Package org.apache.avro.mapred

Examples of org.apache.avro.mapred.FsInput

65
66
67
68
69
70
71
72
73
74
75
  @Override
  public Iterator<T> read(FileSystem fs, final Path path) {
    this.mapFn.initialize();
    try {
      FsInput fsi = new FsInput(path, fs.getConf());
      final DataFileReader<T> reader = new DataFileReader<T>(fsi, recordReader);
      return new AutoClosingIterator<T>(reader, new UnmodifiableIterator<T>() {
        @Override
        public boolean hasNext() {
          return reader.hasNext();
View Full Code Here
72
73
74
75
76
77
78
79
80
81
    if(latest != null) {
      gdr.setExpected(latest);
    }

    this.reader = new DataFileReader<GenericRecord>(new FsInput(split.getPath(), job), gdr);
    this.reader.sync(split.getStart());
    this.start = reader.tell();
    this.stop = split.getStart() + split.getLength();
  }
View Full Code Here
358
359
360
361
362
363
364
365
366
367
368
        if (fstat == null || fstat.length == 0) {
          throw new IllegalArgumentException("No valid files found in directory: " + path);
        }
        path = fstat[0].getPath();
      }
      reader = new DataFileReader(new FsInput(path, conf), new GenericDatumReader<GenericRecord>());
      return reader.getSchema();
    } catch (IOException e) {
      throw new RuntimeException("Error reading schema from path: "  + path, e);
    } finally {
      if (reader != null) {
View Full Code Here
142
143
144
145
146
147
148
149
150
151
      throws IOException {
    Configuration conf = getConf();
    if (!BaseSqoopTestCase.isOnPhysicalCluster()) {
      conf.set(CommonArgs.FS_DEFAULT_NAME, CommonArgs.LOCAL_FS);
    }
    FsInput fsInput = new FsInput(filename, conf);
    DatumReader<GenericRecord> datumReader =
      new GenericDatumReader<GenericRecord>();
    return new DataFileReader<GenericRecord>(fsInput, datumReader);
  }
View Full Code Here
223
224
225
226
227
228
229
230
      Schema recordSchema = AvroKeyValue.getSchema(mKeySchema, options.getValueSchema());
      DatumReader<GenericRecord> datumReader =
        model.createDatumReader(recordSchema);
      mDataFileReader =
        new DataFileReader<GenericRecord>
        (new FsInput(dataFilePath, options.getConfiguration()), datumReader);
     
    }
View Full Code Here
305
306
307
308
309
310
311
312
313
314
315
    private <K> NavigableMap<K, Long> loadIndexFile(
        Configuration conf, Path path, Schema keySchema) throws IOException {
      DatumReader<GenericRecord> datumReader = model.createDatumReader(
          AvroKeyValue.getSchema(keySchema, Schema.create(Schema.Type.LONG)));
      DataFileReader<GenericRecord> fileReader = new DataFileReader<GenericRecord>(
          new FsInput(path, conf), datumReader);

      NavigableMap<K, Long> index;
      if (Schema.create(Schema.Type.STRING).equals(keySchema)) {
        // Because Avro STRING types are mapped to the Java CharSequence class that does not
        // mandate the implementation of Comparable, we need to specify a special
View Full Code Here
203
204
205
206
207
208
209
210
211
212
213
    verify(context);

  Configuration conf = new Configuration();
  conf.set("fs.default.name", "file:///");
  Path avroFile = new Path("target/temp.avro");
  DataFileReader<GenericData.Record> avroFileReader = new DataFileReader<GenericData.Record>(new FsInput(avroFile,
      conf), new SpecificDatumReader<GenericData.Record>());
   
 
  avroFileReader.seek(pointTwo);
    // Verify that the second record was written;
View Full Code Here
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
    // Check that the results from the MapReduce were as expected.
    FileSystem fileSystem = FileSystem.get(job.getConfiguration());
    FileStatus[] outputFiles = fileSystem.globStatus(outputPath.suffix("/myavro-r-00000.avro"));
    Assert.assertEquals(1, outputFiles.length);
    DataFileReader<GenericData.Record> reader = new DataFileReader<GenericData.Record>(
        new FsInput(outputFiles[0].getPath(), job.getConfiguration()),
        new GenericDatumReader<GenericData.Record>(STATS_SCHEMA));
    Map<String, Integer> counts = new HashMap<String, Integer>();
    for (GenericData.Record record : reader) {
      counts.put(((Utf8) record.get("name")).toString(), (Integer) record.get("count"));
    }
    reader.close();

    Assert.assertEquals(3, counts.get("apple").intValue());
    Assert.assertEquals(2, counts.get("banana").intValue());
    Assert.assertEquals(1, counts.get("carrot").intValue());

    outputFiles = fileSystem.globStatus(outputPath.suffix("/myavro1-r-00000.avro"));
    Assert.assertEquals(1, outputFiles.length);
    reader = new DataFileReader<GenericData.Record>(
        new FsInput(outputFiles[0].getPath(), job.getConfiguration()),
        new GenericDatumReader<GenericData.Record>(STATS_SCHEMA_2));
    counts = new HashMap<String, Integer>();
    for (GenericData.Record record : reader) {
      counts.put(((Utf8) record.get("name1")).toString(), (Integer) record.get("count1"));
    }
    reader.close();

    Assert.assertEquals(3, counts.get("apple").intValue());
    Assert.assertEquals(2, counts.get("banana").intValue());
    Assert.assertEquals(1, counts.get("carrot").intValue());
 
    outputFiles = fileSystem.globStatus(outputPath.suffix("/testnewwrite-r-00000.avro"));
    Assert.assertEquals(1, outputFiles.length);
    reader = new DataFileReader<GenericData.Record>(
        new FsInput(outputFiles[0].getPath(), job.getConfiguration()),
            new GenericDatumReader<GenericData.Record>(STATS_SCHEMA));
    counts = new HashMap<String, Integer>();
    for (GenericData.Record record : reader) {
       counts.put(((Utf8) record.get("name")).toString(), (Integer) record.get("count"));
    }
    reader.close();
   
    Assert.assertEquals(3, counts.get("apple").intValue());
    Assert.assertEquals(2, counts.get("banana").intValue());
    Assert.assertEquals(1, counts.get("carrot").intValue());
       
    outputFiles = fileSystem.globStatus(outputPath.suffix("/testnewwrite2-r-00000.avro"));
    Assert.assertEquals(1, outputFiles.length);
    reader = new DataFileReader<GenericData.Record>(
        new FsInput(outputFiles[0].getPath(), job.getConfiguration()),
        new GenericDatumReader<GenericData.Record>(STATS_SCHEMA_2));
    counts = new HashMap<String, Integer>();
    for (GenericData.Record record : reader) {
     counts.put(((Utf8) record.get("name1")).toString(), (Integer) record.get("count1"));
    }
    reader.close();
    Assert.assertEquals(3, counts.get("apple").intValue());
    Assert.assertEquals(2, counts.get("banana").intValue());
    Assert.assertEquals(1, counts.get("carrot").intValue());
   
    outputFiles = fileSystem.globStatus(outputPath.suffix("/testwritenonschema-r-00000.avro"));
    Assert.assertEquals(1, outputFiles.length);
    reader = new DataFileReader<GenericData.Record>(
        new FsInput(outputFiles[0].getPath(), job.getConfiguration()),
        new GenericDatumReader<GenericData.Record>(STATS_SCHEMA));
    counts = new HashMap<String, Integer>();
    for (GenericData.Record record : reader) {
      counts.put(((Utf8) record.get("name")).toString(), (Integer) record.get("count"));
    }
View Full Code Here
318
319
320
321
322
323
324
325
326
327
328
    Assert.assertTrue(job.waitForCompletion(true));
    FileSystem fileSystem = FileSystem.get(job.getConfiguration());
    FileStatus[] outputFiles = fileSystem.globStatus(outputPath.suffix("/myavro3-*"));
    Assert.assertEquals(1, outputFiles.length);
    DataFileReader<TextStats> reader = new DataFileReader<TextStats>(
        new FsInput(outputFiles[0].getPath(), job.getConfiguration()),
        new SpecificDatumReader<TextStats>());
    Map<String, Integer> counts = new HashMap<String, Integer>();
    for (TextStats record : reader) {
      counts.put(record.name.toString(), record.count);
    }
View Full Code Here
360
361
362
363
364
365
366
367
368
369
370
    // Check that the results from the MapReduce were as expected.
    FileSystem fileSystem = FileSystem.get(job.getConfiguration());
    FileStatus[] outputFiles = fileSystem.globStatus(outputPath.suffix("/myavro3-*"));
    Assert.assertEquals(1, outputFiles.length);
    DataFileReader<TextStats> reader = new DataFileReader<TextStats>(
        new FsInput(outputFiles[0].getPath(), job.getConfiguration()),
        new SpecificDatumReader<TextStats>());
    Map<String, Integer> counts = new HashMap<String, Integer>();
    for (TextStats record : reader) {
      counts.put(record.name.toString(), record.count);
    }
View Full Code Here
TOP

Related Classes of org.apache.avro.mapred.FsInput

  • com.cloudera.sqoop.TestAvroImport
  • com.cloudera.sqoop.testutil.LobAvroImportTestCase
  • com.datasalt.pangool.tuplemr.avro.AvroRecordReader
  • com.datasalt.pangool.tuplemr.mapred.lib.input.TupleInputFormat$TupleInputReader
  • com.linkedin.camus.sweeper.mapreduce.CamusSweeperAvroKeyJob
  • com.linkedin.haivvreo.AvroGenericRecordReader
  • com.linkedin.pig.FastAvroRecordReader
  • MaxWidgetIdGenericAvro
  • org.apache.avro.hadoop.file.SortedKeyValueFile$Reader
  • org.apache.avro.hadoop.file.TestSortedKeyValueFile
    • 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.