Examples with TableSchema com.google.cloud.bigquery.storage.v1.TableSchema used on opensource projects

Example 6 with TableSchema

use of com.google.cloud.bigquery.storage.v1.TableSchema in project beam by apache.

the class BigQueryStorageSourceBase method split.

@Override
public List<BigQueryStorageStreamSource<T>> split(long desiredBundleSizeBytes, PipelineOptions options) throws Exception {
    BigQueryOptions bqOptions = options.as(BigQueryOptions.class);
    Table targetTable = getTargetTable(bqOptions);
    ReadSession.Builder readSessionBuilder = ReadSession.newBuilder();
    if (targetTable != null) {
        readSessionBuilder.setTable(BigQueryHelpers.toTableResourceName(targetTable.getTableReference()));
    } else {
        // If the table does not exist targetTable will be null.
        // Construct the table id if we can generate it. For error recording/logging.
        @Nullable String tableReferenceId = getTargetTableId(bqOptions);
        if (tableReferenceId != null) {
            readSessionBuilder.setTable(tableReferenceId);
        }
    }
    if (selectedFieldsProvider != null || rowRestrictionProvider != null) {
        ReadSession.TableReadOptions.Builder tableReadOptionsBuilder = ReadSession.TableReadOptions.newBuilder();
        if (selectedFieldsProvider != null) {
            tableReadOptionsBuilder.addAllSelectedFields(selectedFieldsProvider.get());
        }
        if (rowRestrictionProvider != null) {
            tableReadOptionsBuilder.setRowRestriction(rowRestrictionProvider.get());
        }
        readSessionBuilder.setReadOptions(tableReadOptionsBuilder);
    }
    if (format != null) {
        readSessionBuilder.setDataFormat(format);
    }
    int streamCount = 0;
    if (desiredBundleSizeBytes > 0) {
        long tableSizeBytes = (targetTable != null) ? targetTable.getNumBytes() : 0;
        streamCount = (int) Math.min(tableSizeBytes / desiredBundleSizeBytes, MAX_SPLIT_COUNT);
    }
    streamCount = Math.max(streamCount, MIN_SPLIT_COUNT);
    CreateReadSessionRequest createReadSessionRequest = CreateReadSessionRequest.newBuilder().setParent(BigQueryHelpers.toProjectResourceName(bqOptions.getBigQueryProject() == null ? bqOptions.getProject() : bqOptions.getBigQueryProject())).setReadSession(readSessionBuilder).setMaxStreamCount(streamCount).build();
    ReadSession readSession;
    try (StorageClient client = bqServices.getStorageClient(bqOptions)) {
        readSession = client.createReadSession(createReadSessionRequest);
        LOG.info("Sent BigQuery Storage API CreateReadSession request '{}'; received response '{}'.", createReadSessionRequest, readSession);
    }
    if (readSession.getStreamsList().isEmpty()) {
        // The underlying table is empty or all rows have been pruned.
        return ImmutableList.of();
    }
    Schema sessionSchema;
    if (readSession.getDataFormat() == DataFormat.ARROW) {
        org.apache.arrow.vector.types.pojo.Schema schema = ArrowConversion.arrowSchemaFromInput(readSession.getArrowSchema().getSerializedSchema().newInput());
        org.apache.beam.sdk.schemas.Schema beamSchema = ArrowConversion.ArrowSchemaTranslator.toBeamSchema(schema);
        sessionSchema = AvroUtils.toAvroSchema(beamSchema);
    } else if (readSession.getDataFormat() == DataFormat.AVRO) {
        sessionSchema = new Schema.Parser().parse(readSession.getAvroSchema().getSchema());
    } else {
        throw new IllegalArgumentException("data is not in a supported dataFormat: " + readSession.getDataFormat());
    }
    TableSchema trimmedSchema = BigQueryAvroUtils.trimBigQueryTableSchema(targetTable.getSchema(), sessionSchema);
    List<BigQueryStorageStreamSource<T>> sources = Lists.newArrayList();
    for (ReadStream readStream : readSession.getStreamsList()) {
        sources.add(BigQueryStorageStreamSource.create(readSession, readStream, trimmedSchema, parseFn, outputCoder, bqServices));
    }
    return ImmutableList.copyOf(sources);
}

Example 7 with TableSchema

use of com.google.cloud.bigquery.storage.v1.TableSchema in project beam by apache.

the class BigQueryIOStorageQueryTest method doQuerySourceInitialSplit.

private void doQuerySourceInitialSplit(long bundleSize, int requestedStreamCount, int expectedStreamCount) throws Exception {
    TableReference sourceTableRef = BigQueryHelpers.parseTableSpec("project:dataset.table");
    fakeDatasetService.createDataset(sourceTableRef.getProjectId(), sourceTableRef.getDatasetId(), "asia-northeast1", "Fake plastic tree^H^H^H^Htables", null);
    fakeDatasetService.createTable(new Table().setTableReference(sourceTableRef).setLocation("asia-northeast1"));
    Table queryResultTable = new Table().setSchema(new TableSchema().setFields(ImmutableList.of(new TableFieldSchema().setName("name").setType("STRING"), new TableFieldSchema().setName("number").setType("INTEGER")))).setNumBytes(1024L * 1024L);
    String encodedQuery = FakeBigQueryServices.encodeQueryResult(queryResultTable);
    fakeJobService.expectDryRunQuery(options.getProject(), encodedQuery, new JobStatistics().setQuery(new JobStatistics2().setTotalBytesProcessed(1024L * 1024L).setReferencedTables(ImmutableList.of(sourceTableRef))));
    String stepUuid = "testStepUuid";
    TableReference tempTableReference = createTempTableReference(options.getProject(), BigQueryResourceNaming.createJobIdPrefix(options.getJobName(), stepUuid, JobType.QUERY), Optional.empty());
    CreateReadSessionRequest expectedRequest = CreateReadSessionRequest.newBuilder().setParent("projects/" + options.getProject()).setReadSession(ReadSession.newBuilder().setTable(BigQueryHelpers.toTableResourceName(tempTableReference))).setMaxStreamCount(requestedStreamCount).build();
    Schema sessionSchema = SchemaBuilder.record("__root__").fields().name("name").type().nullable().stringType().noDefault().name("number").type().nullable().longType().noDefault().endRecord();
    ReadSession.Builder builder = ReadSession.newBuilder().setAvroSchema(AvroSchema.newBuilder().setSchema(sessionSchema.toString())).setDataFormat(DataFormat.AVRO);
    for (int i = 0; i < expectedStreamCount; i++) {
        builder.addStreams(ReadStream.newBuilder().setName("stream-" + i));
    }
    StorageClient fakeStorageClient = mock(StorageClient.class);
    when(fakeStorageClient.createReadSession(expectedRequest)).thenReturn(builder.build());
    BigQueryStorageQuerySource<TableRow> querySource = BigQueryStorageQuerySource.create(stepUuid, ValueProvider.StaticValueProvider.of(encodedQuery), /* flattenResults = */
    true, /* useLegacySql = */
    true, /* priority = */
    QueryPriority.BATCH, /* location = */
    null, /* queryTempDataset = */
    null, /* kmsKey = */
    null, null, new TableRowParser(), TableRowJsonCoder.of(), new FakeBigQueryServices().withDatasetService(fakeDatasetService).withJobService(fakeJobService).withStorageClient(fakeStorageClient));
    List<? extends BoundedSource<TableRow>> sources = querySource.split(bundleSize, options);
    assertEquals(expectedStreamCount, sources.size());
}

Example 8 with TableSchema

use of com.google.cloud.bigquery.storage.v1.TableSchema in project beam by apache.

the class BigQueryIOStorageReadTest method testTableSourceInitialSplit_EmptyTable.

@Test
public void testTableSourceInitialSplit_EmptyTable() throws Exception {
    fakeDatasetService.createDataset("foo.com:project", "dataset", "", "", null);
    TableReference tableRef = BigQueryHelpers.parseTableSpec("foo.com:project:dataset.table");
    Table table = new Table().setTableReference(tableRef).setNumBytes(1024L * 1024L).setSchema(new TableSchema());
    fakeDatasetService.createTable(table);
    CreateReadSessionRequest expectedRequest = CreateReadSessionRequest.newBuilder().setParent("projects/project-id").setReadSession(ReadSession.newBuilder().setTable("projects/foo.com:project/datasets/dataset/tables/table")).setMaxStreamCount(1024).build();
    ReadSession emptyReadSession = ReadSession.newBuilder().build();
    StorageClient fakeStorageClient = mock(StorageClient.class);
    when(fakeStorageClient.createReadSession(expectedRequest)).thenReturn(emptyReadSession);
    BigQueryStorageTableSource<TableRow> tableSource = BigQueryStorageTableSource.create(ValueProvider.StaticValueProvider.of(tableRef), null, null, new TableRowParser(), TableRowJsonCoder.of(), new FakeBigQueryServices().withDatasetService(fakeDatasetService).withStorageClient(fakeStorageClient));
    List<? extends BoundedSource<TableRow>> sources = tableSource.split(1024L, options);
    assertTrue(sources.isEmpty());
}

Example 9 with TableSchema

use of com.google.cloud.bigquery.storage.v1.TableSchema in project beam by apache.

the class BigQueryHllSketchCompatibilityIT method writeSketchToBigQuery.

private void writeSketchToBigQuery(List<String> testData, String expectedChecksum) {
    String tableSpec = String.format("%s.%s", DATASET_ID, SKETCH_TABLE_ID);
    String query = String.format("SELECT HLL_COUNT.EXTRACT(%s) FROM %s", SKETCH_FIELD_NAME, tableSpec);
    TableSchema tableSchema = new TableSchema().setFields(Collections.singletonList(new TableFieldSchema().setName(SKETCH_FIELD_NAME).setType(SKETCH_FIELD_TYPE)));
    TestPipelineOptions options = TestPipeline.testingPipelineOptions().as(TestPipelineOptions.class);
    Pipeline p = Pipeline.create(options);
    // until we have a stub class for BigQuery TableRow
    @SuppressWarnings("nullness") SerializableFunction<byte[], TableRow> formatFn = sketch -> new TableRow().set(SKETCH_FIELD_NAME, sketch.length == 0 ? null : sketch);
    p.apply(Create.of(testData).withType(TypeDescriptor.of(String.class))).apply(HllCount.Init.forStrings().globally()).apply(BigQueryIO.<byte[]>write().to(tableSpec).withSchema(tableSchema).withFormatFunction(formatFn).withWriteDisposition(BigQueryIO.Write.WriteDisposition.WRITE_TRUNCATE));
    p.run().waitUntilFinish();
    // BigqueryMatcher will send a query to retrieve the estimated count and verifies its
    // correctness using checksum.
    assertThat(createQueryUsingStandardSql(APP_NAME, PROJECT_ID, query), queryResultHasChecksum(expectedChecksum));
}

Example 10 with TableSchema

use of com.google.cloud.bigquery.storage.v1.TableSchema in project beam by apache.

the class BigQueryHllSketchCompatibilityIT method prepareDatasetAndDataTables.

@BeforeClass
public static void prepareDatasetAndDataTables() throws Exception {
    BIGQUERY_CLIENT.createNewDataset(PROJECT_ID, DATASET_ID);
    TableSchema dataTableSchema = new TableSchema().setFields(Collections.singletonList(new TableFieldSchema().setName(DATA_FIELD_NAME).setType(DATA_FIELD_TYPE)));
    Table dataTableNonEmpty = new Table().setSchema(dataTableSchema).setTableReference(new TableReference().setProjectId(PROJECT_ID).setDatasetId(DATASET_ID).setTableId(DATA_TABLE_ID_NON_EMPTY));
    BIGQUERY_CLIENT.createNewTable(PROJECT_ID, DATASET_ID, dataTableNonEmpty);
    // Prepopulates dataTableNonEmpty with TEST_DATA
    List<Map<String, Object>> rows = TEST_DATA.stream().map(v -> Collections.singletonMap(DATA_FIELD_NAME, (Object) v)).collect(Collectors.toList());
    BIGQUERY_CLIENT.insertDataToTable(PROJECT_ID, DATASET_ID, DATA_TABLE_ID_NON_EMPTY, rows);
    Table dataTableEmpty = new Table().setSchema(dataTableSchema).setTableReference(new TableReference().setProjectId(PROJECT_ID).setDatasetId(DATASET_ID).setTableId(DATA_TABLE_ID_EMPTY));
    BIGQUERY_CLIENT.createNewTable(PROJECT_ID, DATASET_ID, dataTableEmpty);
}