[
https://issues.apache.org/jira/browse/GOBBLIN-1335?focusedWorklogId=535888&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-535888
]
ASF GitHub Bot logged work on GOBBLIN-1335:
-------------------------------------------
Author: ASF GitHub Bot
Created on: 14/Jan/21 05:23
Start Date: 14/Jan/21 05:23
Worklog Time Spent: 10m
Work Description: ZihanLi58 commented on a change in pull request #3172:
URL: https://github.com/apache/incubator-gobblin/pull/3172#discussion_r557050597
##########
File path:
gobblin-iceberg/src/main/java/org/apache/gobblin/iceberg/Utils/IcebergUtils.java
##########
@@ -0,0 +1,281 @@
+/*
+ * 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.gobblin.iceberg.Utils;
+
+import com.google.common.base.Preconditions;
+import com.google.common.collect.Lists;
+import java.io.IOException;
+import java.nio.ByteBuffer;
+import java.util.ArrayList;
+import java.util.Collections;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.concurrent.atomic.AtomicInteger;
+import java.util.stream.Collectors;
+import lombok.extern.slf4j.Slf4j;
+import org.apache.gobblin.metadata.IntegerBytesPair;
+import org.apache.gobblin.metadata.IntegerLongPair;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.fs.Path;
+import org.apache.hadoop.hive.metastore.api.FieldSchema;
+import org.apache.hadoop.hive.serde2.typeinfo.TypeInfo;
+import org.apache.hadoop.hive.serde2.typeinfo.TypeInfoUtils;
+import org.apache.iceberg.DataFile;
+import org.apache.iceberg.DataFiles;
+import org.apache.iceberg.Metrics;
+import org.apache.iceberg.PartitionSpec;
+import org.apache.iceberg.Schema;
+import org.apache.iceberg.StructLike;
+import org.apache.iceberg.avro.AvroSchemaUtil;
+import org.apache.iceberg.exceptions.RuntimeIOException;
+import org.apache.iceberg.types.Conversions;
+import org.apache.iceberg.types.TypeUtil;
+import org.apache.iceberg.types.Types;
+
+
+@Slf4j
+public class IcebergUtils {
+
+ private static final String AVRO_SCHEMA_URL = "avro.schema.url";
+ private static final String AVRO_SCHEMA_LITERAL = "avro.schema.literal";
+ private static final String[] RESTRICTED_PROPERTIES =
+ new String[]{AVRO_SCHEMA_URL, AVRO_SCHEMA_LITERAL};
+
+ private IcebergUtils() {
+ }
+ /**
+ * Calculate the {@Link PartitionSpec} used to create iceberg table
+ */
+ public static PartitionSpec getPartitionSpec(Schema tableSchema, Schema
partitionSchema) {
+ //TODO: Add more information into partition spec e.g. day, year, month,
kafka partition ids, offset ranges for better consuming
+ PartitionSpec.Builder builder = PartitionSpec.builderFor(tableSchema);
+ partitionSchema.asStruct().fields().forEach(f ->
builder.identity(f.name()));
+ return builder.build();
+ }
+
+ /**
+ * Given a avro schema string and a hive table,
+ * calculate the iceberg table schema and partition schema.
+ * (Since we use 'datepartition' as partition column, which is not included
inside the data schema,
+ * we'll need to add that column to data schema to construct table schema
+ */
+ public static IcebergDataAndPartitionSchema getIcebergSchema(String schema,
+ org.apache.hadoop.hive.metastore.api.Table table) {
+
+ org.apache.iceberg.shaded.org.apache.avro.Schema icebergDataSchema =
+ new
org.apache.iceberg.shaded.org.apache.avro.Schema.Parser().parse(schema);
+ Types.StructType dataStructType =
AvroSchemaUtil.convert(icebergDataSchema).asStructType();
+ List<Types.NestedField> dataFields =
Lists.newArrayList(dataStructType.fields());
+ org.apache.iceberg.shaded.org.apache.avro.Schema icebergPartitionSchema =
+ parseSchemaFromCols(table.getPartitionKeys(), table.getDbName(),
table.getTableName(), true);
+ Types.StructType partitionStructType =
AvroSchemaUtil.convert(icebergPartitionSchema).asStructType();
+ List<Types.NestedField> partitionFields = partitionStructType.fields();
+ Preconditions.checkArgument(partitionFields.stream().allMatch(f ->
f.type().isPrimitiveType()),
+ "Only primitive fields are supported for partition columns");
+ dataFields.addAll(partitionFields);
+ Types.StructType updatedStructType = Types.StructType.of(dataFields);
+ updatedStructType =
+ (Types.StructType) TypeUtil.assignFreshIds(updatedStructType, new
AtomicInteger(0)::incrementAndGet);
+ return new IcebergDataAndPartitionSchema(new
org.apache.iceberg.Schema(updatedStructType.fields()),
+ new org.apache.iceberg.Schema(partitionFields));
+ }
+
+ private static org.apache.iceberg.shaded.org.apache.avro.Schema
parseSchemaFromCols(List<FieldSchema> cols,
+ String namespace, String recordName, boolean mkFieldsOptional) {
+ final List<String> colNames = new ArrayList<>(cols.size());
+ final List<TypeInfo> colsTypeInfo = new ArrayList<>(cols.size());
+ cols.forEach(fs -> {
+ colNames.add(fs.getName());
+ colsTypeInfo.add(TypeInfoUtils.getTypeInfoFromTypeString(fs.getType()));
+ });
+ final TypeInfoToSchemaParser parser =
+ new TypeInfoToSchemaParser(namespace, mkFieldsOptional,
Collections.emptyMap());
+ return new org.apache.iceberg.shaded.org.apache.avro.Schema.Parser().parse(
+ parser.parseSchemaFromFieldsTypeInfo("", recordName, colNames,
colsTypeInfo).toString());
+ }
+
+ /**
+ * Given a Hive table, get all the properties of the table, and drop
unneeded ones and transfer to a map
+ */
+ public static Map<String, String>
getTableProperties(org.apache.hadoop.hive.metastore.api.Table table) {
+ final Map<String, String> parameters = getRawTableProperties(table);
+ // drop unneeded parameters
+ for (String k : RESTRICTED_PROPERTIES) {
+ parameters.remove(k);
+ }
+ return parameters;
+ }
+
+ private static Map<String, String>
getRawTableProperties(org.apache.hadoop.hive.metastore.api.Table table) {
+ final Map<String, String> parameters = new HashMap<>();
+ // lowest to highest priority of updating tableProperties
+ parameters.putAll(table.getSd().getSerdeInfo().getParameters());
+ parameters.putAll(table.getSd().getParameters());
+ parameters.putAll(table.getParameters());
+ return parameters;
+ }
+
+ /**
+ * Get the iceberg partition value for given partition strings
+ */
+ public static StructLike getPartition(Types.StructType partitionType,
List<String> partitionValues) {
+ //TODO parse partitionValue as per partitionSchema
+ return new StructLike() {
+ @Override
+ public int size() {
+ return partitionValues.size();
+ }
+
+ @Override
+ public <T> T get(int pos, Class<T> javaClass) {
+ return partitionValue(partitionType.fields().get(pos),
partitionValues.get(pos));
+ }
+
+ @Override
+ public <T> void set(int pos, T value) {
+ throw new UnsupportedOperationException();
+ }
+ };
+ }
+
+ private static <T> T partitionValue(Types.NestedField partitionField, String
colAsString) {
+ Preconditions.checkState(partitionField.type().isPrimitiveType(),
"Partition column {} is not of primitive type",
+ partitionField);
+ return (T) Conversions.fromPartitionString(partitionField.type(),
colAsString);
+ }
+
+ /**
+ * Transfer list of {@Link IntegerLongPair} from origin id to long, to
Map<Integer, Long> from real column id to long
+ * This method is mainly used to get parse the file metrics from GMCE
+ * @param list list of {@Link IntegerLongPair}
+ * @param schemaIdMap A map from origin ID (defined by data pipeline) to the
real iceberg table column id
+ * @return A map from real id to long as the file metrics
+ */
+ public static Map<Integer, Long> getMapFromIntegerLongPairs(
+ List<IntegerLongPair> list, Map<Integer, Integer> schemaIdMap) {
+ //If schemaIdMap is not set, we directly return null to avoid set wrong
file metrics
+ if (list == null || list.size() == 0 || schemaIdMap == null) {
+ return null;
+ }
+ try {
+ return list.stream().collect(Collectors.toMap(t ->
schemaIdMap.get(t.getKey()), IntegerLongPair::getValue));
+ } catch (Exception e) {
+ return null;
+ }
+ }
+
+ /**
+ * Transfer list of {@Link IntegerBytesPair} from origin id to bytes, to
Map<Integer, ByteBuffer> from real column id to ByteBuffer
+ * This method is mainly used to get parse the file metrics from GMCE
+ * @param list list of {@Link IntegerBytesPair} from origin id to bytes
+ * @param schemaIdMap A map from origin ID (defined by data pipeline) to the
real iceberg table column id
+ * @return A map from real id to ByteBuffer as the file metrics
+ */
+ public static Map<Integer, ByteBuffer> getMapFromIntegerBytesPairs(
+ List<IntegerBytesPair> list, Map<Integer, Integer> schemaIdMap) {
+ //If schemaWithOriginId is not set, we directly return null to avoid set
wrong file metrics
+ if (list == null || list.size() == 0 || schemaIdMap == null) {
+ return null;
+ }
+ try {
+ return list.stream().collect(Collectors.toMap(t ->
schemaIdMap.get(t.getKey()), IntegerBytesPair::getValue));
+ } catch (Exception e) {
+ return null;
+ }
+ }
+
+ /**
+ * Method to get DataFile without format and metrics information
+ * This method is mainly used to get the file to be deleted
+ */
+ public static DataFile getIcebergDataFileWithoutMetric(String file,
PartitionSpec partitionSpec,
+ StructLike partition) {
+ //Use raw Path to support federation.
+ String rawPath = new Path(file).toUri().getRawPath();
+ //Just want to remove the old files, so set the record number and file
size to a random value
+ DataFiles.Builder dataFileBuilder =
+
DataFiles.builder(partitionSpec).withPath(rawPath).withFileSizeInBytes(0).withRecordCount(0);
+
+ if (partition != null) {
+ dataFileBuilder.withPartition(partition);
+ }
+ return dataFileBuilder.build();
+ }
+
+ /**
+ * Method to get DataFile with format and metrics information
+ * This method is mainly used to get the file to be added
+ */
+ public static DataFile
getIcebergDataFileWithMetric(org.apache.gobblin.metadata.DataFile file,
Review comment:
DataFile here refers to 'org.apache.iceberg.DataFile', and the parameter
is 'org.apache.gobblin.metadata.DataFile', how should I keep consistent with
other files? Not sure how other files are dealing with this type of import
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 535888)
Time Spent: 2h (was: 1h 50m)
> Publish GMCE(GobblinMetadataChangeEvent) publisher and iceberg retention job
> to Gobblin OSS
> -------------------------------------------------------------------------------------------
>
> Key: GOBBLIN-1335
> URL: https://issues.apache.org/jira/browse/GOBBLIN-1335
> Project: Apache Gobblin
> Issue Type: Task
> Reporter: Zihan Li
> Priority: Major
> Time Spent: 2h
> Remaining Estimate: 0h
>
> Publish GMCE() publisher and iceberg retention job to Gobblin OSS, which
> contains the file information for added/deleted/rewrited files, and will be
> then used for iceberg registration.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
