Java Code Examples for org.kitesdk.data.URIBuilder#build()

The following examples show how to use org.kitesdk.data.URIBuilder#build() . You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.
Example 1
Source File: TestCompactCommandCluster.java    From kite with Apache License 2.0 5 votes vote down vote up 
@Before
public void createDatasets() throws Exception {
  repoUri = "hdfs://" + getDFS().getUri().getAuthority() + "/tmp/data";
  TestUtil.run("delete", unpartitioned, "-r", repoUri, "-d", "target/data");

  File csvFile = temp.newFile("users.csv");
  csvFile.delete();
  String csv = csvFile.toString();
  BufferedWriter writer = Files.newWriter(
      csvFile, CSVSchemaCommand.SCHEMA_CHARSET);

  writer.append("id,username,email\n");
  numRecords = 30;
  for(int i = 0; i < numRecords; i++) {
    writer.append(i+",test"+i+",test"+i+"@example.com\n");
  }
  writer.close();

  TestUtil.run("-v", "csv-schema", csv, "-o", avsc, "--class", "User");
  TestUtil.run("create", unpartitioned, "-s", avsc,
      "-r", repoUri, "-d", "target/data");

  URI dsUri = URIBuilder.build("repo:" + repoUri, "default", partitioned);
  Datasets.<Object, Dataset<Object>>create(dsUri, new DatasetDescriptor.Builder()
      .partitionStrategy(new PartitionStrategy.Builder()
          .hash("id", 2)
          .build())
      .schema(SchemaBuilder.record("User").fields()
          .requiredLong("id")
          .optionalString("username")
          .optionalString("email")
          .endRecord())
      .build(), Object.class);


  TestUtil.run("csv-import", csv, unpartitioned, "-r", repoUri, "-d", "target/data");
  TestUtil.run("csv-import", csv, partitioned, "-r", repoUri, "-d", "target/data");
}
Example 2
Source File: TestTransformCommandCluster.java    From kite with Apache License 2.0 5 votes vote down vote up 
@Test
@SuppressWarnings("unchecked")
public void testPartitionedCopyWithNumWriters() throws Exception {
  command.repoURI = repoUri;
  command.numWriters = 3;
  command.datasets = Lists.newArrayList(source, "dest_partitioned");
  URI dsUri = URIBuilder.build("repo:" + repoUri, "default", "dest_partitioned");
  Datasets.<Object, Dataset<Object>>create(dsUri, new DatasetDescriptor.Builder()
      .partitionStrategy(new PartitionStrategy.Builder()
          .hash("id", 2)
          .build())
      .schema(SchemaBuilder.record("User").fields()
          .requiredLong("id")
          .optionalString("username")
          .optionalString("email")
          .endRecord())
      .build(), Object.class);

  int rc = command.run();
  Assert.assertEquals("Should return success", 0, rc);

  DatasetRepository repo = DatasetRepositories.repositoryFor("repo:" + repoUri);
  FileSystemDataset<GenericData.Record> ds =
      (FileSystemDataset<GenericData.Record>) repo.<GenericData.Record>
          load("default", "dest_partitioned");
  int size = DatasetTestUtilities.datasetSize(ds);
  Assert.assertEquals("Should contain copied records", 6, size);

  Assert.assertEquals("Should produce 2 partitions",
      2, Iterators.size(ds.pathIterator()));

  verify(console).info("Added {} records to \"{}\"", 6l, "dest_partitioned");
  verifyNoMoreInteractions(console);
}
Example 3
Source File: AbstractRefinableView.java    From kite with Apache License 2.0 5 votes vote down vote up 
@Override
public URI getUri() {
  URIBuilder builder = new URIBuilder(dataset.getUri());
  for (Map.Entry<String, String> entry : constraints.toQueryMap().entrySet()) {
    builder.with(entry.getKey(), entry.getValue());
  }
  return builder.build();
}