use of org.apache.druid.data.input.InputEntityReader in project druid by druid-io.
the class KafkaInputFormatTest method testWithHeaderKeyAndValue.
@Test
public void testWithHeaderKeyAndValue() throws IOException {
final byte[] key = StringUtils.toUtf8("{\n" + " \"key\": \"sampleKey\"\n" + "}");
final byte[] payload = StringUtils.toUtf8("{\n" + " \"timestamp\": \"2021-06-25\",\n" + " \"bar\": null,\n" + " \"foo\": \"x\",\n" + " \"baz\": 4,\n" + " \"o\": {\n" + " \"mg\": 1\n" + " }\n" + "}");
Headers headers = new RecordHeaders(SAMPLE_HEADERS);
inputEntity = new KafkaRecordEntity(new ConsumerRecord<byte[], byte[]>("sample", 0, 0, timestamp, null, null, 0, 0, key, payload, headers));
final InputEntityReader reader = format.createReader(new InputRowSchema(new TimestampSpec("timestamp", "iso", null), new DimensionsSpec(DimensionsSpec.getDefaultSchemas(ImmutableList.of("bar", "foo", "kafka.newheader.encoding", "kafka.newheader.kafkapkc", "kafka.newts.timestamp"))), ColumnsFilter.all()), inputEntity, null);
final int numExpectedIterations = 1;
try (CloseableIterator<InputRow> iterator = reader.read()) {
int numActualIterations = 0;
while (iterator.hasNext()) {
final InputRow row = iterator.next();
// Payload verifications
Assert.assertEquals(DateTimes.of("2021-06-25"), row.getTimestamp());
Assert.assertEquals("x", Iterables.getOnlyElement(row.getDimension("foo")));
Assert.assertEquals("4", Iterables.getOnlyElement(row.getDimension("baz")));
Assert.assertEquals("4", Iterables.getOnlyElement(row.getDimension("root_baz")));
Assert.assertEquals("1", Iterables.getOnlyElement(row.getDimension("path_omg")));
Assert.assertEquals("1", Iterables.getOnlyElement(row.getDimension("jq_omg")));
// Header verification
Assert.assertEquals("application/json", Iterables.getOnlyElement(row.getDimension("kafka.newheader.encoding")));
Assert.assertEquals("pkc-bar", Iterables.getOnlyElement(row.getDimension("kafka.newheader.kafkapkc")));
Assert.assertEquals(String.valueOf(DateTimes.of("2021-06-24").getMillis()), Iterables.getOnlyElement(row.getDimension("kafka.newts.timestamp")));
Assert.assertEquals("2021-06-25", Iterables.getOnlyElement(row.getDimension("timestamp")));
// Key verification
Assert.assertEquals("sampleKey", Iterables.getOnlyElement(row.getDimension("kafka.newkey.key")));
Assert.assertTrue(row.getDimension("root_baz2").isEmpty());
Assert.assertTrue(row.getDimension("path_omg2").isEmpty());
Assert.assertTrue(row.getDimension("jq_omg2").isEmpty());
numActualIterations++;
}
Assert.assertEquals(numExpectedIterations, numActualIterations);
}
}
use of org.apache.druid.data.input.InputEntityReader in project druid by druid-io.
the class KafkaInputFormatTest method testWithOutKeyAndHeaderSpecs.
@Test
public void testWithOutKeyAndHeaderSpecs() throws IOException {
final byte[] payload = StringUtils.toUtf8("{\n" + " \"timestamp\": \"2021-06-24\",\n" + " \"bar\": null,\n" + " \"foo\": \"x\",\n" + " \"baz\": 4,\n" + " \"o\": {\n" + " \"mg\": 1\n" + " }\n" + "}");
Headers headers = new RecordHeaders(SAMPLE_HEADERS);
inputEntity = new KafkaRecordEntity(new ConsumerRecord<byte[], byte[]>("sample", 0, 0, timestamp, null, null, 0, 0, null, payload, headers));
KafkaInputFormat localFormat = new KafkaInputFormat(null, null, // Value Format
new JsonInputFormat(new JSONPathSpec(true, ImmutableList.of(new JSONPathFieldSpec(JSONPathFieldType.ROOT, "root_baz", "baz"), new JSONPathFieldSpec(JSONPathFieldType.ROOT, "root_baz2", "baz2"), new JSONPathFieldSpec(JSONPathFieldType.PATH, "path_omg", "$.o.mg"), new JSONPathFieldSpec(JSONPathFieldType.PATH, "path_omg2", "$.o.mg2"), new JSONPathFieldSpec(JSONPathFieldType.JQ, "jq_omg", ".o.mg"), new JSONPathFieldSpec(JSONPathFieldType.JQ, "jq_omg2", ".o.mg2"))), // make sure JsonReader is used
null, // make sure JsonReader is used
null, // make sure JsonReader is used
false), "kafka.newheader.", "kafka.newkey.", "kafka.newts.");
final InputEntityReader reader = localFormat.createReader(new InputRowSchema(new TimestampSpec("timestamp", "iso", null), new DimensionsSpec(DimensionsSpec.getDefaultSchemas(ImmutableList.of("bar", "foo", "kafka.newts.timestamp"))), ColumnsFilter.all()), inputEntity, null);
final int numExpectedIterations = 1;
try (CloseableIterator<InputRow> iterator = reader.read()) {
int numActualIterations = 0;
while (iterator.hasNext()) {
final InputRow row = iterator.next();
// Key verification
Assert.assertTrue(row.getDimension("kafka.newkey.key").isEmpty());
Assert.assertEquals("x", Iterables.getOnlyElement(row.getDimension("foo")));
Assert.assertEquals("4", Iterables.getOnlyElement(row.getDimension("baz")));
Assert.assertEquals("4", Iterables.getOnlyElement(row.getDimension("root_baz")));
Assert.assertEquals("1", Iterables.getOnlyElement(row.getDimension("path_omg")));
Assert.assertEquals("1", Iterables.getOnlyElement(row.getDimension("jq_omg")));
numActualIterations++;
}
Assert.assertEquals(numExpectedIterations, numActualIterations);
}
}
use of org.apache.druid.data.input.InputEntityReader in project druid by druid-io.
the class KafkaInputFormatTest method testWithOutKey.
@Test
public // Headers cannot be null, so testing only no key use case!
void testWithOutKey() throws IOException {
final byte[] payload = StringUtils.toUtf8("{\n" + " \"timestamp\": \"2021-06-24\",\n" + " \"bar\": null,\n" + " \"foo\": \"x\",\n" + " \"baz\": 4,\n" + " \"o\": {\n" + " \"mg\": 1\n" + " }\n" + "}");
Headers headers = new RecordHeaders(SAMPLE_HEADERS);
inputEntity = new KafkaRecordEntity(new ConsumerRecord<byte[], byte[]>("sample", 0, 0, timestamp, null, null, 0, 0, null, payload, headers));
final InputEntityReader reader = format.createReader(new InputRowSchema(new TimestampSpec("timestamp", "iso", null), new DimensionsSpec(DimensionsSpec.getDefaultSchemas(ImmutableList.of("bar", "foo", "kafka.newheader.encoding", "kafka.newheader.kafkapkc", "kafka.newts.timestamp"))), ColumnsFilter.all()), inputEntity, null);
final int numExpectedIterations = 1;
try (CloseableIterator<InputRow> iterator = reader.read()) {
int numActualIterations = 0;
while (iterator.hasNext()) {
final InputRow row = iterator.next();
// Key verification
Assert.assertTrue(row.getDimension("kafka.newkey.key").isEmpty());
numActualIterations++;
}
Assert.assertEquals(numExpectedIterations, numActualIterations);
}
}
Aggregations