honeymoose
/
OpenSearch
mirror of https://github.com/honeymoose/OpenSearch.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Fielddata: Remove BytesValues.WithOrdinals.currentOrd and copyShared. 

These methods don't exist in Lucene's sorted set doc values.

Relates to #6524
This commit is contained in:
Adrien Grand 2014-06-18 02:58:30 +02:00
parent 9e624942d8
commit 8ccfca3a2f
24 changed files with 36 additions and 167 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
src/main/java/org/elasticsearch/index/fielddata/AtomicFieldData.java
View File

@ -19,10 +19,8 @@
package org.elasticsearch.index.fielddata;
import org.apache.lucene.index.TermsEnum;
import org.apache.lucene.util.BytesRef;
import org.elasticsearch.index.fielddata.ScriptDocValues.Strings;
import org.elasticsearch.index.fielddata.plain.AtomicFieldDataWithOrdinalsTermsEnum;
/**
* The thread safe {@link org.apache.lucene.index.AtomicReader} level cache of the data.
@ -90,19 +88,9 @@ public interface AtomicFieldData<Script extends ScriptDocValues> extends RamUsag
public long getMaxOrd() {
return 0;
}
@Override
public long currentOrd() {
return MISSING_ORDINAL;
}
};
}
@Override
public TermsEnum getTermsEnum() {
return new AtomicFieldDataWithOrdinalsTermsEnum(this);
}
};
/**
@ -111,11 +99,6 @@ public interface AtomicFieldData<Script extends ScriptDocValues> extends RamUsag
*/
BytesValues.WithOrdinals getBytesValues();
/**
* Returns a terms enum to iterate over all the underlying values.
*/
TermsEnum getTermsEnum();
}
/**

24
src/main/java/org/elasticsearch/index/fielddata/BytesValues.java
View File

@ -20,8 +20,10 @@
package org.elasticsearch.index.fielddata;
import org.apache.lucene.index.SortedSetDocValues;
import org.apache.lucene.index.TermsEnum;
import org.apache.lucene.util.BytesRef;
import org.elasticsearch.ElasticsearchIllegalStateException;
import org.elasticsearch.index.fielddata.plain.BytesValuesWithOrdinalsTermsEnum;
/**
* A state-full lightweight per document set of <code>byte[]</code> values.
@ -60,15 +62,6 @@ public abstract class BytesValues {
return multiValued;
}
/**
* Converts the current shared {@link BytesRef} to a stable instance. Note,
* this calls makes the bytes safe for *reads*, not writes (into the same BytesRef). For example,
* it makes it safe to be placed in a map.
*/
public BytesRef copyShared() {
return BytesRef.deepCopyOf(scratch);
}
/**
* Sets iteration to the specified docID and returns the number of
* values for this document ID,
@ -139,12 +132,6 @@ public abstract class BytesValues {
*/
public abstract long nextOrd();
/**
* Returns the current ordinal in the iteration
* @return the current ordinal in the iteration
*/
public abstract long currentOrd();
/**
* Returns the value for the given ordinal.
* @param ord the ordinal to lookup.
@ -157,6 +144,13 @@ public abstract class BytesValues {
public BytesRef nextValue() {
return getValueByOrd(nextOrd());
}
/**
* Returns a terms enum to iterate over all the underlying values.
*/
public TermsEnum getTermsEnum() {
return new BytesValuesWithOrdinalsTermsEnum(this);
}
}
/**

10
src/main/java/org/elasticsearch/index/fielddata/ordinals/GlobalOrdinalMapping.java
View File

@ -43,11 +43,6 @@ public class GlobalOrdinalMapping extends BytesValues.WithOrdinals {
int readerIndex;
@Override
public BytesRef copyShared() {
return bytesValues[readerIndex].copyShared();
}
@Override
public long getMaxOrd() {
return ordinalMap.getValueCount();
@ -67,11 +62,6 @@ public class GlobalOrdinalMapping extends BytesValues.WithOrdinals {
return getGlobalOrd(values.nextOrd());
}
@Override
public long currentOrd() {
return getGlobalOrd(values.currentOrd());
}
@Override
public int setDocument(int docId) {
return values.setDocument(docId);

2
src/main/java/org/elasticsearch/index/fielddata/ordinals/InternalGlobalOrdinalsBuilder.java
View File

@ -49,7 +49,7 @@ public class InternalGlobalOrdinalsBuilder extends AbstractIndexComponent implem
final TermsEnum[] subs = new TermsEnum[indexReader.leaves().size()];
for (int i = 0; i < indexReader.leaves().size(); ++i) {
atomicFD[i] = indexFieldData.load(indexReader.leaves().get(i));
subs[i] = atomicFD[i].getTermsEnum();
subs[i] = atomicFD[i].getBytesValues().getTermsEnum();
}
final OrdinalMap ordinalMap = new OrdinalMap(null, subs);
final long memorySizeInBytes = ordinalMap.ramBytesUsed();

7
src/main/java/org/elasticsearch/index/fielddata/ordinals/InternalGlobalOrdinalsIndexFieldData.java
View File

@ -20,14 +20,12 @@ package org.elasticsearch.index.fielddata.ordinals;
import org.apache.lucene.index.AtomicReaderContext;
import org.apache.lucene.index.MultiDocValues.OrdinalMap;
import org.apache.lucene.index.TermsEnum;
import org.elasticsearch.common.settings.Settings;
import org.elasticsearch.index.Index;
import org.elasticsearch.index.fielddata.AtomicFieldData;
import org.elasticsearch.index.fielddata.BytesValues;
import org.elasticsearch.index.fielddata.FieldDataType;
import org.elasticsearch.index.fielddata.ScriptDocValues;
import org.elasticsearch.index.fielddata.plain.AtomicFieldDataWithOrdinalsTermsEnum;
import org.elasticsearch.index.mapper.FieldMapper;
/**
@ -86,11 +84,6 @@ final class InternalGlobalOrdinalsIndexFieldData extends GlobalOrdinalsIndexFiel
throw new UnsupportedOperationException("Script values not supported on global ordinals");
}
@Override
public TermsEnum getTermsEnum() {
return new AtomicFieldDataWithOrdinalsTermsEnum(this);
}
@Override
public void close() {
}

17
src/main/java/org/elasticsearch/index/fielddata/ordinals/MultiOrdinals.java
View File

@ -93,7 +93,6 @@ public class MultiOrdinals extends Ordinals {
private final AppendingPackedLongBuffer ords;
private long offset;
private long limit;
private long currentOrd;
private final ValuesHolder values;
MultiDocs(MultiOrdinals ordinals, ValuesHolder values) {
@ -114,16 +113,16 @@ public class MultiOrdinals extends Ordinals {
final long startOffset = docId > 0 ? endOffsets.get(docId - 1) : 0;
final long endOffset = endOffsets.get(docId);
if (startOffset == endOffset) {
return currentOrd = MISSING_ORDINAL; // ord for missing values
return MISSING_ORDINAL; // ord for missing values
} else {
return currentOrd = ords.get(startOffset);
return ords.get(startOffset);
}
}
@Override
public long nextOrd() {
assert offset < limit;
return currentOrd = ords.get(offset++);
return ords.get(offset++);
}
@Override
@ -135,19 +134,9 @@ public class MultiOrdinals extends Ordinals {
return (int) (endOffset - startOffset);
}
@Override
public long currentOrd() {
return currentOrd;
}
@Override
public BytesRef getValueByOrd(long ord) {
return values.getValueByOrd(ord);
}
@Override
public BytesRef copyShared() {
return values.copy(scratch);
}
}
}

10
src/main/java/org/elasticsearch/index/fielddata/ordinals/SinglePackedOrdinals.java
View File

@ -95,19 +95,9 @@ public class SinglePackedOrdinals extends Ordinals {
return 1 + (int) Math.min(currentOrdinal, 0);
}
@Override
public long currentOrd() {
return currentOrdinal;
}
@Override
public BytesRef getValueByOrd(long ord) {
return values.getValueByOrd(ord);
}
@Override
public BytesRef copyShared() {
return values.copy(scratch);
}
}
}

1
src/main/java/org/elasticsearch/index/fielddata/plain/FSTBytesAtomicFieldData.java
View File

@ -18,7 +18,6 @@
*/
package org.elasticsearch.index.fielddata.plain;
import org.apache.lucene.index.TermsEnum;
import org.apache.lucene.util.BytesRef;
import org.apache.lucene.util.IntsRef;
import org.apache.lucene.util.fst.FST;

11
src/main/java/org/elasticsearch/index/fielddata/plain/IndexIndexFieldData.java
View File

@ -21,7 +21,6 @@ package org.elasticsearch.index.fielddata.plain;
import org.apache.lucene.index.AtomicReaderContext;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.TermsEnum;
import org.apache.lucene.util.BytesRef;
import org.elasticsearch.common.settings.Settings;
import org.elasticsearch.index.Index;
@ -75,11 +74,6 @@ public class IndexIndexFieldData implements IndexFieldData.WithOrdinals<AtomicFi
return 1;
}
@Override
public long currentOrd() {
return BytesValues.WithOrdinals.MIN_ORDINAL;
}
@Override
public BytesRef getValueByOrd(long ord) {
return scratch;
@ -114,11 +108,6 @@ public class IndexIndexFieldData implements IndexFieldData.WithOrdinals<AtomicFi
public void close() {
}
@Override
public TermsEnum getTermsEnum() {
return new AtomicFieldDataWithOrdinalsTermsEnum(this);
}
}
private final FieldMapper.Names names;

1
src/main/java/org/elasticsearch/index/fielddata/plain/PagedBytesAtomicFieldData.java
View File

@ -18,7 +18,6 @@
*/
package org.elasticsearch.index.fielddata.plain;
import org.apache.lucene.index.TermsEnum;
import org.apache.lucene.util.BytesRef;
import org.apache.lucene.util.PagedBytes;
import org.apache.lucene.util.packed.MonotonicAppendingLongBuffer;

3
src/main/java/org/elasticsearch/index/fielddata/plain/ParentChildAtomicFieldData.java
View File

@ -68,8 +68,7 @@ public class ParentChildAtomicFieldData implements AtomicFieldData {
int numValues = values.setDocument(docId);
assert numValues <= 1 : "Per doc/type combination only a single value is allowed";
if (numValues == 1) {
values.nextValue();
terms[counter++] = values.copyShared();
terms[counter++] = BytesRef.deepCopyOf(values.nextValue());
}
}
assert counter <= 2 : "A single doc can potentially be both parent and child, so the maximum allowed values is 2";

10
src/main/java/org/elasticsearch/index/fielddata/plain/SortedSetDVAtomicFieldData.java
View File

@ -96,7 +96,6 @@ abstract class SortedSetDVAtomicFieldData {
private final SortedSetDocValues values;
private long[] ords;
private int ordIndex = Integer.MAX_VALUE;
private long currentOrdinal = -1;
SortedSetValues(SortedSetDocValues values) {
super(DocValues.unwrapSingleton(values) == null);
@ -112,13 +111,13 @@ abstract class SortedSetDVAtomicFieldData {
@Override
public long getOrd(int docId) {
values.setDocument(docId);
return currentOrdinal = values.nextOrd();
return values.nextOrd();
}
@Override
public long nextOrd() {
assert ordIndex < ords.length;
return currentOrdinal = ords[ordIndex++];
return ords[ordIndex++];
}
@Override
@ -135,11 +134,6 @@ abstract class SortedSetDVAtomicFieldData {
return i;
}
@Override
public long currentOrd() {
return currentOrdinal;
}
@Override
public BytesRef getValueByOrd(long ord) {
values.lookupOrd(ord, scratch);

2
src/main/java/org/elasticsearch/index/percolator/QueriesLoaderCollector.java
View File

@ -74,7 +74,7 @@ final class QueriesLoaderCollector extends Collector {
// id is only used for logging, if we fail we log the id in the catch statement
final Query parseQuery = percolator.parsePercolatorDocument(null, fieldsVisitor.source());
if (parseQuery != null) {
queries.put(idValues.copyShared(), parseQuery);
queries.put(BytesRef.deepCopyOf(id), parseQuery);
} else {
logger.warn("failed to add query [{}] - parser returned null", id);
}

13
src/main/java/org/elasticsearch/index/search/child/ParentIdsFilter.java
View File

@ -23,14 +23,17 @@ import org.apache.lucene.queries.TermFilter;
import org.apache.lucene.search.DocIdSet;
import org.apache.lucene.search.DocIdSetIterator;
import org.apache.lucene.search.Filter;
import org.apache.lucene.util.*;
import org.apache.lucene.util.Bits;
import org.apache.lucene.util.BytesRef;
import org.apache.lucene.util.FixedBitSet;
import org.apache.lucene.util.LongBitSet;
import org.elasticsearch.common.lease.Releasables;
import org.elasticsearch.common.lucene.search.AndFilter;
import org.elasticsearch.common.util.BytesRefHash;
import org.elasticsearch.common.util.LongHash;
import org.elasticsearch.index.fielddata.BytesValues;
import org.elasticsearch.index.mapper.Uid;
import org.elasticsearch.index.mapper.internal.UidFieldMapper;
import org.elasticsearch.common.util.BytesRefHash;
import org.elasticsearch.search.internal.SearchContext;
import java.io.IOException;
@ -49,8 +52,7 @@ final class ParentIdsFilter extends Filter {
String parentType, BytesValues.WithOrdinals globalValues,
LongBitSet parentOrds, long numFoundParents) {
if (numFoundParents == 1) {
globalValues.getValueByOrd(parentOrds.nextSetBit(0));
BytesRef id = globalValues.copyShared();
BytesRef id = globalValues.getValueByOrd(parentOrds.nextSetBit(0));
if (nonNestedDocsFilter != null) {
List<Filter> filters = Arrays.asList(
new TermFilter(new Term(UidFieldMapper.NAME, Uid.createUidAsBytes(parentType, id))),
@ -83,8 +85,7 @@ final class ParentIdsFilter extends Filter {
String parentType, BytesValues.WithOrdinals globalValues,
LongHash parentIdxs, long numFoundParents) {
if (numFoundParents == 1) {
globalValues.getValueByOrd(parentIdxs.get(0));
BytesRef id = globalValues.copyShared();
BytesRef id = globalValues.getValueByOrd(parentIdxs.get(0));
if (nonNestedDocsFilter != null) {
List<Filter> filters = Arrays.asList(
new TermFilter(new Term(UidFieldMapper.NAME, Uid.createUidAsBytes(parentType, id))),

2
src/main/java/org/elasticsearch/percolator/PercolatorService.java
View File

@ -765,7 +765,7 @@ public class PercolatorService extends AbstractComponent {
final int numValues = values.setDocument(localDocId);
assert numValues == 1;
BytesRef bytes = values.nextValue();
matches.add(values.copyShared());
matches.add(BytesRef.deepCopyOf(bytes));
if (hls != null) {
Query query = context.percolateQueries().get(bytes);
context.parsedQuery(new ParsedQuery(query, ImmutableMap.<String, Filter>of()));

4
src/main/java/org/elasticsearch/percolator/QueryCollector.java
View File

@ -212,7 +212,7 @@ abstract class QueryCollector extends Collector {
}
if (collector.exists()) {
if (!limit || counter < size) {
matches.add(values.copyShared());
matches.add(BytesRef.deepCopyOf(current));
if (context.highlight() != null) {
highlightPhase.hitExecute(context, context.hitContext());
hls.add(context.hitContext().hit().getHighlightFields());
@ -334,7 +334,7 @@ abstract class QueryCollector extends Collector {
}
if (collector.exists()) {
if (!limit || counter < size) {
matches.add(values.copyShared());
matches.add(BytesRef.deepCopyOf(current));
scores.add(scorer.score());
if (context.highlight() != null) {
highlightPhase.hitExecute(context, context.hitContext());

10
src/main/java/org/elasticsearch/search/aggregations/bucket/terms/GlobalOrdinalsStringTermsAggregator.java
View File

@ -376,19 +376,9 @@ public class GlobalOrdinalsStringTermsAggregator extends AbstractStringTermsAggr
return numAcceptedOrds;
}
@Override
public long currentOrd() {
return currentOrd;
}
@Override
public BytesRef getValueByOrd(long ord) {
return inner.getValueByOrd(ord);
}
@Override
public BytesRef copyShared() {
return inner.copyShared();
}
}
}

7
src/main/java/org/elasticsearch/search/aggregations/bucket/terms/StringTermsAggregator.java
View File

@ -89,6 +89,7 @@ public class StringTermsAggregator extends AbstractStringTermsAggregator {
}
}
// TODO: use terms enum
/** Returns an iterator over the field data terms. */
private static Iterator<BytesRef> terms(final BytesValues.WithOrdinals bytesValues, boolean reverse) {
if (reverse) {
@ -103,8 +104,7 @@ public class StringTermsAggregator extends AbstractStringTermsAggregator {
@Override
public BytesRef next() {
bytesValues.getValueByOrd(i--);
return bytesValues.copyShared();
return BytesRef.deepCopyOf(bytesValues.getValueByOrd(i--));
}
};
@ -120,8 +120,7 @@ public class StringTermsAggregator extends AbstractStringTermsAggregator {
@Override
public BytesRef next() {
bytesValues.getValueByOrd(i++);
return bytesValues.copyShared();
return BytesRef.deepCopyOf(bytesValues.getValueByOrd(i++));
}
};

2
src/main/java/org/elasticsearch/search/aggregations/bucket/terms/support/IncludeExclude.java
View File

@ -79,7 +79,7 @@ public class IncludeExclude {
* Computes which global ordinals are accepted by this IncludeExclude instance.
*/
public LongBitSet acceptedGlobalOrdinals(BytesValues.WithOrdinals globalOrdinals, ValuesSource.Bytes.WithOrdinals valueSource) {
TermsEnum globalTermsEnum = valueSource.getGlobalTermsEnum();
TermsEnum globalTermsEnum = valueSource.globalBytesValues().getTermsEnum();
LongBitSet acceptedGlobalOrdinals = new LongBitSet(globalOrdinals.getMaxOrd());
try {
for (BytesRef term = globalTermsEnum.next(); term != null; term = globalTermsEnum.next()) {

8
src/main/java/org/elasticsearch/search/aggregations/support/ValuesSource.java
View File

@ -21,7 +21,6 @@ package org.elasticsearch.search.aggregations.support;
import org.apache.lucene.index.AtomicReaderContext;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.IndexReaderContext;
import org.apache.lucene.index.TermsEnum;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.util.ArrayUtil;
import org.apache.lucene.util.BytesRef;
@ -176,8 +175,6 @@ public abstract class ValuesSource {
public abstract long globalMaxOrd(IndexSearcher indexSearcher);
public abstract TermsEnum getGlobalTermsEnum();
public static class FieldData extends WithOrdinals implements ReaderContextAware {
protected final IndexFieldData.WithOrdinals<?> indexFieldData;
@ -262,11 +259,6 @@ public abstract class ValuesSource {
return maxOrd = values.getMaxOrd();
}
}
@Override
public TermsEnum getGlobalTermsEnum() {
return globalAtomicFieldData.getTermsEnum();
}
}
}

2
src/main/java/org/elasticsearch/search/facet/terms/strings/TermsStringOrdinalsFacetExecutor.java
View File

@ -280,7 +280,7 @@ public class TermsStringOrdinalsFacetExecutor extends FacetExecutor {
}
public BytesRef copyCurrent() {
return values.copyShared();
return BytesRef.deepCopyOf(current);
}
@Override

4
src/main/java/org/elasticsearch/search/facet/termsstats/strings/TermsStatsStringFacetExecutor.java
View File

@ -170,7 +170,7 @@ public class TermsStatsStringFacetExecutor extends FacetExecutor {
spare.reset(value, hashCode);
InternalTermsStatsStringFacet.StringEntry stringEntry = entries.get(spare);
if (stringEntry == null) {
HashedBytesRef theValue = new HashedBytesRef(values.copyShared(), hashCode);
HashedBytesRef theValue = new HashedBytesRef(BytesRef.deepCopyOf(value), hashCode);
stringEntry = new InternalTermsStatsStringFacet.StringEntry(theValue, 0, 0, 0, Double.POSITIVE_INFINITY, Double.NEGATIVE_INFINITY);
entries.put(theValue, stringEntry);
}
@ -210,7 +210,7 @@ public class TermsStatsStringFacetExecutor extends FacetExecutor {
spare.reset(value, hashCode);
InternalTermsStatsStringFacet.StringEntry stringEntry = entries.get(spare);
if (stringEntry == null) {
HashedBytesRef theValue = new HashedBytesRef(values.copyShared(), hashCode);
HashedBytesRef theValue = new HashedBytesRef(BytesRef.deepCopyOf(value), hashCode);
stringEntry = new InternalTermsStatsStringFacet.StringEntry(theValue, 1, 0, 0, Double.POSITIVE_INFINITY, Double.NEGATIVE_INFINITY);
entries.put(theValue, stringEntry);
} else {

2
src/test/java/org/elasticsearch/index/fielddata/AbstractStringFieldDataTests.java
View File

@ -506,7 +506,7 @@ public abstract class AbstractStringFieldDataTests extends AbstractFieldDataImpl
IndexFieldData.WithOrdinals ifd = getForField("value");
AtomicFieldData.WithOrdinals afd = ifd.load(atomicReaderContext);
TermsEnum termsEnum = afd.getTermsEnum();
TermsEnum termsEnum = afd.getBytesValues().getTermsEnum();
int size = 0;
while (termsEnum.next() != null) {
size++;

34
src/test/java/org/elasticsearch/search/aggregations/support/FieldDataSourceTests.java
View File

@ -112,40 +112,8 @@ public class FieldDataSourceTests extends ElasticsearchTestCase {
};
}
private static void assertConsistent(BytesValues values) {
final int numDocs = scaledRandomIntBetween(10, 100);
for (int i = 0; i < numDocs; ++i) {
final int valueCount = values.setDocument(i);
for (int j = 0; j < valueCount; ++j) {
final BytesRef term = values.nextValue();
assertTrue(term.bytesEquals(values.copyShared()));
}
}
}
@Test
public void bytesValuesWithScript() {
final BytesValues values = randomBytesValues();
ValuesSource source = new ValuesSource.Bytes() {
@Override
public BytesValues bytesValues() {
return values;
}
@Override
public MetaData metaData() {
throw new UnsupportedOperationException();
}
};
SearchScript script = randomScript();
assertConsistent(new ValuesSource.WithScript.BytesValues(source, script));
}
@Test
public void sortedUniqueBytesValues() {
assertConsistent(new ValuesSource.Bytes.SortedAndUnique.SortedUniqueBytesValues(randomBytesValues()));
assertSortedAndUnique(new ValuesSource.Bytes.SortedAndUnique.SortedUniqueBytesValues(randomBytesValues()));
}
@ -160,7 +128,7 @@ public class FieldDataSourceTests extends ElasticsearchTestCase {
if (j > 0) {
assertThat(BytesRef.getUTF8SortedAsUnicodeComparator().compare(ref.get(ref.size() - 1), term), lessThan(0));
}
ref.add(values.copyShared());
ref.add(BytesRef.deepCopyOf(term));
}
}
}