Github user paul-rogers commented on a diff in the pull request:
https://github.com/apache/drill/pull/1027#discussion_r150081972
--- Diff:
contrib/storage-kafka/src/main/java/org/apache/drill/exec/store/kafka/KafkaRecordReader.java
---
@@ -0,0 +1,178 @@
+/**
+ * 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.drill.exec.store.kafka;
+
+import static
org.apache.drill.exec.store.kafka.DrillKafkaConfig.DRILL_KAFKA_POLL_TIMEOUT;
+
+import java.util.Collection;
+import java.util.Iterator;
+import java.util.List;
+import java.util.Set;
+import java.util.concurrent.TimeUnit;
+
+import org.apache.drill.common.exceptions.ExecutionSetupException;
+import org.apache.drill.common.expression.SchemaPath;
+import org.apache.drill.exec.ExecConstants;
+import org.apache.drill.exec.ops.FragmentContext;
+import org.apache.drill.exec.ops.OperatorContext;
+import org.apache.drill.exec.physical.impl.OutputMutator;
+import org.apache.drill.exec.store.AbstractRecordReader;
+import org.apache.drill.exec.store.kafka.KafkaSubScan.KafkaSubScanSpec;
+import org.apache.drill.exec.store.kafka.decoders.MessageReader;
+import org.apache.drill.exec.store.kafka.decoders.MessageReaderFactory;
+import org.apache.drill.exec.util.Utilities;
+import org.apache.drill.exec.vector.complex.impl.VectorContainerWriter;
+import org.apache.kafka.clients.consumer.ConsumerRecord;
+import org.apache.kafka.clients.consumer.ConsumerRecords;
+import org.apache.kafka.clients.consumer.KafkaConsumer;
+import org.apache.kafka.common.TopicPartition;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import com.google.common.base.Stopwatch;
+import com.google.common.collect.Lists;
+import com.google.common.collect.Sets;
+public class KafkaRecordReader extends AbstractRecordReader {
+ private static final Logger logger =
LoggerFactory.getLogger(KafkaRecordReader.class);
+ public static final long DEFAULT_MESSAGES_PER_BATCH = 4000;
+
+ private VectorContainerWriter writer;
+ private MessageReader messageReader;
+
+ private boolean unionEnabled;
+ private KafkaConsumer<byte[], byte[]> kafkaConsumer;
+ private KafkaStoragePlugin plugin;
+ private KafkaSubScanSpec subScanSpec;
+ private long kafkaPollTimeOut;
+ private long endOffset;
+
+ private long currentOffset;
+ private long totalFetchTime = 0;
+
+ private List<TopicPartition> partitions;
+ private final boolean enableAllTextMode;
+ private final boolean readNumbersAsDouble;
+
+ private Iterator<ConsumerRecord<byte[], byte[]>> messageIter;
+
+ public KafkaRecordReader(KafkaSubScan.KafkaSubScanSpec subScanSpec,
List<SchemaPath> projectedColumns,
+ FragmentContext context, KafkaStoragePlugin plugin) {
+ setColumns(projectedColumns);
+ this.enableAllTextMode =
context.getOptions().getOption(ExecConstants.KAFKA_ALL_TEXT_MODE).bool_val;
+ this.readNumbersAsDouble = context.getOptions()
+
.getOption(ExecConstants.KAFKA_READER_READ_NUMBERS_AS_DOUBLE).bool_val;
+ this.unionEnabled =
context.getOptions().getOption(ExecConstants.ENABLE_UNION_TYPE);
+ this.plugin = plugin;
+ this.subScanSpec = subScanSpec;
+ this.endOffset = subScanSpec.getEndOffset();
+ this.kafkaPollTimeOut =
Long.valueOf(plugin.getConfig().getDrillKafkaProps().getProperty(DRILL_KAFKA_POLL_TIMEOUT));
+ }
+
+ @Override
+ protected Collection<SchemaPath> transformColumns(Collection<SchemaPath>
projectedColumns) {
--- End diff --
Preparing columns is really quite difficult with many cases to handle.
Drill appears to allow projection of the form `a.b`, `a.c` which means that `a`
is a map and we wish to project just `b` and `c` from `a`.
As it turns out, there is an active project to replace the ad-hoc
projection logic in each reader with a common, complete implementation. That
work may be ready by Drill 1.13, so whatever is done here is temporary.
Finally, the code checks for a star query and includes the star column.
But, a star query occurs only when the one and only field is the wildcard
(`*`). So, the transform for a star query would seem to be a no-op.
Finally, a `SELECT COUNT(*)` query will provide an empty projection list,
which this code must handle, if it does not already do so.
---
