Examples with TableFieldSchema com.google.api.services.bigquery.model.TableFieldSchema used on opensource projects

Example 71 with TableFieldSchema

use of com.google.api.services.bigquery.model.TableFieldSchema in project beam by apache.

the class BigQuerySchemaUpdateOptionsIT method testAllowFieldRelaxation.

@Test
public void testAllowFieldRelaxation() throws Exception {
    String tableName = makeTestTable();
    Set<SchemaUpdateOption> schemaUpdateOptions = EnumSet.of(BigQueryIO.Write.SchemaUpdateOption.ALLOW_FIELD_RELAXATION);
    TableSchema newSchema = new TableSchema().setFields(ImmutableList.of(new TableFieldSchema().setName("optional_field").setType("STRING")));
    String value = "hellooo";
    TableRow rowToInsert = new TableRow().set("optional_field", value);
    String testQuery = String.format("SELECT optional_field FROM [%s.%s];", BIG_QUERY_DATASET_ID, tableName);
    List<List<String>> expectedResult = Arrays.asList(Arrays.asList(value));
    runWriteTest(schemaUpdateOptions, tableName, newSchema, rowToInsert, testQuery, expectedResult);
}

Example 72 with TableFieldSchema

use of com.google.api.services.bigquery.model.TableFieldSchema in project beam by apache.

the class BigQueryIOReadTest method testEstimatedSizeWithStreamingBuffer.

@Test
public void testEstimatedSizeWithStreamingBuffer() throws Exception {
    List<TableRow> data = ImmutableList.of(new TableRow().set("name", "a").set("number", 1L), new TableRow().set("name", "b").set("number", 2L), new TableRow().set("name", "c").set("number", 3L), new TableRow().set("name", "d").set("number", 4L), new TableRow().set("name", "e").set("number", 5L), new TableRow().set("name", "f").set("number", 6L));
    TableReference table = BigQueryHelpers.parseTableSpec("project:data_set.table_name");
    fakeDatasetService.createDataset("project", "data_set", "", "", null);
    fakeDatasetService.createTable(new Table().setTableReference(table).setSchema(new TableSchema().setFields(ImmutableList.of(new TableFieldSchema().setName("name").setType("STRING"), new TableFieldSchema().setName("number").setType("INTEGER")))).setStreamingBuffer(new Streamingbuffer().setEstimatedBytes(BigInteger.valueOf(10))));
    fakeDatasetService.insertAll(table, data, null);
    String stepUuid = "testStepUuid";
    BoundedSource<TableRow> bqSource = BigQueryTableSourceDef.create(fakeBqServices, ValueProvider.StaticValueProvider.of(table)).toSource(stepUuid, TableRowJsonCoder.of(), BigQueryIO.TableRowParser.INSTANCE, false);
    PipelineOptions options = PipelineOptionsFactory.create();
    // Each row should have 24 bytes (See StringUtf8Coder in detail):
    // first 1 byte indicating length and following 23 bytes: {"name":"a","number":1}
    // 10 bytes comes from the estimated bytes of the Streamingbuffer
    long expectedSize = 24L * data.size() + 10;
    assertEquals(expectedSize, bqSource.getEstimatedSizeBytes(options));
}

Example 73 with TableFieldSchema

use of com.google.api.services.bigquery.model.TableFieldSchema in project beam by apache.

the class BigQueryIOReadTest method testEstimatedSizeWithoutStreamingBuffer.

@Test
public void testEstimatedSizeWithoutStreamingBuffer() throws Exception {
    List<TableRow> data = ImmutableList.of(new TableRow().set("name", "a").set("number", 1L), new TableRow().set("name", "b").set("number", 2L), new TableRow().set("name", "c").set("number", 3L), new TableRow().set("name", "d").set("number", 4L), new TableRow().set("name", "e").set("number", 5L), new TableRow().set("name", "f").set("number", 6L));
    TableReference table = BigQueryHelpers.parseTableSpec("project:data_set.table_name");
    fakeDatasetService.createDataset("project", "data_set", "", "", null);
    fakeDatasetService.createTable(new Table().setTableReference(table).setSchema(new TableSchema().setFields(ImmutableList.of(new TableFieldSchema().setName("name").setType("STRING"), new TableFieldSchema().setName("number").setType("INTEGER")))));
    fakeDatasetService.insertAll(table, data, null);
    String stepUuid = "testStepUuid";
    BoundedSource<TableRow> bqSource = BigQueryTableSourceDef.create(fakeBqServices, ValueProvider.StaticValueProvider.of(table)).toSource(stepUuid, TableRowJsonCoder.of(), BigQueryIO.TableRowParser.INSTANCE, false);
    PipelineOptions options = PipelineOptionsFactory.create();
    // Each row should have 24 bytes (See StringUtf8Coder in detail):
    // first 1 byte indicating length and following 23 bytes: {"name":"a","number":1}
    long expectedSize = 24L * data.size();
    assertEquals(expectedSize, bqSource.getEstimatedSizeBytes(options));
}

Example 74 with TableFieldSchema

use of com.google.api.services.bigquery.model.TableFieldSchema in project beam by apache.

the class BigQueryIOStorageQueryTest method testQuerySourceInitialSplit_NoReferencedTables.

/**
 * This test simulates the scenario where the SQL text which is executed by the query job doesn't
 * by itself refer to any tables (e.g. "SELECT 17 AS value"), and thus there are no referenced
 * tables when the dry run of the query is performed.
 */
@Test
public void testQuerySourceInitialSplit_NoReferencedTables() throws Exception {
    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())));
    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(1024).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 < 1024; 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(1024, options);
    assertEquals(1024, sources.size());
}

Example 75 with TableFieldSchema

use of com.google.api.services.bigquery.model.TableFieldSchema in project DataflowJavaSDK-examples by GoogleCloudPlatform.

the class WriteToBigQuery method getSchema.

/**
 * Build the output table schema.
 */
protected TableSchema getSchema() {
    List<TableFieldSchema> fields = new ArrayList<>();
    for (Map.Entry<String, FieldInfo<InputT>> entry : fieldInfo.entrySet()) {
        String key = entry.getKey();
        FieldInfo<InputT> fcnInfo = entry.getValue();
        String bqType = fcnInfo.getFieldType();
        fields.add(new TableFieldSchema().setName(key).setType(bqType));
    }
    return new TableSchema().setFields(fields);
}