Package parquet.schema

Examples of parquet.schema.MessageType

367
368
369
370
371
372
373
  }

  MessageType fromParquetSchema(List<SchemaElement> schema) {
    Iterator<SchemaElement> iterator = schema.iterator();
    SchemaElement root = iterator.next();
    return new MessageType(root.getName(), convertChildren(iterator, root.getNum_children()));
  }
View Full Code Here
30
31
32
33
34
35
36
37
  @Override
  public parquet.hadoop.api.ReadSupport.ReadContext init(
      Configuration configuration, Map<String, String> keyValueMetaData,
      MessageType fileSchema) {
    String partialSchemaString = configuration.get(ReadSupport.PARQUET_READ_SCHEMA);
    MessageType requestedProjection = getSchemaForRead(fileSchema, partialSchemaString);
    return new ReadContext(requestedProjection);
  }
View Full Code Here
410
411
412
413
414
415
416
417
            for (HiveColumnHandle column : columns) {
                if (!column.isPartitionKey() && column.getHiveColumnIndex() < messageType.getFieldCount()) {
                    fields.add(messageType.getType(column.getName()));
                }
            }
            MessageType requestedProjection = new MessageType(messageType.getName(), fields.build());
            return new ReadContext(requestedProjection);
        }
View Full Code Here
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
    this.entry = entry;
    setColumns(columns);
  }

  public static MessageType getProjection(MessageType schema, Collection<SchemaPath> columns) {
    MessageType projection = null;
    for (SchemaPath path : columns) {
      List<String> segments = Lists.newArrayList();
      PathSegment rootSegment = path.getRootSegment();
      PathSegment seg = rootSegment;
      String messageName = schema.getName();
      while(seg != null){
        if(seg.isNamed()) {
          segments.add(seg.getNameSegment().getPath());
        }
        seg = seg.getChild();
      }
      String[] pathSegments = new String[segments.size()];
      segments.toArray(pathSegments);
      Type type = null;
      try {
        type = schema.getType(pathSegments);
      } catch (InvalidRecordException e) {
        logger.warn("Invalid record" , e);
      }
      if (type != null) {
        Type t = getType(pathSegments, 0, schema);
        if (projection == null) {
          projection = new MessageType(messageName, t);
        } else {
          projection = projection.union(new MessageType(messageName, t));
        }
      }
    }
    return projection;
  }
View Full Code Here
131
132
133
134
135
136
137
138
139
140
141
  @Override
  public void setup(OutputMutator output) throws ExecutionSetupException {

    try {
      schema = footer.getFileMetaData().getSchema();
      MessageType projection = null;

      if (isStarQuery()) {
        projection = schema;
      } else {
        projection = getProjection(schema, getColumns());
View Full Code Here
123
124
125
126
127
128
129
130
131
132
133
  private void newSchema() throws IOException {
    List<Type> types = Lists.newArrayList();
    for (MaterializedField field : batchSchema) {
      types.add(getType(field));
    }
    schema = new MessageType("root", types);

    Path fileName = new Path(location, prefix + "_" + index + ".parquet");
    parquetFileWriter = new ParquetFileWriter(conf, schema, fileName);
    parquetFileWriter.start();
View Full Code Here
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
    return s;
  }

  private static boolean isComplex(ParquetMetadata footer) {
    MessageType schema = footer.getFileMetaData().getSchema();

    for (Type type : schema.getFields()) {
      if (!type.isPrimitive()) {
        return true;
      }
    }
    for (ColumnDescriptor col : schema.getColumns()) {
      if (col.getMaxRepetitionLevel() > 0) {
        return true;
      }
    }
    return false;
View Full Code Here
TOP

Related Classes of parquet.schema.MessageType

  • com.asakusafw.directio.hive.parquet.DataModelMaterializer
  • com.asakusafw.directio.hive.parquet.DataModelWriteSupport
  • com.facebook.presto.hive.ParquetHiveRecordCursor$PrestoReadSupport
  • org.apache.drill.exec.store.parquet.ParquetRecordWriter
  • org.apache.drill.exec.store.parquet.ParquetScanBatchCreator
  • org.apache.drill.exec.store.parquet.TestFileGenerator
  • org.apache.drill.exec.store.parquet2.DrillParquetReader
  • org.apache.drill.exec.store.ParquetRecordReaderTest
  • org.apache.drill.ParquetSchemaMerge
  • org.apache.hadoop.hive.ql.io.parquet.convert.HiveSchemaConverter
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.