guilload commented on a change in pull request #1103:
URL: https://github.com/apache/iceberg/pull/1103#discussion_r439002574



##########
File path: mr/src/main/java/org/apache/iceberg/mr/mapred/IcebergSerDe.java
##########
@@ -0,0 +1,129 @@
+/*
+ * 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.mapred;
+
+import java.io.IOException;
+import java.io.UncheckedIOException;
+import java.sql.Date;
+import java.sql.Timestamp;
+import java.time.LocalDate;
+import java.time.LocalDateTime;
+import java.time.LocalTime;
+import java.time.OffsetDateTime;
+import java.util.ArrayList;
+import java.util.Collections;
+import java.util.List;
+import java.util.Properties;
+import javax.annotation.Nullable;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hive.serde2.AbstractSerDe;
+import org.apache.hadoop.hive.serde2.SerDeException;
+import org.apache.hadoop.hive.serde2.SerDeStats;
+import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;
+import org.apache.hadoop.io.Writable;
+import org.apache.iceberg.Schema;
+import org.apache.iceberg.SnapshotsTable;
+import org.apache.iceberg.Table;
+import org.apache.iceberg.types.Type;
+import org.apache.iceberg.types.Types;
+
+public class IcebergSerDe extends AbstractSerDe {
+
+  private Schema schema;
+  private ObjectInspector inspector;
+  private List<Object> row;
+
+  @Override
+  public void initialize(@Nullable Configuration configuration, Properties 
serDeProperties) throws SerDeException {
+    Table table = null;
+    try {
+      table = TableResolver.resolveTableFromConfiguration(configuration, 
serDeProperties);
+    } catch (IOException e) {
+      throw new UncheckedIOException("Unable to resolve table from 
configuration: ", e);
+    }
+    this.schema = table.schema();
+    if (table instanceof SnapshotsTable) {
+      try {
+        this.inspector = new 
IcebergObjectInspectorGenerator().createObjectInspector(schema);
+      } catch (Exception e) {
+        throw new SerDeException(e);
+      }
+    } else {
+      List<Types.NestedField> columns = new ArrayList<>(schema.columns());
+      columns.add(Types.NestedField.optional(Integer.MAX_VALUE,
+          SystemTableUtil.snapshotIdVirtualColumnName(serDeProperties), 
Types.LongType.get()));
+      Schema withVirtualColumn = new Schema(columns);
+      try {
+        this.inspector = new 
IcebergObjectInspectorGenerator().createObjectInspector(withVirtualColumn);
+      } catch (Exception e) {
+        throw new SerDeException(e);
+      }
+    }
+  }
+
+  @Override
+  public Class<? extends Writable> getSerializedClass() {
+    return null;
+  }
+
+  @Override
+  public Writable serialize(Object o, ObjectInspector objectInspector) {
+    return null;
+  }
+
+  @Override
+  public SerDeStats getSerDeStats() {
+    return null;
+  }
+
+  @Override
+  public Object deserialize(Writable writable) {

Review comment:
       There's actually no need to return anything other the `Record` object. 
Hive relies on the `ObjectInspector` class to retrieve values from arbitrary 
data structures and convert those values to the classes Hive expects.
   
   You may want to take a look at 
https://cwiki.apache.org/confluence/display/Hive/DeveloperGuide#DeveloperGuide-ObjectInspector
 and then read the implementation of a few object inspectors in the Hive 
codebase.
   
   So all we have to do here is return the `Record` object and ensure we have 
the right object inspectors for the `Record` class and other data types that 
are not represented in the same fashion in Iceberg vs. Hive.
   
   As a matter of fact, I started working on an object inspector for the 
`Record` class last year. Why don't I submit a PR on top off this one to get 
you started?

##########
File path: 
mr/src/main/java/org/apache/iceberg/mr/mapred/IcebergSchemaToTypeInfo.java
##########
@@ -0,0 +1,114 @@
+/*
+ * 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.mapred;
+
+import java.util.ArrayList;
+import java.util.List;
+import org.apache.hadoop.hive.serde.serdeConstants;
+import org.apache.hadoop.hive.serde2.SerDeException;
+import org.apache.hadoop.hive.serde2.typeinfo.HiveDecimalUtils;
+import org.apache.hadoop.hive.serde2.typeinfo.TypeInfo;
+import org.apache.hadoop.hive.serde2.typeinfo.TypeInfoFactory;
+import org.apache.iceberg.Schema;
+import org.apache.iceberg.relocated.com.google.common.collect.ImmutableMap;
+import org.apache.iceberg.types.Type;
+import org.apache.iceberg.types.Types;
+
+/**
+ * Class to convert Iceberg types to Hive TypeInfo
+ */
+final class IcebergSchemaToTypeInfo {
+
+  private IcebergSchemaToTypeInfo() {
+
+  }
+
+  private static final ImmutableMap<Object, Object> primitiveTypeToTypeInfo = 
ImmutableMap.builder()
+      .put(Types.BooleanType.get(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.BOOLEAN_TYPE_NAME))
+      .put(Types.IntegerType.get(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.INT_TYPE_NAME))
+      .put(Types.LongType.get(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.BIGINT_TYPE_NAME))
+      .put(Types.FloatType.get(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.FLOAT_TYPE_NAME))
+      .put(Types.DoubleType.get(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.DOUBLE_TYPE_NAME))
+      .put(Types.BinaryType.get(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.BINARY_TYPE_NAME))
+      .put(Types.StringType.get(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.STRING_TYPE_NAME))
+      .put(Types.DateType.get(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.DATE_TYPE_NAME))
+      .put(Types.TimestampType.withoutZone(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.BIGINT_TYPE_NAME))
+      .put(Types.TimestampType.withZone(), 
TypeInfoFactory.getPrimitiveTypeInfo(serdeConstants.BIGINT_TYPE_NAME))
+      .build();
+
+  public static List<TypeInfo> getColumnTypes(Schema schema) throws Exception {
+    List<Types.NestedField> fields = schema.columns();
+    List<TypeInfo> types = new ArrayList<>(fields.size());
+    for (Types.NestedField field : fields) {
+      types.add(generateTypeInfo(field.type()));
+    }
+    return types;
+  }
+
+  private static TypeInfo generateTypeInfo(Type type) throws Exception {

Review comment:
       I already have a type visitor somewhere from `Schema` to 
`ObjectInspector`. I can also submit that one on my PR so you can focus on the 
remaining things to do.

##########
File path: mr/src/main/java/org/apache/iceberg/mr/mapred/IcebergSerDe.java
##########
@@ -0,0 +1,129 @@
+/*
+ * 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.mapred;
+
+import java.io.IOException;
+import java.io.UncheckedIOException;
+import java.sql.Date;
+import java.sql.Timestamp;
+import java.time.LocalDate;
+import java.time.LocalDateTime;
+import java.time.LocalTime;
+import java.time.OffsetDateTime;
+import java.util.ArrayList;
+import java.util.Collections;
+import java.util.List;
+import java.util.Properties;
+import javax.annotation.Nullable;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hive.serde2.AbstractSerDe;
+import org.apache.hadoop.hive.serde2.SerDeException;
+import org.apache.hadoop.hive.serde2.SerDeStats;
+import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;
+import org.apache.hadoop.io.Writable;
+import org.apache.iceberg.Schema;
+import org.apache.iceberg.SnapshotsTable;
+import org.apache.iceberg.Table;
+import org.apache.iceberg.types.Type;
+import org.apache.iceberg.types.Types;
+
+public class IcebergSerDe extends AbstractSerDe {
+
+  private Schema schema;
+  private ObjectInspector inspector;
+  private List<Object> row;
+
+  @Override
+  public void initialize(@Nullable Configuration configuration, Properties 
serDeProperties) throws SerDeException {
+    Table table = null;
+    try {
+      table = TableResolver.resolveTableFromConfiguration(configuration, 
serDeProperties);
+    } catch (IOException e) {
+      throw new UncheckedIOException("Unable to resolve table from 
configuration: ", e);
+    }
+    this.schema = table.schema();
+    if (table instanceof SnapshotsTable) {
+      try {
+        this.inspector = new 
IcebergObjectInspectorGenerator().createObjectInspector(schema);
+      } catch (Exception e) {
+        throw new SerDeException(e);
+      }
+    } else {
+      List<Types.NestedField> columns = new ArrayList<>(schema.columns());
+      columns.add(Types.NestedField.optional(Integer.MAX_VALUE,
+          SystemTableUtil.snapshotIdVirtualColumnName(serDeProperties), 
Types.LongType.get()));
+      Schema withVirtualColumn = new Schema(columns);
+      try {
+        this.inspector = new 
IcebergObjectInspectorGenerator().createObjectInspector(withVirtualColumn);
+      } catch (Exception e) {
+        throw new SerDeException(e);
+      }
+    }
+  }
+
+  @Override
+  public Class<? extends Writable> getSerializedClass() {
+    return null;
+  }
+
+  @Override
+  public Writable serialize(Object o, ObjectInspector objectInspector) {
+    return null;
+  }
+
+  @Override
+  public SerDeStats getSerDeStats() {
+    return null;
+  }
+
+  @Override
+  public Object deserialize(Writable writable) {
+    IcebergWritable icebergWritable = (IcebergWritable) writable;
+    List<Types.NestedField> fields = icebergWritable.schema().columns();
+
+    if (row == null || row.size() != fields.size()) {
+      row = new ArrayList<Object>(fields.size());
+    } else {
+      row.clear();
+    }
+    for (int i = 0; i < fields.size(); i++) {
+      Object obj = ((IcebergWritable) writable).record().get(i);
+      Type fieldType = fields.get(i).type();
+      if (fieldType.equals(Types.DateType.get())) {
+        row.add(Date.valueOf((LocalDate) obj));

Review comment:
       See this is where object inspectors come in handy, they'll take care of 
the conversion for us. Plus, you can clearly see that this current approach 
won't work for nested data structures.




----------------------------------------------------------------
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]



---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to