apache · lukasnalezenec · Apr 28, 2015 · Apr 28, 2015 · julienledem · May 1, 2015
diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
@@ -27,4 +27,4 @@ Here are a few tips to get your contribution in:
   4. Make sure that your code passes the unit tests. You can run the tests with `mvn test` in the root directory. 
   5. Add new unit tests for your code. 
 
-If you’d like to report a bug but don’t have time to fix it, you can still post it to our [issue tracker](https://issues.apache.org/jira/browse/PARQUET), or email the mailing list (dev@parquet.incubator.apache.org).
+If you’d like to report a bug but don’t have time to fix it, you can still post it to our [issue tracker](https://issues.apache.org/jira/browse/PARQUET), or email the mailing list (dev@parquet.apache.org).
diff --git a/NOTICE b/NOTICE
@@ -1,5 +1,5 @@
 
-Apache Parquet MR (Incubating)
+Apache Parquet MR
 Copyright 2014 The Apache Software Foundation
 
 This product includes software developed at

diff --git a/dev/README.md b/dev/README.md
@@ -26,8 +26,8 @@ Merging a pull request requires being a committer on the project.
 * How to merge a Pull request:
 have an apache and apache-github remote setup
 ```
-git remote add apache-github git@github.com:apache/incubator-parquet-mr.git
-git remote add apache https://git-wip-us.apache.org/repos/asf/incubator-parquet-mr.git
+git remote add apache-github git@github.com:apache/parquet-mr.git
+git remote add apache https://git-wip-us.apache.org/repos/asf/parquet-mr.git
 ```
 run the following command
 ```
@@ -42,19 +42,19 @@ example output:
 ```
 Which pull request would you like to merge? (e.g. 34):
 ```
-Type the pull request number (from https://github.com/apache/incubator-parquet-mr/pulls) and hit enter.
+Type the pull request number (from https://github.com/apache/parquet-mr/pulls) and hit enter.
 ```
 === Pull Request #X ===
 title	Blah Blah Blah
 source	repo/branch
 target	master
-url	https://api.github.com/repos/apache/incubator-parquet-mr/pulls/X
+url	https://api.github.com/repos/apache/parquet-mr/pulls/X
 
 Proceed with merging pull request #3? (y/n): 
 ```
 If this looks good, type y and hit enter.
 ```
-From git-wip-us.apache.org:/repos/asf/incubator-parquet-mr.git
+From git-wip-us.apache.org:/repos/asf/parquet-mr.git
  * [new branch]      master     -> PR_TOOL_MERGE_PR_3_MASTER
 Switched to branch 'PR_TOOL_MERGE_PR_3_MASTER'
 
@@ -68,7 +68,7 @@ Delta compression using up to 4 threads.
 Compressing objects: 100% (26/26), done.
 Writing objects: 100% (36/36), 5.32 KiB, done.
 Total 36 (delta 17), reused 0 (delta 0)
-To git-wip-us.apache.org:/repos/asf/incubator-parquet-mr.git
+To git-wip-us.apache.org:/repos/asf/parquet-mr.git
    b767ac4..485658a  PR_TOOL_MERGE_PR_X_MASTER -> master
 Restoring head pointer to b767ac4e
 Note: checking out 'b767ac4e'.

diff --git a/dev/source-release.sh b/dev/source-release.sh
@@ -28,7 +28,7 @@ if [ -z "$2" ]; then
   exit
 fi
 
-version=$1-incubating
+version=$1
 rc=$2
 
 if [ -d tmp/ ]; then
@@ -68,7 +68,7 @@ svn co --depth=empty https://dist.apache.org/repos/dist/dev/incubator/parquet tm
 mkdir -p tmp/$tagrc
 cp ${tarball}* tmp/$tagrc
 svn add tmp/$tagrc
-echo "svn ci -m 'Apache Parquet MR (Incubating) $version RC${rc}' tmp/$tagrc"
+echo "svn ci -m 'Apache Parquet MR $version RC${rc}' tmp/$tagrc"
 
 # clean up
 rm -rf tmp

diff --git a/parquet-avro/pom.xml b/parquet-avro/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-avro</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Avro (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Avro</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
     <avro.version>1.7.6</avro.version>

diff --git a/parquet-cascading/pom.xml b/parquet-cascading/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-cascading</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Cascading (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Cascading</name>
+  <url>https://parquet.apache.org</url>
 
   <repositories>
     <repository>

diff --git a/parquet-column/pom.xml b/parquet-column/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-column</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Column (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Column</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
   </properties>

diff --git a/parquet-common/pom.xml b/parquet-common/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-common</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Common (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Common</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
   </properties>

diff --git a/parquet-encoding/pom.xml b/parquet-encoding/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-encoding</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Encodings (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Encodings</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
   </properties>

diff --git a/parquet-generator/pom.xml b/parquet-generator/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-generator</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Generator (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Generator</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
   </properties>

diff --git a/parquet-hadoop-bundle/pom.xml b/parquet-hadoop-bundle/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-hadoop-bundle</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Hadoop Bundle (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Hadoop Bundle</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
   </properties>

diff --git a/parquet-hadoop/pom.xml b/parquet-hadoop/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-hadoop</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Hadoop (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Hadoop</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
   </properties>

diff --git a/parquet-hive-bundle/pom.xml b/parquet-hive-bundle/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-hive-bundle</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Hive Bundle (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Hive Bundle</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
   </properties>

diff --git a/parquet-hive/parquet-hive-binding/parquet-hive-0.10-binding/pom.xml b/parquet-hive/parquet-hive-binding/parquet-hive-0.10-binding/pom.xml
@@ -27,7 +27,7 @@
   <modelVersion>4.0.0</modelVersion>
 
   <artifactId>parquet-hive-0.10-binding</artifactId>
-  <name>Apache Parquet Hive 0.10 Binding (Incubating)</name>
+  <name>Apache Parquet Hive 0.10 Binding</name>
   <packaging>jar</packaging>
 
   <properties>

diff --git a/parquet-hive/parquet-hive-binding/parquet-hive-0.12-binding/pom.xml b/parquet-hive/parquet-hive-binding/parquet-hive-0.12-binding/pom.xml
@@ -27,7 +27,7 @@
   <modelVersion>4.0.0</modelVersion>
 
   <artifactId>parquet-hive-0.12-binding</artifactId>
-  <name>Apache Parquet Hive 0.12 Binding (Incubating)</name>
+  <name>Apache Parquet Hive 0.12 Binding</name>
   <packaging>jar</packaging>
 
   <properties>

diff --git a/parquet-hive/parquet-hive-binding/parquet-hive-binding-bundle/pom.xml b/parquet-hive/parquet-hive-binding/parquet-hive-binding-bundle/pom.xml
@@ -29,7 +29,7 @@
   <modelVersion>4.0.0</modelVersion>
 
   <artifactId>parquet-hive-binding-bundle</artifactId>
-  <name>Apache Parquet Hive Binding Bundle (Incubating)</name>
+  <name>Apache Parquet Hive Binding Bundle</name>
   <packaging>jar</packaging>
 
   <dependencies>

diff --git a/parquet-hive/parquet-hive-binding/parquet-hive-binding-factory/pom.xml b/parquet-hive/parquet-hive-binding/parquet-hive-binding-factory/pom.xml
@@ -27,7 +27,7 @@
   <modelVersion>4.0.0</modelVersion>
 
   <artifactId>parquet-hive-binding-factory</artifactId>
-  <name>Apache Parquet Hive Binding Factory (Incubating)</name>
+  <name>Apache Parquet Hive Binding Factory</name>
   <packaging>jar</packaging>
 
   <dependencies>

diff --git a/parquet-hive/parquet-hive-binding/parquet-hive-binding-interface/pom.xml b/parquet-hive/parquet-hive-binding/parquet-hive-binding-interface/pom.xml
@@ -27,7 +27,7 @@
   <modelVersion>4.0.0</modelVersion>
 
   <artifactId>parquet-hive-binding-interface</artifactId>
-  <name>Apache Parquet Hive Binding Interface (Incubating)</name>
+  <name>Apache Parquet Hive Binding Interface</name>
   <packaging>jar</packaging>
 
   <dependencies>

diff --git a/parquet-hive/parquet-hive-binding/pom.xml b/parquet-hive/parquet-hive-binding/pom.xml
@@ -27,7 +27,7 @@
   <modelVersion>4.0.0</modelVersion>
 
   <artifactId>parquet-hive-binding</artifactId>
-  <name>Apache Parquet Hive Binding Parent (Incubating)</name>
+  <name>Apache Parquet Hive Binding Parent</name>
   <packaging>pom</packaging>
 
   <!-- provides javax.jdo:jdo2-api:jar:2.3-ec -->

diff --git a/parquet-hive/pom.xml b/parquet-hive/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-hive</artifactId>
   <packaging>pom</packaging>
 
-  <name>Apache Parquet Hive (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Hive</name>
+  <url>https://parquet.apache.org</url>
 
   <developers>
     <developer>

diff --git a/parquet-jackson/pom.xml b/parquet-jackson/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-jackson</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Jackson (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Jackson</name>
+  <url>https://parquet.apache.org</url>
 
   <dependencies>
     <dependency>

diff --git a/parquet-pig-bundle/pom.xml b/parquet-pig-bundle/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-pig-bundle</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Pig Bundle (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Pig Bundle</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
   </properties>

diff --git a/parquet-pig/pom.xml b/parquet-pig/pom.xml
@@ -29,8 +29,8 @@
   <artifactId>parquet-pig</artifactId>
   <packaging>jar</packaging>
 
-  <name>Apache Parquet Pig (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Pig</name>
+  <url>https://parquet.apache.org</url>
 
   <properties>
   </properties>

diff --git a/parquet-protobuf/pom.xml b/parquet-protobuf/pom.xml
@@ -35,8 +35,8 @@
   </properties>
 
 
-  <name>Apache Parquet Protobuf (Incubating)</name>
-  <url>https://parquet.incubator.apache.org</url>
+  <name>Apache Parquet Protobuf</name>
+  <url>https://parquet.apache.org</url>
 
   <dependencies>
     <dependency>

diff --git a/parquet-protobuf/src/main/java/org/apache/parquet/proto/ProtoReadSupport.java b/parquet-protobuf/src/main/java/org/apache/parquet/proto/ProtoReadSupport.java
@@ -46,6 +46,16 @@ public static void setRequestedProjection(Configuration configuration, String re
     configuration.set(PB_REQUESTED_PROJECTION, requestedProjection);
   }
 
+  /**
+   * Set name of protobuf class to be used for reading data.
+   * If no class is set, value from file header is used.
+   * Note that the value in header is present only if the file was written
+   * using parquet-protobuf project, it will fail otherwise.
+   * */
+  public static void setProtobufClass(Configuration configuration, String protobufClass) {
+    configuration.set(PB_CLASS, protobufClass);
+  }
+
   @Override
   public ReadContext init(InitContext context) {
     String requestedProjectionString = context.getConfiguration().get(PB_REQUESTED_PROJECTION);
@@ -63,16 +73,22 @@ public ReadContext init(InitContext context) {
 
   @Override
   public RecordMaterializer<T> prepareForRead(Configuration configuration, Map<String, String> keyValueMetaData, MessageType fileSchema, ReadContext readContext) {
-    String strProtoClass = keyValueMetaData.get(PB_CLASS);
+    String headerProtoClass = keyValueMetaData.get(PB_CLASS);
+    String configuredProtoClass = configuration.get(PB_CLASS);
+
+    if (configuredProtoClass != null) {
+      LOG.debug("Replacing class " + headerProtoClass + " by " + configuredProtoClass);
+      headerProtoClass = configuredProtoClass;
+    }
 
-    if (strProtoClass == null) {
+    if (headerProtoClass == null) {
       throw new RuntimeException("I Need parameter " + PB_CLASS + " with Protocol Buffer class");
     }
 
-    LOG.debug("Reading data with Protocol Buffer class" + strProtoClass);
+    LOG.debug("Reading data with Protocol Buffer class " + headerProtoClass);
 
     MessageType requestedSchema = readContext.getRequestedSchema();
-    Class<? extends Message> protobufClass = Protobufs.getProtobufClass(strProtoClass);
+    Class<? extends Message> protobufClass = Protobufs.getProtobufClass(headerProtoClass);
     return new ProtoRecordMaterializer(requestedSchema, protobufClass);
   }
 

diff --git a/parquet-protobuf/src/main/resources/META-INF/NOTICE b/parquet-protobuf/src/main/resources/META-INF/NOTICE
@@ -1,5 +1,5 @@
 
-Apache Parquet MR (Incubating)
+Apache Parquet MR
 Copyright 2014 The Apache Software Foundation
 
 This product includes software developed at

diff --git a/parquet-protobuf/src/test/java/org/apache/parquet/proto/ProtoInputOutputFormatTest.java b/parquet-protobuf/src/test/java/org/apache/parquet/proto/ProtoInputOutputFormatTest.java
@@ -19,24 +19,21 @@
 package org.apache.parquet.proto;
 
 import com.google.protobuf.Message;
-import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.Path;
-import org.junit.Test;
-import org.apache.parquet.Log;
+import org.apache.parquet.proto.test.TestProtobuf;
+import org.apache.parquet.proto.test.TestProtobuf.FirstCustomClassMessage;
+import org.apache.parquet.proto.test.TestProtobuf.SecondCustomClassMessage;
 import org.apache.parquet.proto.utils.ReadUsingMR;
 import org.apache.parquet.proto.utils.WriteUsingMR;
-import org.apache.parquet.proto.test.TestProtobuf;
+import org.junit.Test;
 
 import java.util.List;
 
-import static org.junit.Assert.assertEquals;
-import static org.junit.Assert.assertTrue;
+import static org.junit.Assert.*;
 
 
 public class ProtoInputOutputFormatTest {
 
-  private static final Log LOG = Log.getLog(ProtoInputOutputFormatTest.class);
-
   /**
    * Writes Protocol Buffer using first MR job, reads written file using
    * second job and compares input and output.
@@ -95,6 +92,34 @@ public void testProjection() throws Exception {
     assertTrue("Found data outside projection.", readDocument.getNameCount() == 0);
   }
 
+  /**
+   * When user specified protobuffer class in configuration,
+   * It should replace class specified in header.
+   * */
+  @Test
+  public void testCustomProtoClass() throws Exception {
+    FirstCustomClassMessage.Builder inputMessage;
+    inputMessage = FirstCustomClassMessage.newBuilder();
+    inputMessage.setString("writtenString");
+
+    Path outputPath = new WriteUsingMR().write(new Message[]{inputMessage.build()});
+    ReadUsingMR readUsingMR = new ReadUsingMR();
+    String customClass = SecondCustomClassMessage.class.getName();
+    ProtoReadSupport.setProtobufClass(readUsingMR.getConfiguration(), customClass);
+    List<Message> result = readUsingMR.read(outputPath);
+
+    assertEquals(1, result.size());
+    Message msg = result.get(0);
+    assertFalse("Class from header returned.",
+            msg instanceof FirstCustomClassMessage);
+    assertTrue("Custom class was not used",
+            msg instanceof SecondCustomClassMessage);
+
+    String stringValue;
+    stringValue = ((SecondCustomClassMessage) msg).getString();
+    assertEquals("writtenString", stringValue);
+  }
+
   /**
    * Runs job that writes input to file and then job reading data back.
    */