Copilot commented on code in PR #6095:
URL: https://github.com/apache/hive/pull/6095#discussion_r2384169202
##########
iceberg/iceberg-handler/src/main/java/org/apache/iceberg/mr/hive/BaseHiveIcebergMetaHook.java:
##########
@@ -217,31 +219,73 @@ private void validateCatalogConfigsDefined() {
}
}
+ /**
+ * Persists the table's write sort order based on the HMS property
'default-sort-order'
+ * that is populated by the DDL layer.
+ *
+ * Behaviour:
+ * - If the JSON represents Z-order, we remove DEFAULT_SORT_ORDER
+ * as Iceberg does not have Z-order support in its spec.
+ * So, we persist Z-order metadata in 'sort.order' and 'sort.columns' to
be used by Hive Writer.
+ * - Otherwise, the JSON is a list of SortFields; we convert it to Iceberg
+ * SortOrder JSON and keep it in DEFAULT_SORT_ORDER for Iceberg to use it.
+ */
private void setSortOrder(org.apache.hadoop.hive.metastore.api.Table
hmsTable, Schema schema,
Properties properties) {
- String sortOderJSONString =
hmsTable.getParameters().get(TableProperties.DEFAULT_SORT_ORDER);
- SortFields sortFields = null;
- if (!Strings.isNullOrEmpty(sortOderJSONString)) {
+ String sortOrderJSONString =
hmsTable.getParameters().get(TableProperties.DEFAULT_SORT_ORDER);
+ if (!Strings.isNullOrEmpty(sortOrderJSONString)) {
try {
- sortFields = JSON_OBJECT_MAPPER.reader().readValue(sortOderJSONString,
SortFields.class);
+ if (isZOrderJSON(sortOrderJSONString)) {
+ properties.remove(TableProperties.DEFAULT_SORT_ORDER);
+ ZorderFields zorderFields =
JSON_OBJECT_MAPPER.reader().readValue(sortOrderJSONString, ZorderFields.class);
+ if (zorderFields != null &&
!zorderFields.getZOrderFields().isEmpty()) {
+ setZOrderSortOrder(zorderFields, properties);
+ }
+ } else {
+ SortFields sortFields =
JSON_OBJECT_MAPPER.reader().readValue(sortOrderJSONString, SortFields.class);
+ if (sortFields != null && !sortFields.getSortFields().isEmpty()) {
+ SortOrder.Builder sortOrderBuilder = SortOrder.builderFor(schema);
+ sortFields.getSortFields().forEach(fieldDesc -> {
+ NullOrder nullOrder = fieldDesc.getNullOrdering() ==
NullOrdering.NULLS_FIRST ?
+ NullOrder.NULLS_FIRST : NullOrder.NULLS_LAST;
+ SortDirection sortDirection = fieldDesc.getDirection() ==
SortFieldDesc.SortDirection.ASC ?
+ SortDirection.ASC : SortDirection.DESC;
+ sortOrderBuilder.sortBy(fieldDesc.getColumnName(),
sortDirection, nullOrder);
+ });
+ properties.put(TableProperties.DEFAULT_SORT_ORDER,
SortOrderParser.toJson(sortOrderBuilder.build()));
+ }
+ }
} catch (Exception e) {
- LOG.warn("Can not read write order json: {}", sortOderJSONString, e);
- return;
- }
- if (sortFields != null && !sortFields.getSortFields().isEmpty()) {
- SortOrder.Builder sortOderBuilder = SortOrder.builderFor(schema);
- sortFields.getSortFields().forEach(fieldDesc -> {
- NullOrder nullOrder = fieldDesc.getNullOrdering() ==
NullOrdering.NULLS_FIRST ?
- NullOrder.NULLS_FIRST : NullOrder.NULLS_LAST;
- SortDirection sortDirection = fieldDesc.getDirection() ==
SortFieldDesc.SortDirection.ASC ?
- SortDirection.ASC : SortDirection.DESC;
- sortOderBuilder.sortBy(fieldDesc.getColumnName(), sortDirection,
nullOrder);
- });
- properties.put(TableProperties.DEFAULT_SORT_ORDER,
SortOrderParser.toJson(sortOderBuilder.build()));
+ LOG.warn("Can not read write order json: {}", sortOrderJSONString, e);
}
}
}
+ /**
+ * Configures the Z-order sort order metadata in the given properties
+ * based on the specified Z-order fields.
+ *
+ * @param zOrderFields the ZorderFields containing columns for Z-order
sorting
+ * @param properties the Properties object to store sort order metadata
+ */
+ private void setZOrderSortOrder(ZorderFields zOrderFields, Properties
properties) {
+ List<String> columnNames = zOrderFields.getZOrderFields().stream()
+ .map(ZOrderFieldDesc::getColumnName)
+ .collect(Collectors.toList());
+
+ LOG.info("Setting Z-order sort order for columns: {}", columnNames);
+
+ properties.put("sort.order", "ZORDER");
+ properties.put("sort.columns", String.join(",", columnNames));
+
+ LOG.info("Z-order sort order configured for Iceberg table with columns:
{}", columnNames);
+ }
+
+ private boolean isZOrderJSON(String jsonString) {
+ return jsonString.contains("zorderFields");
Review Comment:
Using string contains for JSON detection is unreliable and could produce
false positives. Consider parsing the JSON to check for the actual structure or
use a more robust detection method.
```suggestion
try {
com.fasterxml.jackson.databind.JsonNode node =
JSON_OBJECT_MAPPER.readTree(jsonString);
return node.has("zorderFields");
} catch (Exception e) {
return false;
}
```
##########
iceberg/iceberg-handler/src/main/java/org/apache/iceberg/mr/hive/udf/GenericUDFIcebergZorder.java:
##########
@@ -0,0 +1,174 @@
+/*
+ * 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.iceberg.mr.hive.udf;
+
+import java.nio.ByteBuffer;
+import java.nio.charset.StandardCharsets;
+import org.apache.hadoop.hive.ql.exec.Description;
+import org.apache.hadoop.hive.ql.exec.UDFArgumentException;
+import org.apache.hadoop.hive.ql.exec.UDFArgumentTypeException;
+import org.apache.hadoop.hive.ql.metadata.HiveException;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDF;
+import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;
+import org.apache.hadoop.hive.serde2.objectinspector.PrimitiveObjectInspector;
+import
org.apache.hadoop.hive.serde2.objectinspector.primitive.PrimitiveObjectInspectorFactory;
+import org.apache.hadoop.io.BytesWritable;
+import org.apache.iceberg.util.ZOrderByteUtils;
+
+/**
+ * Hive UDF to compute the Z-order value of given input columns using
Iceberg's ZOrderByteUtils.
+ * Supports various primitive types and converts inputs into interleaved
binary representation.
+ */
+@Description(name = "iceberg_zorder",
+ value = "_FUNC_(value) - " +
+ "Returns the z-value calculated by Iceberg ZOrderByteUtils
class")
+public class GenericUDFIcebergZorder extends GenericUDF {
+ private PrimitiveObjectInspector[] argOIs;
+ // For variable-length types (e.g., strings), how many bytes contribute to
z-order
+ private final int varLengthContribution = 8;
+ private transient ByteBuffer[] reUseBuffer;
+ private static final int MAX_OUTPUT_SIZE = Integer.MAX_VALUE;
+
+ /**
+ * Initializes the UDF, validating argument types are primitives and
preparing buffers.
+ */
+ @Override
+ public ObjectInspector initialize(ObjectInspector[] arguments) throws
UDFArgumentException {
+ if (arguments.length < 2) {
+ throw new UDFArgumentException("iceberg_zorder requires at least 2
arguments");
Review Comment:
The minimum argument check should allow for 1 argument since Z-order can be
applied to a single column. Change the condition to `arguments.length < 1`.
```suggestion
if (arguments.length < 1) {
throw new UDFArgumentException("iceberg_zorder requires at least 1
argument");
```
##########
iceberg/iceberg-handler/src/main/java/org/apache/iceberg/mr/hive/udf/GenericUDFIcebergZorder.java:
##########
@@ -0,0 +1,174 @@
+/*
+ * 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.iceberg.mr.hive.udf;
+
+import java.nio.ByteBuffer;
+import java.nio.charset.StandardCharsets;
+import org.apache.hadoop.hive.ql.exec.Description;
+import org.apache.hadoop.hive.ql.exec.UDFArgumentException;
+import org.apache.hadoop.hive.ql.exec.UDFArgumentTypeException;
+import org.apache.hadoop.hive.ql.metadata.HiveException;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDF;
+import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;
+import org.apache.hadoop.hive.serde2.objectinspector.PrimitiveObjectInspector;
+import
org.apache.hadoop.hive.serde2.objectinspector.primitive.PrimitiveObjectInspectorFactory;
+import org.apache.hadoop.io.BytesWritable;
+import org.apache.iceberg.util.ZOrderByteUtils;
+
+/**
+ * Hive UDF to compute the Z-order value of given input columns using
Iceberg's ZOrderByteUtils.
+ * Supports various primitive types and converts inputs into interleaved
binary representation.
+ */
+@Description(name = "iceberg_zorder",
+ value = "_FUNC_(value) - " +
+ "Returns the z-value calculated by Iceberg ZOrderByteUtils
class")
+public class GenericUDFIcebergZorder extends GenericUDF {
+ private PrimitiveObjectInspector[] argOIs;
+ // For variable-length types (e.g., strings), how many bytes contribute to
z-order
+ private final int varLengthContribution = 8;
+ private transient ByteBuffer[] reUseBuffer;
+ private static final int MAX_OUTPUT_SIZE = Integer.MAX_VALUE;
+
+ /**
+ * Initializes the UDF, validating argument types are primitives and
preparing buffers.
+ */
+ @Override
+ public ObjectInspector initialize(ObjectInspector[] arguments) throws
UDFArgumentException {
+ if (arguments.length < 2) {
+ throw new UDFArgumentException("iceberg_zorder requires at least 2
arguments");
+ }
+ argOIs = new PrimitiveObjectInspector[arguments.length];
+ reUseBuffer = new ByteBuffer[arguments.length];
+ for (int i = 0; i < arguments.length; i++) {
+ if (!(arguments[i] instanceof PrimitiveObjectInspector poi)) {
+ throw new UDFArgumentTypeException(i, "Only primitive types supported
for z-order");
+ }
+ argOIs[i] = poi;
+ }
+ return PrimitiveObjectInspectorFactory.writableBinaryObjectInspector;
+ }
+
+ /**
+ * Evaluates the UDF by converting input values to ordered bytes,
interleaving them,
+ * and returning the resulting Z-order binary value.
+ */
+ @Override
+ public Object evaluate(DeferredObject[] arguments) throws HiveException {
+ byte[][] inputs = new byte[arguments.length][];
+ int totalLength = 0;
+
+ for (int i = 0; i < arguments.length; i++) {
+ byte[] orderedBytes = convertToOrderedBytes(arguments[i].get(),
argOIs[i], i);
+ inputs[i] = orderedBytes;
+ totalLength += orderedBytes.length;
+ }
+
+ int outputLength = Math.min(totalLength, MAX_OUTPUT_SIZE);
+ ByteBuffer buffer = ByteBuffer.allocate(outputLength);
+
+ byte[] interleaved = ZOrderByteUtils.interleaveBits(inputs, outputLength,
buffer);
+ return new BytesWritable(interleaved);
+ }
+
+ @Override
+ public String getDisplayString(String[] children) {
+ return "iceberg_zorder(" + String.join(", ", children) + ")";
+ }
+
+ /**
+ * Converts a single input value to its ordered byte representation based on
type.
+ * @return fixed-length byte arrays to be used in interleaving.
+ */
+ private byte[] convertToOrderedBytes(Object value, PrimitiveObjectInspector
oi,
+ int position) throws HiveException {
+ if (value == null) {
+ // For NULL values, we have primitive buffer size of 8 with values of 0
+ return ByteBuffer.wrap(new
byte[ZOrderByteUtils.PRIMITIVE_BUFFER_SIZE]).array();
+ }
Review Comment:
Creating a new ByteBuffer and byte array for each null value is inefficient.
Consider creating a static final byte array for null values to reuse across
calls.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
