Github user chinmaykolhatkar commented on a diff in the pull request:
https://github.com/apache/incubator-apex-malhar/pull/219#discussion_r58046072
--- Diff:
contrib/src/main/java/com/datatorrent/contrib/parquet/ParquetFilePOJOReader.java
---
@@ -0,0 +1,318 @@
+/**
+ * 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 com.datatorrent.contrib.parquet;
+
+import java.lang.reflect.Field;
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Map;
+import java.util.StringTokenizer;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import org.apache.commons.lang3.ClassUtils;
+
+import com.google.common.collect.Lists;
+import com.google.common.collect.Maps;
+
+import com.datatorrent.api.Context;
+import com.datatorrent.api.DefaultOutputPort;
+import com.datatorrent.api.annotation.OutputPortFieldAnnotation;
+import com.datatorrent.lib.util.FieldInfo;
+import com.datatorrent.lib.util.FieldInfo.SupportType;
+import com.datatorrent.lib.util.PojoUtils;
+import com.datatorrent.lib.util.PojoUtils.Setter;
+
+import parquet.example.data.Group;
+import parquet.io.InvalidRecordException;
+import parquet.io.ParquetEncodingException;
+import parquet.schema.PrimitiveType.PrimitiveTypeName;
+
+/**
+ * <p>
+ * ParquetFilePOJOReader
+ * </p>
+ * ParquetFilePOJOReader operator is a concrete implementation of
+ * AbstractParquetFileReader to read Parquet files and emit records as
POJOs.The
+ * POJO class name & field mapping should be provided by the user. If this
+ * mapping is not provided then reflection is used to determine this
mapping. As
+ * of now only primitive types ( INT32, INT64, BOOLEAN, FLOAT, DOUBLE,
BINARY )
+ * are supported.
+ *
+ * Parquet primitive types maps to POJO datatypes as follows : INT32 -> int
+ * INT64 -> long BOOLEAN -> boolean FLOAT -> float DOUBLE -> double BINARY
->
+ * String
+ *
+ * @displayName ParquetFilePOJOReader
+ * @tags parquet,input adapter
+ * @since 3.3.0
+ */
+public class ParquetFilePOJOReader extends
AbstractParquetFileReader<Object>
+{
+
+ /**
+ * POJO class
+ */
+ protected transient Class<?> pojoClass;
+ /**
+ * Map containing setters for fields in POJO
+ */
+ protected transient Map<String, Setter> pojoSetters;
+ /**
+ * String representing Parquet TO POJO field mapping. If not provided,
then
+ * reflection is used to determine the mapping. Format :
+ * PARQUET_FIELD_NAME:POJO_FIELD_NAME:TYPE
+ * E.g.event_id:event_id_v2:INTEGER,org_id:org_id_v2:STRING,long_id:
+ * long_id_v2:
+ *
LONG,css_file_loaded:css_file_loaded_v2:BOOLEAN,float_val:float_val_v2:
+ * FLOAT,double_val:double_val_v2:DOUBLE
+ */
+ protected transient String groupToPOJOFieldsMapping = null;
+ protected transient List<FieldInfo> fieldInfos;
+ protected transient List<ActiveFieldInfo> columnFieldSetters = null;
+ protected static final String FIELD_SEPARATOR = ":";
+ protected static final String RECORD_SEPARATOR = ",";
+ private static final Logger logger =
LoggerFactory.getLogger(ParquetFilePOJOReader.class);
+
+ @OutputPortFieldAnnotation(schemaRequired = true)
+ public final transient DefaultOutputPort<Object> output = new
DefaultOutputPort<Object>()
+ {
+
+ @Override
+ public void setup(Context.PortContext context)
+ {
+ pojoClass = context.getValue(Context.PortContext.TUPLE_CLASS);
+ pojoSetters = Maps.newHashMap();
+ for (Field f : pojoClass.getDeclaredFields()) {
+ try {
+ pojoSetters.put(f.getName(), generateSettersForField(pojoClass,
f.getName()));
+ } catch (NoSuchFieldException | SecurityException e) {
+ logger.error("Failed to generate setters. Exception {}", e);
+ throw new RuntimeException(e);
+ }
+ }
+
+ if (groupToPOJOFieldsMapping == null) {
+ fieldInfos = createFieldInfoMap(generateFieldInfoInputs());
+ } else {
+ fieldInfos = createFieldInfoMap(groupToPOJOFieldsMapping);
+ }
+ initColumnFieldSetters();
+ }
+
+ };
+
+ /**
+ * Converts Group to POJO
+ */
+ @Override
+ protected Object convertGroup(Group group)
+ {
+ Object obj;
+ try {
+ obj = pojoClass.newInstance();
+ } catch (InstantiationException | IllegalAccessException ex) {
+ throw new RuntimeException(ex);
+ }
+ for (int i = 0; i < columnFieldSetters.size(); i++) {
+ try {
+ ParquetFilePOJOReader.ActiveFieldInfo afi =
columnFieldSetters.get(i);
+ int fieldIndex =
schema.getFieldIndex(afi.fieldInfo.getColumnName());
--- End diff --
fieldIndex can be useful for instead of setter map.. Maps are costly to
retrieve data from.
If you could have an array where index is fieldIndex and value at that
index is corresponding setter method, the access would be much faster. This
would improve the processing speed of a tuple.
For eg. PojoUtils.Setters[] pojoSetters;
and while accessing you could do:
pojoSetters[fieldIndex].set(obj, value);
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
