@Test
public void testViewUris() throws IOException {
PartitionStrategy partitionStrategy =
new PartitionStrategy.Builder().hash("username", 2).build();
Dataset<Record> inputDataset =
repo.create(
"ns",
"in",
new DatasetDescriptor.Builder()
.schema(USER_SCHEMA)
.partitionStrategy(partitionStrategy)
.build());
Dataset<Record> outputDataset =
repo.create(
"ns",
"out",
new DatasetDescriptor.Builder()
.schema(USER_SCHEMA)
.partitionStrategy(partitionStrategy)
.build());
writeTestUsers(inputDataset, 10);
URI sourceViewUri =
new URIBuilder(repo.getUri(), "ns", "in").with("username", "test-0").build();
View<Record> inputView = Datasets.<Record, Dataset<Record>>load(sourceViewUri, Record.class);
Assert.assertEquals(1, datasetSize(inputView));
Pipeline pipeline = new MRPipeline(TestCrunchDatasets.class);
PCollection<GenericData.Record> data =
pipeline.read(CrunchDatasets.asSource(sourceViewUri, GenericData.Record.class));
URI targetViewUri = new URIBuilder(repo.getUri(), "ns", "out").with("email", "email-0").build();
pipeline.write(data, CrunchDatasets.asTarget(targetViewUri), Target.WriteMode.APPEND);
pipeline.run();
Assert.assertEquals(1, datasetSize(outputDataset));
}
/**
* List the {@link Dataset} URIs in the repository identified by the URI.
*
* <p>URI formats are defined by {@code Dataset} implementations. The repository URIs you pass to
* this method must begin with {@code repo:}. For example, to list the {@code Dataset} URIs for
* the Hive repository, provide the URI {@code repo:hive}.
*
* @param uri a {@code DatasetRepository} URI
* @return the URIs present in the {@code DatasetRepository}
* @throws NullPointerException if {@code uri} is null
* @throws IllegalArgumentException if {@code uri} is not a repository URI
*/
public static Collection<URI> list(URI uri) {
boolean isRepo = URIBuilder.REPO_SCHEME.equals(uri.getScheme());
Preconditions.checkArgument(isRepo, "Not a repository URI: " + uri);
DatasetRepository repo = Registration.open(URI.create(uri.getRawSchemeSpecificPart()));
// build a URI for each dataset name
URI repoUri = repo.getUri();
List<URI> datasets = Lists.newArrayList();
for (String namespace : repo.namespaces()) {
for (String dataset : repo.datasets(namespace)) {
datasets.add(new URIBuilder(repoUri, namespace, dataset).build());
}
}
return datasets;
}