lucene 4: fix sorting

Author: imotov

src/main/java/org/apache/lucene/search/ShardFieldDocSortedHitQueue.java

@@ -19,13 +19,12 @@
 
 package org.apache.lucene.search;
 
+import org.apache.lucene.util.BytesRef;
 import org.apache.lucene.util.PriorityQueue;
 import org.elasticsearch.ElasticSearchIllegalStateException;
 import org.elasticsearch.search.controller.ShardFieldDoc;
 
 import java.io.IOException;
-import java.text.Collator;
-import java.util.Locale;
 
 /**
  *
@@ -86,8 +85,8 @@ SortField[] getFields() {
  /**
  * Returns whether <code>a</code> is less relevant than <code>b</code>.
  *
- * @param a ScoreDoc
- * @param b ScoreDoc
+ * @param docA ScoreDoc
+ * @param docB ScoreDoc
  * @return <code>true</code> if document <code>a</code> should be sorted after document <code>b</code>.
  */
  @SuppressWarnings("unchecked")
@@ -98,8 +97,8 @@ protected final boolean lessThan(final ShardFieldDoc docA, final ShardFieldDoc d
  for (int i = 0; i < n && c == 0; ++i) {
  final SortField.Type type = fields[i].getType();
  if (type == SortField.Type.STRING) {
- final String s1 = (String) docA.fields[i];
- final String s2 = (String) docB.fields[i];
+ final BytesRef s1 = (BytesRef) docA.fields[i];
+ final BytesRef s2 = (BytesRef) docB.fields[i];
  // null values need to be sorted first, because of how FieldCache.getStringIndex()
  // works - in that routine, any documents without a value in the given field are
  // put first. If both are null, the next SortField is used

src/main/java/org/elasticsearch/common/lucene/Lucene.java

@@ -24,6 +24,7 @@
 import org.apache.lucene.index.*;
 import org.apache.lucene.search.*;
 import org.apache.lucene.store.Directory;
+import org.apache.lucene.util.BytesRef;
 import org.apache.lucene.util.Version;
 import org.elasticsearch.common.Nullable;
 import org.elasticsearch.common.io.stream.StreamInput;
@@ -173,6 +174,8 @@ public static TopDocs readTopDocs(StreamInput in) throws IOException {
  cFields[j] = in.readShort();
  } else if (type == 8) {
  cFields[j] = in.readBoolean();
+ } else if (type == 9) {
+ cFields[j] = in.readBytesRef();
  } else {
  throw new IOException("Can't match type [" + type + "]");
  }
@@ -258,6 +261,9 @@ public static void writeTopDocs(StreamOutput out, TopDocs topDocs, int from) thr
  } else if (type == Boolean.class) {
  out.writeByte((byte) 8);
  out.writeBoolean((Boolean) field);
+ } else if (type == BytesRef.class) {
+ out.writeByte((byte) 9);
+ out.writeBytesRef((BytesRef) field);
  } else {
  throw new IOException("Can't handle sort field value of type [" + type + "]");
  }

src/main/java/org/elasticsearch/index/field/data/DocFieldData.java

@@ -46,11 +46,15 @@ public boolean isEmpty() {
  return !fieldData.hasValue(docId);
  }
 
- public BytesRef stringValue() {
- return fieldData.stringValue(docId);
+ public String stringValue() {
+ BytesRef val = fieldData.stringValue(docId);
+ if (val == null) {
+ return null;
+ }
+ return val.utf8ToString();
  }
 
- public BytesRef getStringValue() {
+ public String getStringValue() {
  return stringValue();
  }
 

src/main/java/org/elasticsearch/index/field/data/strings/StringDocFieldData.java

@@ -31,11 +31,31 @@ public StringDocFieldData(StringFieldData fieldData) {
  super(fieldData);
  }
 
- public BytesRef getValue() {
+ public String getValue() {
+ BytesRef value = fieldData.value(docId);
+ if (value == null) {
+ return null;
+ }
+ return value.utf8ToString();
+ }
+
+ public String[] getValues() {
+ BytesRef[] values = fieldData.values(docId);
+ if (values == null) {
+ return null;
+ }
+ String[] stringValues = new String[values.length];
+ for (int i = 0; i < values.length; i++) {
+ stringValues[i] = values[i].utf8ToString();
+ }
+ return stringValues;
+ }
+
+ public BytesRef getBytesValue() {
  return fieldData.value(docId);
  }
 
- public BytesRef[] getValues() {
+ public BytesRef[] getBytesValues() {
  return fieldData.values(docId);
  }
 }

src/main/java/org/elasticsearch/index/field/data/support/FieldDataLoader.java

@@ -46,7 +46,7 @@ public static <T extends FieldData> T load(AtomicReader reader, String field, Ty
 
  Terms terms = reader.terms(field);
  if (terms == null) {
- return loader.buildSingleValue(field, new int[0]); // Return empty field data if field doesn't exists.
+ return loader.buildSingleValue(field, ordinals.get(0)); // Return empty field data if field doesn't exists.
  }
 
  TermsEnum termsEnum = terms.iterator(null);

src/main/java/org/elasticsearch/search/internal/InternalSearchHit.java

@@ -21,6 +21,7 @@
 
 import com.google.common.collect.ImmutableMap;
 import org.apache.lucene.search.Explanation;
+import org.apache.lucene.util.BytesRef;
 import org.elasticsearch.ElasticSearchParseException;
 import org.elasticsearch.common.Nullable;
 import org.elasticsearch.common.Strings;
@@ -29,6 +30,7 @@
 import org.elasticsearch.common.compress.CompressorFactory;
 import org.elasticsearch.common.io.stream.StreamInput;
 import org.elasticsearch.common.io.stream.StreamOutput;
+import org.elasticsearch.common.text.StringAndBytesText;
 import org.elasticsearch.common.text.Text;
 import org.elasticsearch.common.xcontent.XContentBuilder;
 import org.elasticsearch.common.xcontent.XContentBuilderString;
@@ -295,6 +297,14 @@ public void highlightFields(Map<String, HighlightField> highlightFields) {
  }
 
  public void sortValues(Object[] sortValues) {
+ // LUCENE 4 UPGRADE: There must be a better way
+ if (sortValues != null) {
+ for (int i=0; i<sortValues.length; i++) {
+ if (sortValues[i] instanceof BytesRef) {
+ sortValues[i] = new StringAndBytesText(new BytesArray((BytesRef)sortValues[i]));
+ }
+ }
+ }
  this.sortValues = sortValues;
  }
 
@@ -571,6 +581,9 @@ public void readFrom(StreamInput in, InternalSearchHits.StreamContext context) t
  sortValues[i] = in.readShort();
  } else if (type == 8) {
  sortValues[i] = in.readBoolean();
+ } else if (type == 9) {
+ // LUCENE 4 UPGRADE: There must be a better way
+ sortValues[i] = new StringAndBytesText(new BytesArray(in.readBytesRef()));
  } else {
  throw new IOException("Can't match type [" + type + "]");
  }
@@ -664,6 +677,9 @@ public void writeTo(StreamOutput out, InternalSearchHits.StreamContext context)
  } else if (type == Boolean.class) {
  out.writeByte((byte) 8);
  out.writeBoolean((Boolean) sortValue);
+ } else if (type == BytesRef.class) {
+ out.writeByte((byte) 9);
+ out.writeBytesRef((BytesRef) sortValue);
  } else {
  throw new IOException("Can't handle sort field value of type [" + type + "]");
  }

src/test/java/org/elasticsearch/test/unit/index/field/data/strings/StringFieldDataTests.java

@@ -82,11 +82,11 @@ public void stringFieldDataTests() throws Exception {
  // svalue
  assertThat(sFieldData.hasValue(0), equalTo(true));
  assertThat(sFieldData.value(0).utf8ToString(), equalTo("zzz"));
- assertThat(sFieldData.docFieldData(0).getValue().utf8ToString(), equalTo("zzz"));
+ assertThat(sFieldData.docFieldData(0).getValue(), equalTo("zzz"));
  assertThat(sFieldData.values(0).length, equalTo(1));
  assertThat(sFieldData.docFieldData(0).getValues().length, equalTo(1));
  assertThat(sFieldData.values(0)[0].utf8ToString(), equalTo("zzz"));
- assertThat(sFieldData.docFieldData(0).getValues()[0].utf8ToString(), equalTo("zzz"));
+ assertThat(sFieldData.docFieldData(0).getValues()[0], equalTo("zzz"));
 
  assertThat(sFieldData.hasValue(1), equalTo(true));
  assertThat(sFieldData.value(1).utf8ToString(), equalTo("xxx"));