Examples with TimestampSpec org.apache.druid.data.input.impl.TimestampSpec used on opensource projects

Example 76 with TimestampSpec

use of org.apache.druid.data.input.impl.TimestampSpec in project druid by druid-io.

the class OrcReaderTest method testOrcSplitElim.

// This test is migrated from OrcHadoopInputRowParserTest
@Test
public void testOrcSplitElim() throws IOException {
    final InputEntityReader reader = createReader(new TimestampSpec("ts", "millis", null), new DimensionsSpec(null), new OrcInputFormat(new JSONPathSpec(true, null), null, new Configuration()), "example/orc_split_elim.orc");
    try (CloseableIterator<InputRow> iterator = reader.read()) {
        int actualRowCount = 0;
        Assert.assertTrue(iterator.hasNext());
        final InputRow row = iterator.next();
        actualRowCount++;
        Assert.assertEquals(DateTimes.of("1969-12-31T16:00:00.0Z"), row.getTimestamp());
        Assert.assertEquals("2", Iterables.getOnlyElement(row.getDimension("userid")));
        Assert.assertEquals("foo", Iterables.getOnlyElement(row.getDimension("string1")));
        Assert.assertEquals("0.8", Iterables.getOnlyElement(row.getDimension("subtype")));
        Assert.assertEquals("1.2", Iterables.getOnlyElement(row.getDimension("decimal1")));
        while (iterator.hasNext()) {
            actualRowCount++;
            iterator.next();
        }
        Assert.assertEquals(25000, actualRowCount);
    }
}

Example 77 with TimestampSpec

use of org.apache.druid.data.input.impl.TimestampSpec in project hive by apache.

the class TestDruidRecordWriter method testWrite.

// Test is failing due to Guava dependency, Druid 0.13.0 should have less dependency on Guava
@Ignore
@Test
public void testWrite() throws IOException, SegmentLoadingException {
    final String dataSourceName = "testDataSource";
    final File segmentOutputDir = temporaryFolder.newFolder();
    final File workingDir = temporaryFolder.newFolder();
    Configuration config = new Configuration();
    final InputRowParser inputRowParser = new MapInputRowParser(new TimeAndDimsParseSpec(new TimestampSpec(DruidConstants.DEFAULT_TIMESTAMP_COLUMN, "auto", null), new DimensionsSpec(ImmutableList.of(new StringDimensionSchema("host")), null, null)));
    final Map<String, Object> parserMap = objectMapper.convertValue(inputRowParser, new TypeReference<Map<String, Object>>() {
    });
    DataSchema dataSchema = new DataSchema(dataSourceName, parserMap, new AggregatorFactory[] { new LongSumAggregatorFactory("visited_sum", "visited_sum"), new HyperUniquesAggregatorFactory("unique_hosts", "unique_hosts") }, new UniformGranularitySpec(Granularities.DAY, Granularities.NONE, ImmutableList.of(INTERVAL_FULL)), null, objectMapper);
    IndexSpec indexSpec = new IndexSpec(new RoaringBitmapSerdeFactory(true), null, null, null);
    RealtimeTuningConfig tuningConfig = new RealtimeTuningConfig(null, null, null, null, temporaryFolder.newFolder(), null, null, null, null, indexSpec, null, null, 0, 0, null, null, 0L, null, null);
    LocalFileSystem localFileSystem = FileSystem.getLocal(config);
    DataSegmentPusher dataSegmentPusher = new LocalDataSegmentPusher(new LocalDataSegmentPusherConfig() {

        @Override
        public File getStorageDirectory() {
            return segmentOutputDir;
        }
    });
    Path segmentDescriptorPath = new Path(workingDir.getAbsolutePath(), DruidStorageHandler.SEGMENTS_DESCRIPTOR_DIR_NAME);
    DruidRecordWriter druidRecordWriter = new DruidRecordWriter(dataSchema, tuningConfig, dataSegmentPusher, 20, segmentDescriptorPath, localFileSystem);
    List<DruidWritable> druidWritables = expectedRows.stream().map(input -> new DruidWritable(ImmutableMap.<String, Object>builder().putAll(input).put(Constants.DRUID_TIMESTAMP_GRANULARITY_COL_NAME, Granularities.DAY.bucketStart(new DateTime((long) input.get(DruidConstants.DEFAULT_TIMESTAMP_COLUMN))).getMillis()).build())).collect(Collectors.toList());
    for (DruidWritable druidWritable : druidWritables) {
        druidRecordWriter.write(druidWritable);
    }
    druidRecordWriter.close(false);
    List<DataSegment> dataSegmentList = DruidStorageHandlerUtils.getCreatedSegments(segmentDescriptorPath, config);
    Assert.assertEquals(1, dataSegmentList.size());
    File tmpUnzippedSegmentDir = temporaryFolder.newFolder();
    new LocalDataSegmentPuller().getSegmentFiles(dataSegmentList.get(0), tmpUnzippedSegmentDir);
    final QueryableIndex queryableIndex = DruidStorageHandlerUtils.INDEX_IO.loadIndex(tmpUnzippedSegmentDir);
    QueryableIndexStorageAdapter adapter = new QueryableIndexStorageAdapter(queryableIndex);
    Firehose firehose = new IngestSegmentFirehose(ImmutableList.of(new WindowedStorageAdapter(adapter, adapter.getInterval())), null, ImmutableList.of("host"), ImmutableList.of("visited_sum", "unique_hosts"), null);
    List<InputRow> rows = Lists.newArrayList();
    while (firehose.hasMore()) {
        rows.add(firehose.nextRow());
    }
    verifyRows(expectedRows, rows);
}

Example 78 with TimestampSpec

use of org.apache.druid.data.input.impl.TimestampSpec in project druid by druid-io.

the class FirehoseFactoryToInputSourceAdaptorTest method testUnimplementedInputFormat.

@Test
public void testUnimplementedInputFormat() throws IOException {
    final List<String> lines = new ArrayList<>();
    for (int i = 0; i < 10; i++) {
        lines.add(StringUtils.format("%d,name_%d,%d", 20190101 + i, i, i + 100));
    }
    final TestFirehoseFactory firehoseFactory = new TestFirehoseFactory(lines);
    final StringInputRowParser inputRowParser = new StringInputRowParser(new CSVParseSpec(new TimestampSpec(null, "yyyyMMdd", null), new DimensionsSpec(DimensionsSpec.getDefaultSchemas(Arrays.asList("timestamp", "name", "score"))), ",", Arrays.asList("timestamp", "name", "score"), false, 0), StringUtils.UTF8_STRING);
    final FirehoseFactoryToInputSourceAdaptor inputSourceAdaptor = new FirehoseFactoryToInputSourceAdaptor(firehoseFactory, inputRowParser);
    final InputSourceReader reader = inputSourceAdaptor.reader(new InputRowSchema(inputRowParser.getParseSpec().getTimestampSpec(), inputRowParser.getParseSpec().getDimensionsSpec(), ColumnsFilter.all()), null, null);
    final List<InputRow> result = new ArrayList<>();
    try (CloseableIterator<InputRow> iterator = reader.read()) {
        while (iterator.hasNext()) {
            result.add(iterator.next());
        }
    }
    Assert.assertEquals(10, result.size());
    for (int i = 0; i < 10; i++) {
        Assert.assertEquals(DateTimes.of(StringUtils.format("2019-01-%02d", 1 + i)), result.get(i).getTimestamp());
        Assert.assertEquals(StringUtils.format("name_%d", i), Iterables.getOnlyElement(result.get(i).getDimension("name")));
        Assert.assertEquals(StringUtils.format("%d", i + 100), Iterables.getOnlyElement(result.get(i).getDimension("score")));
    }
}

Example 79 with TimestampSpec

use of org.apache.druid.data.input.impl.TimestampSpec in project druid by druid-io.

the class OssInputSourceTest method testCompressedReader.

@Test
public void testCompressedReader() throws IOException {
    EasyMock.reset(OSSCLIENT);
    expectListObjects(PREFIXES.get(0), ImmutableList.of(EXPECTED_COMPRESSED_URIS.get(0)), CONTENT);
    expectListObjects(EXPECTED_COMPRESSED_URIS.get(1), ImmutableList.of(EXPECTED_COMPRESSED_URIS.get(1)), CONTENT);
    expectGetObjectCompressed(EXPECTED_COMPRESSED_URIS.get(0));
    expectGetObjectCompressed(EXPECTED_COMPRESSED_URIS.get(1));
    EasyMock.replay(OSSCLIENT);
    OssInputSource inputSource = new OssInputSource(OSSCLIENT, INPUT_DATA_CONFIG, null, ImmutableList.of(PREFIXES.get(0), EXPECTED_COMPRESSED_URIS.get(1)), null, null);
    InputRowSchema someSchema = new InputRowSchema(new TimestampSpec("time", "auto", null), new DimensionsSpec(DimensionsSpec.getDefaultSchemas(ImmutableList.of("dim1", "dim2"))), ColumnsFilter.all());
    InputSourceReader reader = inputSource.reader(someSchema, new CsvInputFormat(ImmutableList.of("time", "dim1", "dim2"), "|", false, null, 0), temporaryFolder.newFolder());
    CloseableIterator<InputRow> iterator = reader.read();
    while (iterator.hasNext()) {
        InputRow nextRow = iterator.next();
        Assert.assertEquals(NOW, nextRow.getTimestamp());
        Assert.assertEquals("hello", nextRow.getDimension("dim1").get(0));
        Assert.assertEquals("world", nextRow.getDimension("dim2").get(0));
    }
    EasyMock.verify(OSSCLIENT);
}

Example 80 with TimestampSpec

use of org.apache.druid.data.input.impl.TimestampSpec in project druid by druid-io.

the class FlattenJSONBenchmarkUtil method getForcedPathParser.

public Parser getForcedPathParser() {
    List<JSONPathFieldSpec> fields = new ArrayList<>();
    fields.add(JSONPathFieldSpec.createNestedField("ts", "$['ts']"));
    fields.add(JSONPathFieldSpec.createNestedField("d1", "$['d1']"));
    fields.add(JSONPathFieldSpec.createNestedField("d2", "$['d2']"));
    fields.add(JSONPathFieldSpec.createNestedField("e1.d1", "$['e1.d1']"));
    fields.add(JSONPathFieldSpec.createNestedField("e1.d2", "$['e1.d2']"));
    fields.add(JSONPathFieldSpec.createNestedField("e2.d3", "$['e2.d3']"));
    fields.add(JSONPathFieldSpec.createNestedField("e2.d4", "$['e2.d4']"));
    fields.add(JSONPathFieldSpec.createNestedField("e2.d5", "$['e2.d5']"));
    fields.add(JSONPathFieldSpec.createNestedField("e2.d6", "$['e2.d6']"));
    fields.add(JSONPathFieldSpec.createNestedField("e2.ad1[0]", "$['e2.ad1[0]']"));
    fields.add(JSONPathFieldSpec.createNestedField("e2.ad1[1]", "$['e2.ad1[1]']"));
    fields.add(JSONPathFieldSpec.createNestedField("e2.ad1[2]", "$['e2.ad1[2]']"));
    fields.add(JSONPathFieldSpec.createNestedField("ae1[0].d1", "$['ae1[0].d1']"));
    fields.add(JSONPathFieldSpec.createNestedField("ae1[1].d1", "$['ae1[1].d1']"));
    fields.add(JSONPathFieldSpec.createNestedField("ae1[2].e1.d2", "$['ae1[2].e1.d2']"));
    fields.add(JSONPathFieldSpec.createNestedField("m3", "$['m3']"));
    fields.add(JSONPathFieldSpec.createNestedField("m4", "$['m4']"));
    fields.add(JSONPathFieldSpec.createNestedField("e3.m1", "$['e3.m1']"));
    fields.add(JSONPathFieldSpec.createNestedField("e3.m2", "$['e3.m2']"));
    fields.add(JSONPathFieldSpec.createNestedField("e3.m3", "$['e3.m3']"));
    fields.add(JSONPathFieldSpec.createNestedField("e3.m4", "$['e3.m4']"));
    fields.add(JSONPathFieldSpec.createNestedField("e3.am1[0]", "$['e3.am1[0]']"));
    fields.add(JSONPathFieldSpec.createNestedField("e3.am1[1]", "$['e3.am1[1]']"));
    fields.add(JSONPathFieldSpec.createNestedField("e3.am1[2]", "$['e3.am1[2]']"));
    fields.add(JSONPathFieldSpec.createNestedField("e3.am1[3]", "$['e3.am1[3]']"));
    fields.add(JSONPathFieldSpec.createNestedField("e4.e4.m4", "$['e4.e4.m4']"));
    JSONPathSpec flattenSpec = new JSONPathSpec(false, fields);
    JSONParseSpec spec = new JSONParseSpec(new TimestampSpec("ts", "iso", null), DimensionsSpec.EMPTY, flattenSpec, null, null);
    return spec.makeParser();
}