceof MultiValueSource;
boolean s2MV = sources.get(1) instanceof MultiValueSource;
if (s1MV && s2MV) {
mvr.mv1 = (MultiValueSource) sources.get(0);
mvr.mv2 = (MultiValueSource) sources.get(1);
} else if (s1MV ||
s2MV) {
//if one is a MultiValueSource, than the other one needs to be too.
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, "Illegal number of sources. There must be an even number of sources");
} else {
mvr.mv1 = new ToMultiValueSource(Collections.singletonList(sources.get(0)));
mvr.mv2 = new ToMultiValueSource(Collections.singletonList(sources.get(1)));
}
} else {
int dim = sources.size() / 2;
List<ValueSource> sources1 = new ArrayList<ValueSource>(dim);
List<ValueSource> sources2 = new ArrayList<ValueSource>(dim);
//Get dim value sources for the first vector
splitSources(dim, sources, sources1, sources2);
mvr.mv1 = new ToMultiValueSource(sources1);
mvr.mv2 = new ToMultiValueSource(sources2);
}
return mvr;
}
private static class MVResult {
MultiValueSource mv1;
MultiValueSource mv2;
}
}

9
src/java/org/apache/solr/search/function/DocValues.java
View File

@ -48,6 +48,15 @@ public abstract class DocValues {
public String strVal(int doc) { throw new UnsupportedOperationException(); }
public abstract String toString(int doc);
//For Functions that can work with multiple values from the same document. This does not apply to all functions
public void byteVal(int doc, byte [] vals) { throw new UnsupportedOperationException(); }
public void shortVal(int doc, short [] vals) { throw new UnsupportedOperationException(); }
public void floatVal(int doc, float [] vals) { throw new UnsupportedOperationException(); }
public void intVal(int doc, int [] vals) { throw new UnsupportedOperationException(); }
public void longVal(int doc, long [] vals) { throw new UnsupportedOperationException(); }
public void doubleVal(int doc, double [] vals) { throw new UnsupportedOperationException(); }
public void strVal(int doc, String [] vals) { throw new UnsupportedOperationException(); }
public Explanation explain(int doc) {
return new Explanation(floatVal(doc), toString(doc));

43
src/java/org/apache/solr/search/function/distance/DistanceUtils.java
View File

@ -1,4 +1,6 @@
package org.apache.solr.search.function.distance;
import org.apache.solr.common.SolrException;
/**
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
@ -50,5 +52,44 @@ public class DistanceUtils {
return result;
}
/**
* Given a string containing <i>dimension</i> values encoded in it, separated by commas, return a String array of length <i>dimension</i>
* containing the values.
* @param out A preallocated array. Must be size dimension. If it is not it will be resized.
* @param externalVal The value to parse
* @param dimension The expected number of values for the point
* @return An array of the values that make up the point (aka vector)
*
* @throws {@link SolrException} if the dimension specified does not match the number of values in the externalValue.
*/
public static String[] parsePoint(String[] out, String externalVal, int dimension) {
//TODO: Should we support sparse vectors?
if (out==null || out.length != dimension) out=new String[dimension];
int idx = externalVal.indexOf(',');
int end = idx;
int start = 0;
int i = 0;
if (idx == -1 && dimension == 1 && externalVal.length() > 0){//we have a single point, dimension better be 1
out[0] = externalVal.trim();
i = 1;
}
else if (idx > 0) {//if it is zero, that is an error
//Parse out a comma separated list of point values, as in: 73.5,89.2,7773.4
for (; i < dimension; i++){
while (start<end && externalVal.charAt(start)==' ') start++;
while (end>start && externalVal.charAt(end-1)==' ') end--;
out[i] = externalVal.substring(start, end);
start = idx+1;
end = externalVal.indexOf(',', start);
if (end == -1){
end = externalVal.length();
}
}
}
if (i != dimension){
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, "incompatible dimension (" + dimension +
") and values (" + externalVal + "). Only " + i + " values specified");
}
return out;
}
}

93
src/java/org/apache/solr/search/function/distance/HaversineFunction.java
View File

@ -18,6 +18,8 @@ package org.apache.solr.search.function.distance;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.search.Searcher;
import org.apache.solr.common.SolrException;
import org.apache.solr.search.MultiValueSource;
import org.apache.solr.search.function.DocValues;
import org.apache.solr.search.function.ValueSource;
@ -29,28 +31,31 @@ import java.util.Map;
* Calculate the Haversine formula (distance) between any two points on a sphere
* Takes in four value sources: (latA, lonA); (latB, lonB).
* <p/>
* Assumes the value sources are in radians
* Assumes the value sources are in radians unless
* <p/>
* See http://en.wikipedia.org/wiki/Great-circle_distance and
* http://en.wikipedia.org/wiki/Haversine_formula for the actual formula and
* also http://www.movable-type.co.uk/scripts/latlong.html
*
* @see org.apache.solr.search.function.RadianFunction
*/
public class HaversineFunction extends ValueSource {
private ValueSource x1;
private ValueSource y1;
private ValueSource x2;
private ValueSource y2;
private MultiValueSource p1;
private MultiValueSource p2;
private boolean convertToRadians = false;
private double radius;
public HaversineFunction(ValueSource x1, ValueSource y1, ValueSource x2, ValueSource y2, double radius) {
this.x1 = x1;
this.y1 = y1;
this.x2 = x2;
this.y2 = y2;
public HaversineFunction(MultiValueSource p1, MultiValueSource p2, double radius) {
this(p1, p2, radius, false);
}
public HaversineFunction(MultiValueSource p1, MultiValueSource p2, double radius, boolean convertToRads){
this.p1 = p1;
this.p2 = p2;
if (p1.dimension() != 2 || p2.dimension() != 2) {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, "Illegal dimension for value sources");
}
this.radius = radius;
this.convertToRadians = convertToRads;
}
protected String name() {
@ -59,28 +64,40 @@ public class HaversineFunction extends ValueSource {
/**
* @param doc The doc to score
* @param x1DV
* @param y1DV
* @param x2DV
* @param y2DV
* @param p1DV
* @param p2DV
* @return The haversine distance formula
*/
protected double distance(int doc, DocValues x1DV, DocValues y1DV, DocValues x2DV, DocValues y2DV) {
double x1 = x1DV.doubleVal(doc); //in radians
double y1 = y1DV.doubleVal(doc);
double x2 = x2DV.doubleVal(doc);
double y2 = y2DV.doubleVal(doc);
protected double distance(int doc, DocValues p1DV, DocValues p2DV) {
double[] p1D = new double[2];
double[] p2D = new double[2];
p1DV.doubleVal(doc, p1D);
p2DV.doubleVal(doc, p2D);
double x1;
double y1;
double x2;
double y2;
if (convertToRadians) {
x1 = p1D[0] * DistanceUtils.DEGREES_TO_RADIANS;
y1 = p1D[1] * DistanceUtils.DEGREES_TO_RADIANS;
x2 = p2D[0] * DistanceUtils.DEGREES_TO_RADIANS;
y2 = p2D[1] * DistanceUtils.DEGREES_TO_RADIANS;
} else {
x1 = p1D[0];
y1 = p1D[1];
x2 = p2D[0];
y2 = p2D[1];
}
return DistanceUtils.haversine(x1, y1, x2, y2, radius);
}
@Override
public DocValues getValues(Map context, IndexReader reader) throws IOException {
final DocValues x1DV = x1.getValues(context, reader);
final DocValues y1DV = y1.getValues(context, reader);
final DocValues x2DV = x2.getValues(context, reader);
final DocValues y2DV = y2.getValues(context, reader);
final DocValues vals1 = p1.getValues(context, reader);
final DocValues vals2 = p2.getValues(context, reader);
return new DocValues() {
public float floatVal(int doc) {
return (float) doubleVal(doc);
@ -95,7 +112,7 @@ public class HaversineFunction extends ValueSource {
}
public double doubleVal(int doc) {
return (double) distance(doc, x1DV, y1DV, x2DV, y2DV);
return (double) distance(doc, vals1, vals2);
}
public String strVal(int doc) {
@ -106,8 +123,7 @@ public class HaversineFunction extends ValueSource {
public String toString(int doc) {
StringBuilder sb = new StringBuilder();
sb.append(name()).append('(');
sb.append(x1DV.toString(doc)).append(',').append(y1DV.toString(doc)).append(',')
.append(x2DV.toString(doc)).append(',').append(y2DV.toString(doc));
sb.append(vals1.toString(doc)).append(',').append(vals2.toString(doc));
sb.append(')');
return sb.toString();
}
@ -116,10 +132,9 @@ public class HaversineFunction extends ValueSource {
@Override
public void createWeight(Map context, Searcher searcher) throws IOException {
x1.createWeight(context, searcher);
x2.createWeight(context, searcher);
y1.createWeight(context, searcher);
y2.createWeight(context, searcher);
p1.createWeight(context, searcher);
p2.createWeight(context, searcher);
}
@Override
@ -127,20 +142,16 @@ public class HaversineFunction extends ValueSource {
if (this.getClass() != o.getClass()) return false;
HaversineFunction other = (HaversineFunction) o;
return this.name().equals(other.name())
&& x1.equals(other.x1) &&
y1.equals(other.y1) &&
x2.equals(other.x2) &&
y2.equals(other.y2) && radius == other.radius;
&& p1.equals(other.p1) &&
p2.equals(other.p2) && radius == other.radius;
}
@Override
public int hashCode() {
int result;
long temp;
result = x1.hashCode();
result = 31 * result + y1.hashCode();
result = 31 * result + x2.hashCode();
result = 31 * result + y2.hashCode();
result = p1.hashCode();
result = 31 * result + p2.hashCode();
result = 31 * result + name().hashCode();
temp = Double.doubleToRawLongBits(radius);
result = 31 * result + (int) (temp ^ (temp >>> 32));
@ -150,7 +161,7 @@ public class HaversineFunction extends ValueSource {
public String description() {
StringBuilder sb = new StringBuilder();
sb.append(name()).append('(');
sb.append(x1).append(',').append(y1).append(',').append(x2).append(',').append(y2);
sb.append(p1).append(',').append(p2);
sb.append(')');
return sb.toString();
}

18
src/java/org/apache/solr/search/function/distance/SquaredEuclideanFunction.java
View File

@ -18,6 +18,7 @@ package org.apache.solr.search.function.distance;
import org.apache.solr.search.function.DocValues;
import org.apache.solr.search.function.ValueSource;
import org.apache.solr.search.MultiValueSource;
import java.util.List;
@ -30,8 +31,8 @@ import java.util.List;
public class SquaredEuclideanFunction extends VectorDistanceFunction {
protected String name = "sqedist";
public SquaredEuclideanFunction(List<ValueSource> sources1, List<ValueSource> sources2) {
super(-1, sources1, sources2);//overriding distance, so power doesn't matter here
public SquaredEuclideanFunction(MultiValueSource source1, MultiValueSource source2) {
super(-1, source1, source2);//overriding distance, so power doesn't matter here
}
@ -43,11 +44,16 @@ public class SquaredEuclideanFunction extends VectorDistanceFunction {
/**
* @param doc The doc to score
*/
protected double distance(int doc, DocValues[] docValues1, DocValues[] docValues2) {
protected double distance(int doc, DocValues dv1, DocValues dv2) {
double result = 0;
for (int i = 0; i < docValues1.length; i++) {
result += Math.pow(docValues1[i].doubleVal(doc) - docValues2[i].doubleVal(doc), 2);
}
double [] vals1 = new double[source1.dimension()];
double [] vals2 = new double[source1.dimension()];
dv1.doubleVal(doc, vals1);
dv2.doubleVal(doc, vals2);
for (int i = 0; i < vals1.length; i++) {
double v = vals1[i] - vals2[i];
result += v * v;
}
return result;
}

122
src/java/org/apache/solr/search/function/distance/VectorDistanceFunction.java
View File

@ -21,9 +21,9 @@ import org.apache.lucene.search.Searcher;
import org.apache.solr.common.SolrException;
import org.apache.solr.search.function.DocValues;
import org.apache.solr.search.function.ValueSource;
import org.apache.solr.search.MultiValueSource;
import java.io.IOException;
import java.util.List;
import java.util.Map;
@ -41,18 +41,18 @@ import java.util.Map;
* @see SquaredEuclideanFunction for the special case
*/
public class VectorDistanceFunction extends ValueSource {
protected List<ValueSource> sources1, sources2;
protected MultiValueSource source1, source2;
protected float power;
protected float oneOverPower;
public VectorDistanceFunction(float power, List<ValueSource> sources1, List<ValueSource> sources2) {
this.power = power;
this.oneOverPower = 1 / power;
this.sources1 = sources1;
this.sources2 = sources2;
if ((sources1.size() != sources2.size())) {
public VectorDistanceFunction(float power, MultiValueSource source1, MultiValueSource source2) {
if ((source1.dimension() != source2.dimension())) {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, "Illegal number of sources");
}
this.power = power;
this.oneOverPower = 1 / power;
this.source1 = source1;
this.source2 = source2;
}
protected String name() {
@ -63,37 +63,39 @@ public class VectorDistanceFunction extends ValueSource {
* Calculate the distance
*
* @param doc The current doc
* @param docValues1 The values from the first set of value sources
* @param docValues2 The values from the second set of value sources
* @param dv1 The values from the first MultiValueSource
* @param dv2 The values from the second MultiValueSource
* @return The distance
*/
protected double distance(int doc, DocValues[] docValues1, DocValues[] docValues2) {
protected double distance(int doc, DocValues dv1, DocValues dv2) {
double result = 0;
//Handle some special cases:
double [] vals1 = new double[source1.dimension()];
double [] vals2 = new double[source1.dimension()];
dv1.doubleVal(doc, vals1);
dv2.doubleVal(doc, vals2);
if (power == 0) {
for (int i = 0; i < docValues1.length; i++) {
//sparseness measure
result += docValues1[i].doubleVal(doc) - docValues2[i].doubleVal(doc) == 0 ? 0 : 1;
for (int i = 0; i < vals1.length; i++) {
result += vals1[i] - vals2[i] == 0 ? 0 :1;
}
} else if (power == 1.0) {
for (int i = 0; i < docValues1.length; i++) {
result += docValues1[i].doubleVal(doc) - docValues2[i].doubleVal(doc);
for (int i = 0; i < vals1.length; i++) {
result += vals1[i] - vals2[i];
}
} else if (power == 2.0) {
for (int i = 0; i < docValues1.length; i++) {
double v = docValues1[i].doubleVal(doc) - docValues2[i].doubleVal(doc);
for (int i = 0; i < vals1.length; i++) {
double v = vals1[i] - vals2[i];
result += v * v;
}
result = Math.sqrt(result);
} else if (power == Integer.MAX_VALUE || Double.isInfinite(power)) {//infininte norm?
for (int i = 0; i < docValues1.length; i++) {
//TODO: is this the correct infinite norm?
result = Math.max(docValues1[i].doubleVal(doc) - docValues2[i].doubleVal(doc), result);
for (int i = 0; i < vals1.length; i++) {
result = Math.max(vals1[i], vals2[i]);
}
} else {
for (int i = 0; i < docValues1.length; i++) {
result += Math.pow(docValues1[i].doubleVal(doc) - docValues2[i].doubleVal(doc), power);
for (int i = 0; i < vals1.length; i++) {
result += Math.pow(vals1[i] - vals2[i], power);
}
result = Math.pow(result, oneOverPower);
}
@ -103,19 +105,24 @@ public class VectorDistanceFunction extends ValueSource {
@Override
public DocValues getValues(Map context, IndexReader reader) throws IOException {
final DocValues[] valsArr1 = new DocValues[sources1.size()];
int i = 0;
for (ValueSource source : sources1) {
valsArr1[i++] = source.getValues(context, reader);
}
final DocValues[] valsArr2 = new DocValues[sources2.size()];
i = 0;
for (ValueSource source : sources2) {
valsArr2[i++] = source.getValues(context, reader);
}
final DocValues vals1 = source1.getValues(context, reader);
final DocValues vals2 = source2.getValues(context, reader);
return new DocValues() {
@Override
public byte byteVal(int doc) {
return (byte) doubleVal(doc);
}
@Override
public short shortVal(int doc) {
return (short)doubleVal(doc);
}
public float floatVal(int doc) {
return (float) doubleVal(doc);
}
@ -129,7 +136,7 @@ public class VectorDistanceFunction extends ValueSource {
}
public double doubleVal(int doc) {
return distance(doc, valsArr1, valsArr2);
return distance(doc, vals1, vals2);
}
public String strVal(int doc) {
@ -141,18 +148,8 @@ public class VectorDistanceFunction extends ValueSource {
StringBuilder sb = new StringBuilder();
sb.append(name()).append('(').append(power).append(',');
boolean firstTime = true;
for (DocValues vals : valsArr1) {
if (firstTime) {
firstTime = false;
} else {
sb.append(',');
}
sb.append(vals.toString(doc));
}
for (DocValues vals : valsArr2) {
sb.append(',');//we will always have valsArr1, else there is an error
sb.append(vals.toString(doc));
}
sb.append(vals1.toString(doc)).append(',');
sb.append(vals2.toString(doc));
sb.append(')');
return sb.toString();
}
@ -161,12 +158,8 @@ public class VectorDistanceFunction extends ValueSource {
@Override
public void createWeight(Map context, Searcher searcher) throws IOException {
for (ValueSource source : sources1) {
source.createWeight(context, searcher);
}
for (ValueSource source : sources2) {
source.createWeight(context, searcher);
}
source1.createWeight(context, searcher);
source2.createWeight(context, searcher);
}
@Override
@ -177,16 +170,16 @@ public class VectorDistanceFunction extends ValueSource {
VectorDistanceFunction that = (VectorDistanceFunction) o;
if (Float.compare(that.power, power) != 0) return false;
if (!sources1.equals(that.sources1)) return false;
if (!sources2.equals(that.sources2)) return false;
if (!source1.equals(that.source1)) return false;
if (!source2.equals(that.source2)) return false;
return true;
}
@Override
public int hashCode() {
int result = sources1.hashCode();
result = 31 * result + sources2.hashCode();
int result = source1.hashCode();
result = 31 * result + source2.hashCode();
result = 31 * result + Float.floatToRawIntBits(power);
return result;
}
@ -195,19 +188,8 @@ public class VectorDistanceFunction extends ValueSource {
public String description() {
StringBuilder sb = new StringBuilder();
sb.append(name()).append('(').append(power).append(',');
boolean firstTime = true;
for (ValueSource source : sources1) {
if (firstTime) {
firstTime = false;
} else {
sb.append(',');
}
sb.append(source);
}
for (ValueSource source : sources2) {
sb.append(',');//we will always have sources1, else there is an error
sb.append(source);
}
sb.append(source1).append(',');
sb.append(source2);
sb.append(')');
return sb.toString();
}

74
src/java/org/apache/solr/update/DocumentBuilder.java
View File

@ -57,19 +57,35 @@ public class DocumentBuilder {
// we don't check for a null val ourselves because a solr.FieldType
// might actually want to map it to something. If createField()
// returns null, then we don't store the field.
Field field = sfield.createField(val, boost);
if (field != null) {
if (!sfield.multiValued()) {
String oldValue = map.put(sfield.getName(), val);
if (oldValue != null) {
throw new SolrException( SolrException.ErrorCode.BAD_REQUEST,"ERROR: multiple values encountered for non multiValued field " + sfield.getName()
+ ": first='" + oldValue + "' second='" + val + "'");
if (sfield.isPolyField()) {
Fieldable[] fields = sfield.createFields(val, boost);
if (fields != null && fields.length > 0) {
if (!sfield.multiValued()) {
String oldValue = map.put(sfield.getName(), val);
if (oldValue != null) {
throw new SolrException(SolrException.ErrorCode.BAD_REQUEST, "ERROR: multiple values encountered for non multiValued field " + sfield.getName()
+ ": first='" + oldValue + "' second='" + val + "'");
}
}
// Add each field
for (Fieldable field : fields) {
doc.add(field);
}
}
} else {
Field field = sfield.createField(val, boost);
if (field != null) {
if (!sfield.multiValued()) {
String oldValue = map.put(sfield.getName(), val);
if (oldValue != null) {
throw new SolrException( SolrException.ErrorCode.BAD_REQUEST,"ERROR: multiple values encountered for non multiValued field " + sfield.getName()
+ ": first='" + oldValue + "' second='" + val + "'");
}
}
}
// field.setBoost(boost);
doc.add(field);
}
}
/**
@ -147,7 +163,7 @@ public class DocumentBuilder {
for (SchemaField field : schema.getRequiredFields()) {
if (doc.getField(field.getName() ) == null) {
if (field.getDefaultValue() != null) {
doc.add( field.createField( field.getDefaultValue(), 1.0f ) );
addField(doc, field, field.getDefaultValue(), 1.0f);
} else {
if (missingFields==null) {
missingFields = new ArrayList<String>(1);
@ -176,6 +192,19 @@ public class DocumentBuilder {
Document ret = doc; doc=null;
return ret;
}
private static void addField(Document doc, SchemaField field, String val, float boost) {
if (field.isPolyField()) {
Fieldable[] farr = field.getType().createFields(field, val, boost);
for (Fieldable f : farr) {
if (f != null) doc.add(f); // null fields are not added
}
} else {
Field f = field.createField(val, boost);
if (f != null) doc.add(f); // null fields are not added
}
}
/**
@ -230,7 +259,9 @@ public class DocumentBuilder {
isBinaryField = true;
BinaryField binaryField = (BinaryField) sfield.getType();
Field f = binaryField.createField(sfield,v,boost);
if(f != null) out.add(f);
if(f != null){
out.add(f);
}
used = true;
} else {
// TODO!!! HACK -- date conversion
@ -243,10 +274,7 @@ public class DocumentBuilder {
if (sfield != null) {
used = true;
Field f = sfield.createField(val, boost);
if (f != null) { // null fields are not added
out.add(f);
}
addField(out, sfield, val, boost);
}
}
@ -263,17 +291,21 @@ public class DocumentBuilder {
}
used = true;
Field f = null;
//Don't worry about poly fields here
Fieldable [] fields = null;
if (isBinaryField) {
if (destinationField.getType() instanceof BinaryField) {
BinaryField binaryField = (BinaryField) destinationField.getType();
f = binaryField.createField(destinationField, v, boost);
//TODO: safe to assume that binary fields only create one?
fields = new Field[]{binaryField.createField(destinationField, v, boost)};
}
} else {
f = destinationField.createField(cf.getLimitedValue(val), boost);
fields = destinationField.createFields(cf.getLimitedValue(val), boost);
}
if (f != null) { // null fields are not added
out.add(f);
if (fields != null) { // null fields are not added
for (Fieldable f : fields) {
out.add(f);
}
}
}
@ -297,7 +329,7 @@ public class DocumentBuilder {
for (SchemaField field : schema.getRequiredFields()) {
if (out.getField(field.getName() ) == null) {
if (field.getDefaultValue() != null) {
out.add( field.createField( field.getDefaultValue(), 1.0f ) );
addField(out, field, field.getDefaultValue(), 1.0f);
}
else {
String id = schema.printableUniqueKey( out );

11
src/java/org/apache/solr/util/AbstractSolrTestCase.java
View File

@ -231,6 +231,17 @@ public abstract class AbstractSolrTestCase extends TestCase {
}
}
public void assertQEx(String message, SolrQueryRequest req, SolrException.ErrorCode code ) {
try {
h.query(req);
fail( message );
} catch (SolrException e) {
assertEquals( code.code, e.code() );
} catch (Exception e2) {
throw new RuntimeException("Exception during query", e2);
}
}
/**
* @see TestHarness#optimize

196
src/test/org/apache/solr/schema/PolyFieldTest.java Normal file
View File

@ -0,0 +1,196 @@
package org.apache.solr.schema;
/**
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF 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.
*/
import org.apache.lucene.document.Field;
import org.apache.lucene.document.Fieldable;
import org.apache.lucene.document.Document;
import org.apache.lucene.search.BooleanClause;
import org.apache.lucene.search.BooleanQuery;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.TopDocs;
import org.apache.lucene.index.IndexReader;
import org.apache.solr.core.SolrCore;
import org.apache.solr.util.AbstractSolrTestCase;
import org.apache.solr.common.SolrException;
import java.util.Map;
import java.util.Random;
/**
* Test a whole slew of things related to PolyFields
*/
public class PolyFieldTest extends AbstractSolrTestCase {
@Override
public String getSchemaFile() {
return "schema.xml";
}
@Override
public String getSolrConfigFile() {
return "solrconfig.xml";
}
public void testSchemaBasics() throws Exception {
IndexSchema schema = h.getCore().getSchema();
SchemaField home = schema.getField("home");
assertNotNull(home);
assertTrue(home.isPolyField());
SchemaField[] dynFields = schema.getDynamicFieldPrototypes();
boolean seen = false;
for (SchemaField dynField : dynFields) {
if (dynField.getName().equals("*" + FieldType.POLY_FIELD_SEPARATOR + "double")) {
seen = true;
}
}
assertTrue("Didn't find the expected dynamic field", seen);
FieldType homeFT = schema.getFieldType("home");
assertEquals(home.getType(), homeFT);
FieldType xy = schema.getFieldTypeByName("xy");
assertNotNull(xy);
assertTrue(xy instanceof PointType);
assertTrue(xy.isPolyField());
home = schema.getFieldOrNull("home_0" + FieldType.POLY_FIELD_SEPARATOR + "double");
assertNotNull(home);
home = schema.getField("home");
assertNotNull(home);
homeFT = schema.getPolyFieldType("home");
assertNotNull(homeFT);
home = schema.getField("homed");//sub field suffix
assertNotNull(home);
assertTrue(home.isPolyField());
try {
FieldType bad = schema.getPolyFieldType("foo");
assertTrue(false);
} catch (Exception e) {
}
try {
FieldType bad = schema.getPolyFieldTypeNoEx("foo");
assertNull(bad);
} catch (Exception e) {
assertTrue(false);
}
}
public void testPointFieldType() throws Exception {
SolrCore core = h.getCore();
IndexSchema schema = core.getSchema();
SchemaField home = schema.getField("home");
assertNotNull(home);
assertTrue("home is not a poly field", home.isPolyField());
FieldType tmp = home.getType();
assertTrue(tmp instanceof PointType);
PointType pt = (PointType) tmp;
assertEquals(pt.getDimension(), 2);
double[] xy = new double[]{35.0, -79.34};
String point = xy[0] + "," + xy[1];
Fieldable[] fields = home.createFields(point, 2);
assertEquals(fields.length, 3);//should be 3, we have a stored field
//first two fields contain the values, third is just stored and contains the original
for (int i = 0; i < 3; i++) {
boolean hasValue = fields[1].tokenStreamValue() != null
|| fields[1].getBinaryValue() != null
|| fields[1].stringValue() != null;
assertTrue("Doesn't have a value: " + fields[1], hasValue);
}
/*assertTrue("first field " + fields[0].tokenStreamValue() + " is not 35.0", pt.getSubType().toExternal(fields[0]).equals(String.valueOf(xy[0])));
assertTrue("second field is not -79.34", pt.getSubType().toExternal(fields[1]).equals(String.valueOf(xy[1])));
assertTrue("third field is not '35.0,-79.34'", pt.getSubType().toExternal(fields[2]).equals(point));*/
home = schema.getField("home_ns");
assertNotNull(home);
fields = home.createFields(point, 2);
assertEquals(fields.length, 2);//should be 2, since we aren't storing
home = schema.getField("home_ns");
assertNotNull(home);
try {
fields = home.createFields("35.0,foo", 2);
assertTrue(false);
} catch (Exception e) {
//
}
}
public void testSearching() throws Exception {
for (int i = 0; i < 50; i++) {
assertU(adoc("id", "" + i, "home", i + "," + (i * 100), "homed", (i * 1000) + "," + (i * 10000)));
}
assertU(commit());
IndexReader reader = h.getCore().getSearcher().get().getReader();
/*for (int i = 0; i < 50; i++){
Document doc = reader.document(i);
System.out.println("Doc: " + doc.get("homed_0___double"));
}*/
assertQ(req("fl", "*,score", "q", "*:*"), "//*[@numFound='50']");
assertQ(req("fl", "*,score", "q", "home:1,100"),
"//*[@numFound='1']",
"//str[@name='home'][.='1,100']");
assertQ(req("fl", "*,score", "q", "homed:1000,10000"),
"//*[@numFound='1']",
"//str[@name='homed'][.='1000,10000']");
assertQ(req("fl", "*,score", "q",
"{!func}sqedist(home, toMultiVS(0, 0))"),
"\"//*[@numFound='50']\"");
assertQ(req("fl", "*,score", "q",
"{!func}dist(2, home, toMultiVS(0, 0))"),
"\"//*[@numFound='50']\"");
assertQ(req("fl", "*,score", "q",
"home:[10,10000 TO 30,30000]"),
"\"//*[@numFound='3']\"");
assertQ(req("fl", "*,score", "q",
"homed:[1,1000 TO 2000,35000]"),
"\"//*[@numFound='2']\"");
//bad
assertQEx("Query should throw an exception due to incorrect dimensions", req("fl", "*,score", "q",
"homed:[1 TO 2000]"), SolrException.ErrorCode.BAD_REQUEST);
}
public void testSearchDetails() throws Exception {
SolrCore core = h.getCore();
IndexSchema schema = core.getSchema();
double[] xy = new double[]{35.0, -79.34};
String point = xy[0] + "," + xy[1];
//How about some queries?
//don't need a parser for this path currently. This may change
assertU(adoc("id", "0", "home_ns", point));
assertU(commit());
SchemaField home = schema.getField("home_ns");
PointType pt = (PointType) home.getType();
assertEquals(pt.getDimension(), 2);
Query q = pt.getFieldQuery(null, home, point);
assertNotNull(q);
assertTrue(q instanceof BooleanQuery);
//should have two clauses, one for 35.0 and the other for -79.34
BooleanQuery bq = (BooleanQuery) q;
BooleanClause[] clauses = bq.getClauses();
assertEquals(clauses.length, 2);
}
}

32
src/test/org/apache/solr/search/function/distance/DistanceFunctionTest.java
View File

@ -16,9 +16,9 @@ package org.apache.solr.search.function.distance;
* limitations under the License.
*/
import org.apache.lucene.spatial.geohash.GeoHashUtils;
import org.apache.solr.common.SolrException;
import org.apache.solr.util.AbstractSolrTestCase;
import org.apache.lucene.spatial.geohash.GeoHashUtils;
/**
@ -44,20 +44,21 @@ public class DistanceFunctionTest extends AbstractSolrTestCase {
assertU(adoc("id", "2", "x_td", "0", "y_td", String.valueOf(Math.PI / 2), "gh_s", GeoHashUtils.encode(32.7693246, -78.9289094)));
assertU(adoc("id", "3", "x_td", String.valueOf(Math.PI / 2), "y_td", String.valueOf(Math.PI / 2), "gh_s", GeoHashUtils.encode(32.7693246, -80.9289094)));
assertU(adoc("id", "4", "x_td", String.valueOf(Math.PI / 4), "y_td", String.valueOf(Math.PI / 4), "gh_s", GeoHashUtils.encode(32.7693246, -81.9289094)));
assertU(adoc("id", "5", "x_td", "45.0", "y_td", "45.0",
"gh_s", GeoHashUtils.encode(32.7693246, -81.9289094)));
assertU(commit());
//Get the haversine distance between the point 0,0 and the docs above assuming a radius of 1
assertQ(req("fl", "*,score", "q", "{!func}hsin(x_td, y_td, 0, 0, 1)", "fq", "id:1"), "//float[@name='score']='0.0'");
assertQ(req("fl", "*,score", "q", "{!func}hsin(x_td, y_td, 0, 0, 1)", "fq", "id:2"), "//float[@name='score']='" + (float) (Math.PI / 2) + "'");
assertQ(req("fl", "*,score", "q", "{!func}hsin(x_td, y_td, 0, 0, 1)", "fq", "id:3"), "//float[@name='score']='" + (float) (Math.PI / 2) + "'");
assertQ(req("fl", "*,score", "q", "{!func}hsin(x_td, y_td, 0, 0, 1)", "fq", "id:4"), "//float[@name='score']='1.0471976'");
assertQ(req("fl", "*,score", "q", "{!func}hsin(1, x_td, y_td, 0, 0)", "fq", "id:1"), "//float[@name='score']='0.0'");
assertQ(req("fl", "*,score", "q", "{!func}hsin(1, x_td, y_td, 0, 0)", "fq", "id:2"), "//float[@name='score']='" + (float) (Math.PI / 2) + "'");
assertQ(req("fl", "*,score", "q", "{!func}hsin(1, x_td, y_td, 0, 0)", "fq", "id:3"), "//float[@name='score']='" + (float) (Math.PI / 2) + "'");
assertQ(req("fl", "*,score", "q", "{!func}hsin(1, x_td, y_td, 0, 0)", "fq", "id:4"), "//float[@name='score']='1.0471976'");
assertQ(req("fl", "*,score", "q", "{!func}hsin(1, x_td, y_td, 0, 0, true)", "fq", "id:5"), "//float[@name='score']='1.0471976'");
//Geo Hash Haversine
//Can verify here: http://www.movable-type.co.uk/scripts/latlong.html, but they use a slightly different radius for the earth, so just be close
assertQ(req("fl", "*,score", "q", "{!func}ghhsin(gh_s, \"" + GeoHashUtils.encode(32, -79) +
"\"," + Constants.EARTH_RADIUS_KM +
")", "fq", "id:1"), "//float[@name='score']='122.30894'");
assertQ(req("fl", "*,score", "q", "{!func}ghhsin(gh_s, geohash(32, -79)," + Constants.EARTH_RADIUS_KM +
")", "fq", "id:1"), "//float[@name='score']='122.30894'");
assertQ(req("fl", "*,score", "q", "{!func}ghhsin(" + Constants.EARTH_RADIUS_KM + ", gh_s, \"" + GeoHashUtils.encode(32, -79) +
"\",)", "fq", "id:1"), "//float[@name='score']='122.30894'");
assertQ(req("fl", "*,score", "q", "{!func}ghhsin(" + Constants.EARTH_RADIUS_KM + ", gh_s, geohash(32, -79))", "fq", "id:1"), "//float[@name='score']='122.30894'");
}
public void testVector() throws Exception {
@ -66,6 +67,8 @@ public class DistanceFunctionTest extends AbstractSolrTestCase {
assertU(adoc("id", "3", "x_td", "1", "y_td", "1", "z_td", "1", "w_td", "1"));
assertU(adoc("id", "4", "x_td", "1", "y_td", "0", "z_td", "0", "w_td", "0"));
assertU(adoc("id", "5", "x_td", "2.3", "y_td", "5.5", "z_td", "7.9", "w_td", "-2.4"));
assertU(adoc("id", "6", "point", "1.0,0.0"));
assertU(adoc("id", "7", "point", "5.5,10.9"));
assertU(commit());
//two dimensions, notice how we only pass in 4 value sources
assertQ(req("fl", "*,score", "q", "{!func}sqedist(x_td, y_td, 0, 0)", "fq", "id:1"), "//float[@name='score']='0.0'");
@ -111,6 +114,15 @@ public class DistanceFunctionTest extends AbstractSolrTestCase {
assertQ(req("fl", "*,score", "q", "{!func}dist(1, x_td, y_td, 0, 0)", "fq", "id:3"), "//float[@name='score']='" + (float) 2.0 + "'");
assertQ(req("fl", "*,score", "q", "{!func}dist(1, x_td, y_td, 0, 0)", "fq", "id:4"), "//float[@name='score']='1.0'");
assertQ(req("fl", "*,score", "q", "{!func}dist(1, x_td, y_td, 0, 0)", "fq", "id:5"), "//float[@name='score']='" + (float) (2.3 + 5.5) + "'");
//Do point tests:
assertQ(req("fl", "*,score", "q", "{!func}dist(1, toMultiVS(x_td, y_td), toMultiVS(0, 0))", "fq", "id:5"),
"//float[@name='score']='" + (float) (2.3 + 5.5) + "'");
assertQ(req("fl", "*,score", "q", "{!func}dist(1, point, toMultiVS(0, 0))", "fq", "id:6"),
"//float[@name='score']='" + 0.0f + "'");
}
}

14
src/test/org/apache/solr/update/DocumentBuilderTest.java
View File

@ -22,6 +22,7 @@ import org.apache.solr.common.SolrException;
import org.apache.solr.common.SolrInputDocument;
import org.apache.solr.core.SolrCore;
import org.apache.solr.util.AbstractSolrTestCase;
import org.apache.solr.schema.FieldType;
/**
*
@ -59,4 +60,17 @@ public class DocumentBuilderTest extends AbstractSolrTestCase {
Document out = DocumentBuilder.toDocument( doc, core.getSchema() );
assertNull( out.get( "name" ) );
}
public void testMultiField() throws Exception {
SolrCore core = h.getCore();
// make sure a null value is not indexed
SolrInputDocument doc = new SolrInputDocument();
doc.addField( "home", "2.2,3.3", 1.0f );
Document out = DocumentBuilder.toDocument( doc, core.getSchema() );
assertNotNull( out.get( "home" ) );//contains the stored value and term vector, if there is one
assertNotNull( out.getField( "home_0" + FieldType.POLY_FIELD_SEPARATOR + "double" ) );
assertNotNull( out.getField( "home_1" + FieldType.POLY_FIELD_SEPARATOR + "double" ) );
}
}

15
src/test/test-files/solr/conf/schema.xml
View File

@ -367,6 +367,12 @@
</fieldtype>
<fieldType name="uuid" class="solr.UUIDField" />
<!-- Try out some point types -->
<fieldType name="xy" class="solr.PointType" dimension="2" subFieldType="double"/>
<fieldType name="tenD" class="solr.PointType" dimension="10" subFieldType="double"/>
<!-- Use the sub field suffix -->
<fieldType name="xyd" class="solr.PointType" dimension="2" subFieldSuffix="*_d"/>
</types>
@ -392,6 +398,15 @@
<field name="shouldbestored" type="unstored" stored="true"/>
<field name="shouldbeunindexed" type="unstored" indexed="false" stored="true"/>
<!-- Test points -->
<!-- Test points -->
<field name="home" type="xy" indexed="true" stored="true" multiValued="false"/>
<field name="homed" type="xyd" indexed="true" stored="true" multiValued="false"/>
<field name="home_ns" type="xy" indexed="true" stored="false" multiValued="false"/>
<field name="work" type="xy" indexed="true" stored="true" multiValued="false"/>
<field name="point10" type="tenD" indexed="true" stored="true" multiValued="false"/>
<!-- test different combinations of indexed and stored -->
<field name="bind" type="boolean" indexed="true" stored="false"/>

10
src/test/test-files/solr/conf/schema11.xml
View File

@ -251,6 +251,10 @@
<fieldType name="tdoubles" class="solr.TrieDoubleField" omitNorms="true" positionIncrementGap="0" precisionStep="0" multiValued="true" />
<fieldType name="tdates" class="solr.TrieDateField" omitNorms="true" positionIncrementGap="0" precisionStep="0" multiValued="true" />
<!-- Poly field -->
<fieldType name="xy" class="solr.PointType" dimension="2" subFieldType="double"/>
<fieldType name="xyd" class="solr.PointType" dimension="2" subFieldSuffix="*_d"/>
</types>
@ -275,7 +279,11 @@
<!-- for testing, a type that does a transform to see if it's correctly done everywhere -->
<field name="id" type="sfloat" indexed="true" stored="true" required="true" />
<field name="text" type="text" indexed="true" stored="false" />
<field name="text" type="text" indexed="true" stored="false" />
<!-- Test a point field for distances -->
<field name="point" type="xy" indexed="true" stored="true" multiValued="false"/>
<field name="pointD" type="xyd" indexed="true" stored="true" multiValued="false"/>
<!-- Dynamic field definitions. If a field name is not found, dynamicFields
will be used if the name matches any of the patterns.