[parquet-thrift] start removing things marked for deprecation

Author: dossett

parquet-thrift/pom.xml

@@ -120,29 +120,11 @@
       <type>test-jar</type>
       <scope>test</scope>
     </dependency>
-    <dependency>
-      <groupId>org.apache.parquet</groupId>
-      <artifactId>parquet-pig</artifactId>
-      <version>1.15.0</version>
-    </dependency>
-    <dependency>
-      <groupId>org.apache.pig</groupId>
-      <artifactId>pig</artifactId>
-      <version>${pig.version}</version>
-      <classifier>${pig.classifier}</classifier>
-      <scope>provided</scope>
-    </dependency>
     <dependency>
       <groupId>javax.annotation</groupId>
       <artifactId>javax.annotation-api</artifactId>
       <version>${javax.annotation.version}</version>
     </dependency>
-    <dependency> <!-- for pig runtime in tests -->
-      <groupId>org.antlr</groupId>
-      <artifactId>antlr-runtime</artifactId>
-      <version>3.5.3</version>
-      <scope>test</scope>
-    </dependency>
     <dependency>
       <groupId>org.apache.thrift</groupId>
       <artifactId>libthrift</artifactId>
@@ -160,13 +142,6 @@
       <version>${slf4j.version}</version>
       <scope>test</scope>
     </dependency>
-    <dependency>
-      <!-- needed for Pig tests -->
-      <groupId>log4j</groupId>
-      <artifactId>log4j</artifactId>
-      <version>1.2.17</version>
-      <scope>test</scope>
-    </dependency>
     <dependency>
       <groupId>org.apache.parquet</groupId>
       <artifactId>parquet-hadoop</artifactId>

parquet-thrift/src/main/java/org/apache/parquet/hadoop/thrift/AbstractThriftWriteSupport.java

@@ -15,7 +15,6 @@
  */
 package org.apache.parquet.hadoop.thrift;
 
-import com.twitter.elephantbird.pig.util.ThriftToPig;
 import java.util.Map;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.parquet.conf.HadoopParquetConfiguration;
@@ -25,13 +24,11 @@
 import org.apache.parquet.io.ColumnIOFactory;
 import org.apache.parquet.io.MessageColumnIO;
 import org.apache.parquet.io.api.RecordConsumer;
-import org.apache.parquet.pig.PigMetaData;
 import org.apache.parquet.schema.MessageType;
 import org.apache.parquet.thrift.ParquetWriteProtocol;
 import org.apache.parquet.thrift.ThriftMetaData;
 import org.apache.parquet.thrift.ThriftSchemaConverter;
 import org.apache.parquet.thrift.struct.ThriftType.StructType;
-import org.apache.thrift.TBase;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
@@ -99,26 +96,10 @@ protected void init(Class<T> thriftClass) {
 
     final Map<String, String> extraMetaData =
         new ThriftMetaData(thriftClass.getName(), thriftStruct).toExtraMetaData();
-    // adding the Pig schema as it would have been mapped from thrift
-    // TODO: make this work for non-tbase types
-    if (isPigLoaded() && TBase.class.isAssignableFrom(thriftClass)) {
-      new PigMetaData(new ThriftToPig((Class<? extends TBase<?, ?>>) thriftClass).toSchema())
-          .addToMetaData(extraMetaData);
-    }
 
     this.writeContext = new WriteContext(schema, extraMetaData);
   }
 
-  protected boolean isPigLoaded() {
-    try {
-      Class.forName("org.apache.pig.impl.logicalLayer.schema.Schema");
-      return true;
-    } catch (ClassNotFoundException e) {
-      LOG.info("Pig is not loaded, pig metadata will not be written");
-      return false;
-    }
-  }
-
   @Override
   public WriteContext init(Configuration configuration) {
     return init(new HadoopParquetConfiguration(configuration));

parquet-thrift/src/main/java/org/apache/parquet/hadoop/thrift/ParquetThriftBytesOutputFormat.java

@@ -72,23 +72,4 @@ public ParquetThriftBytesOutputFormat(
       FieldIgnoredHandler errorHandler) {
     super(new ThriftBytesWriteSupport(configuration, protocolFactory, thriftClass, buffered, errorHandler));
   }
-
-  /**
-   * @param protocolFactory the protocol factory to use to read the bytes
-   * @param thriftClass     thriftClass the class to extract the schema from
-   * @param buffered        whether we should buffer each record
-   * @param errorHandler    handle record corruption and schema incompatible exception
-   * @deprecated Use @link{ParquetThriftBytesOutputFormat(
-   * Configuration configuration, TProtocolFactory protocolFactory,
-   * {@literal Class<\? extends TBase<\?, ?>>} thriftClass, boolean buffered,
-   * FieldIgnoredHandler errorHandler)} instead.
-   */
-  @Deprecated
-  public ParquetThriftBytesOutputFormat(
-      TProtocolFactory protocolFactory,
-      Class<? extends TBase<?, ?>> thriftClass,
-      boolean buffered,
-      FieldIgnoredHandler errorHandler) {
-    this(new Configuration(), protocolFactory, thriftClass, buffered, errorHandler);
-  }
 }

parquet-thrift/src/main/java/org/apache/parquet/hadoop/thrift/ParquetThriftOutputFormat.java

@@ -29,14 +29,14 @@
 public class ParquetThriftOutputFormat<T extends TBase<?, ?>> extends ParquetOutputFormat<T> {
 
   public static void setThriftClass(Job job, Class<? extends TBase<?, ?>> thriftClass) {
-    ThriftWriteSupport.setThriftClass(ContextUtil.getConfiguration(job), thriftClass);
+    TBaseWriteSupport.setThriftClass(ContextUtil.getConfiguration(job), thriftClass);
   }
 
   public static Class<? extends TBase<?, ?>> getThriftClass(Job job) {
-    return ThriftWriteSupport.getThriftClass(ContextUtil.getConfiguration(job));
+    return TBaseWriteSupport.getThriftClass(ContextUtil.getConfiguration(job));
   }
 
   public ParquetThriftOutputFormat() {
-    super(new ThriftWriteSupport<T>());
+    super(new TBaseWriteSupport<T>());
   }
 }

parquet-thrift/src/main/java/org/apache/parquet/hadoop/thrift/ThriftBytesWriteSupport.java

@@ -99,20 +99,6 @@ public ThriftBytesWriteSupport() {
     this.errorHandler = null;
   }
 
-  /**
-   * @deprecated Use @link{ThriftBytesWriteSupport(Configuration configuration,
-   * TProtocolFactory protocolFactory, {@literal Class<? extends TBase<?,?>>} thriftClass,
-   * boolean buffered, FieldIgnoredHandler errorHandler)} instead
-   */
-  @Deprecated
-  public ThriftBytesWriteSupport(
-      TProtocolFactory protocolFactory,
-      Class<? extends TBase<?, ?>> thriftClass,
-      boolean buffered,
-      FieldIgnoredHandler errorHandler) {
-    this(new Configuration(), protocolFactory, thriftClass, buffered, errorHandler);
-  }
-
   public ThriftBytesWriteSupport(
       Configuration configuration,
       TProtocolFactory protocolFactory,

parquet-thrift/src/main/java/org/apache/parquet/hadoop/thrift/ThriftReadSupport.java

@@ -23,7 +23,6 @@
 import java.util.Map;
 import java.util.Set;
 import org.apache.hadoop.conf.Configuration;
-import org.apache.hadoop.mapred.JobConf;
 import org.apache.parquet.Strings;
 import org.apache.parquet.conf.HadoopParquetConfiguration;
 import org.apache.parquet.conf.ParquetConfiguration;
@@ -39,7 +38,6 @@
 import org.apache.parquet.thrift.projection.FieldProjectionFilter;
 import org.apache.parquet.thrift.projection.StrictFieldProjectionFilter;
 import org.apache.parquet.thrift.projection.ThriftProjectionException;
-import org.apache.parquet.thrift.projection.deprecated.DeprecatedFieldProjectionFilter;
 import org.apache.parquet.thrift.struct.ThriftType.StructType;
 import org.apache.thrift.TBase;
 import org.apache.thrift.protocol.TProtocol;
@@ -49,13 +47,6 @@
 public class ThriftReadSupport<T> extends ReadSupport<T> {
   private static final Logger LOG = LoggerFactory.getLogger(ThriftReadSupport.class);
 
-  /**
-   * Deprecated. Use {@link #STRICT_THRIFT_COLUMN_FILTER_KEY}
-   * Accepts a ";" delimited list of globs in the syntax implemented by {@link DeprecatedFieldProjectionFilter}
-   */
-  @Deprecated
-  public static final String THRIFT_COLUMN_FILTER_KEY = "parquet.thrift.column.filter";
-
   /**
    * Accepts a ";" delimited list of glob paths, in the syntax implemented by {@link StrictFieldProjectionFilter}
    */
@@ -73,20 +64,6 @@ public class ThriftReadSupport<T> extends ReadSupport<T> {
 
   protected Class<T> thriftClass;
 
-  /**
-   * A {@link ThriftRecordConverter} builds an object by working with {@link TProtocol}. The default
-   * implementation creates standard Apache Thrift {@link TBase} objects; to support alternatives, such
-   * as <a href="http://github.com/twitter/scrooge">Twiter's Scrooge</a>, a custom converter can be specified.
-   *
-   * @param conf  a mapred jobconf
-   * @param klass a thrift class
-   * @deprecated use {@link #setRecordConverterClass(Configuration, Class)} below
-   */
-  @Deprecated
-  public static void setRecordConverterClass(JobConf conf, Class<?> klass) {
-    setRecordConverterClass((Configuration) conf, klass);
-  }
-
   /**
    * A {@link ThriftRecordConverter} builds an object by working with {@link TProtocol}. The default
    * implementation creates standard Apache Thrift {@link TBase} objects; to support alternatives, such
@@ -99,11 +76,6 @@ public static void setRecordConverterClass(Configuration conf, Class<?> klass) {
     conf.set(RECORD_CONVERTER_CLASS_KEY, klass.getName());
   }
 
-  @Deprecated
-  public static void setProjectionPushdown(JobConf jobConf, String projectionString) {
-    jobConf.set(THRIFT_COLUMN_FILTER_KEY, projectionString);
-  }
-
   public static void setStrictFieldProjectionFilter(Configuration conf, String semicolonDelimitedGlobs) {
     conf.set(STRICT_THRIFT_COLUMN_FILTER_KEY, semicolonDelimitedGlobs);
   }
@@ -113,31 +85,12 @@ public static FieldProjectionFilter getFieldProjectionFilter(Configuration conf)
   }
 
   public static FieldProjectionFilter getFieldProjectionFilter(ParquetConfiguration conf) {
-    String deprecated = conf.get(THRIFT_COLUMN_FILTER_KEY);
     String strict = conf.get(STRICT_THRIFT_COLUMN_FILTER_KEY);
 
-    if (Strings.isNullOrEmpty(deprecated) && Strings.isNullOrEmpty(strict)) {
+    if (Strings.isNullOrEmpty(strict)) {
       return null;
     }
 
-    if (!Strings.isNullOrEmpty(deprecated) && !Strings.isNullOrEmpty(strict)) {
-      throw new ThriftProjectionException("You cannot provide both "
-          + THRIFT_COLUMN_FILTER_KEY
-          + " and "
-          + STRICT_THRIFT_COLUMN_FILTER_KEY
-          + "! "
-          + THRIFT_COLUMN_FILTER_KEY
-          + " is deprecated.");
-    }
-
-    if (!Strings.isNullOrEmpty(deprecated)) {
-      LOG.warn(
-          "Using {} is deprecated. Please see the docs for {}!",
-          THRIFT_COLUMN_FILTER_KEY,
-          STRICT_THRIFT_COLUMN_FILTER_KEY);
-      return new DeprecatedFieldProjectionFilter(deprecated);
-    }
-
     return StrictFieldProjectionFilter.fromSemicolonDelimitedString(strict);
   }
 
@@ -166,8 +119,8 @@ public org.apache.parquet.hadoop.api.ReadSupport.ReadContext init(InitContext co
     if (partialSchemaString != null && projectionFilter != null) {
       throw new ThriftProjectionException(String.format(
           "You cannot provide both a partial schema and field projection filter."
-              + "Only one of (%s, %s, %s) should be set.",
-          PARQUET_READ_SCHEMA, STRICT_THRIFT_COLUMN_FILTER_KEY, THRIFT_COLUMN_FILTER_KEY));
+              + "Only one of (%s, %s) should be set.",
+          PARQUET_READ_SCHEMA, STRICT_THRIFT_COLUMN_FILTER_KEY));
     }
 
     // set requestedProjections only when it's specified
@@ -197,13 +150,6 @@ protected MessageType getProjectedSchema(
         .convert((Class<TBase<?, ?>>) thriftClass);
   }
 
-  @Deprecated
-  @SuppressWarnings("unchecked")
-  protected MessageType getProjectedSchema(FieldProjectionFilter fieldProjectionFilter) {
-    return new ThriftSchemaConverter(new Configuration(), fieldProjectionFilter)
-        .convert((Class<TBase<?, ?>>) thriftClass);
-  }
-
   private void initThriftClassFromMultipleFiles(Map<String, Set<String>> fileMetadata, Configuration conf)
       throws ClassNotFoundException {
     initThriftClassFromMultipleFiles(fileMetadata, new HadoopParquetConfiguration(conf));

parquet-thrift/src/main/java/org/apache/parquet/hadoop/thrift/ThriftWriteSupport.java

@@ -30,19 +30,6 @@
 
 public class TBaseRecordConverter<T extends TBase<?, ?>> extends ThriftRecordConverter<T> {
 
-  /**
-   * This is for compatibility only.
-   *
-   * @param thriftClass            a thrift class
-   * @param requestedParquetSchema the requested Parquet schema
-   * @param thriftType             the thrift type
-   * @deprecated will be removed in 2.x
-   */
-  @Deprecated
-  public TBaseRecordConverter(final Class<T> thriftClass, MessageType requestedParquetSchema, StructType thriftType) {
-    this(thriftClass, requestedParquetSchema, thriftType, (HadoopParquetConfiguration) null);
-  }
-
   @SuppressWarnings("unused")
   public TBaseRecordConverter(
       final Class<T> thriftClass, MessageType requestedParquetSchema, StructType thriftType, Configuration conf) {