private byte[] createAvroData(String name, int age, List<String> emails) throws IOException {
String AVRO_SCHEMA =
"{\n"
+ "\"type\": \"record\",\n"
+ "\"name\": \"Employee\",\n"
+ "\"fields\": [\n"
+ " {\"name\": \"name\", \"type\": \"string\"},\n"
+ " {\"name\": \"age\", \"type\": \"int\"},\n"
+ " {\"name\": \"emails\", \"type\": {\"type\": \"array\", \"items\": \"string\"}},\n"
+ " {\"name\": \"boss\", \"type\": [\"Employee\",\"null\"]}\n"
+ "]}";
Schema schema = new Schema.Parser().parse(AVRO_SCHEMA);
ByteArrayOutputStream out = new ByteArrayOutputStream();
GenericRecord e1 = new GenericData.Record(schema);
e1.put("name", name);
e1.put("age", age);
e1.put("emails", emails);
e1.put("boss", null);
DatumWriter<GenericRecord> datumWriter = new GenericDatumWriter<>(schema);
DataFileWriter<GenericRecord> dataFileWriter = new DataFileWriter<>(datumWriter);
dataFileWriter.create(schema, out);
dataFileWriter.append(e1);
dataFileWriter.close();
return out.toByteArray();
}
private void writeContainer(Record src, OutputStream dst) {
DataFileWriter dataFileWriter = null;
try {
try {
Schema schema = null;
for (Object attachment : src.get(Fields.ATTACHMENT_BODY)) {
Preconditions.checkNotNull(attachment);
GenericContainer datum = (GenericContainer) attachment;
schema = getSchema(datum, schema);
assert schema != null;
if (dataFileWriter == null) { // init
GenericDatumWriter datumWriter = new GenericDatumWriter(schema);
dataFileWriter = new DataFileWriter(datumWriter);
if (codecFactory != null) {
dataFileWriter.setCodec(codecFactory);
}
for (Map.Entry<String, String> entry : metadata.entrySet()) {
dataFileWriter.setMeta(entry.getKey(), entry.getValue());
}
dataFileWriter.create(schema, dst);
}
dataFileWriter.append(datum);
}
if (dataFileWriter != null) {
dataFileWriter.flush();
}
} catch (IOException e) {
throw new MorphlineRuntimeException(e);
}
} finally {
Closeables.closeQuietly(dataFileWriter);
}
}
@Test
public void test() throws IOException {
File file = new File("target/AvroDocument.avro");
Schema schema = AvroDocument._SCHEMA;
{
System.out.println("Writing to: " + file.getAbsolutePath());
DatumWriter<Object> datumWriter = new SpecificDatumWriter(AvroDocument.class);
FileOutputStream outputStream = new FileOutputStream(file);
DataFileWriter<Object> dfw = new DataFileWriter<Object>(schema, outputStream, datumWriter);
AvroDocument d = createTestDocument();
dfw.append(d);
dfw.flush();
dfw.close();
}
{
System.out.println("Reading from: " + file.getAbsolutePath());
DatumReader<Object> datumReader = new SpecificDatumReader(AvroDocument.class);
SeekableInput seekableInput = new SeekableFileInput(file);
DataFileReader<Object> dfr = new DataFileReader<Object>(seekableInput, datumReader);
AvroDocument d = new AvroDocument();
dfr.next(d);
AvroDocumentReader.dumpAvroDocument(d, System.out);
}
}
private static <T> File createFile(File file, Schema schema, T... records) throws IOException {
DatumWriter<T> datumWriter = new GenericDatumWriter<T>(schema);
DataFileWriter<T> fileWriter = new DataFileWriter<T>(datumWriter);
fileWriter.create(schema, file);
for (T record : records) {
fileWriter.append(record);
}
fileWriter.close();
return file;
}
public static void main(String[] args) throws IOException {
User user1 = new User();
user1.setName("Alyssa");
user1.setFavoriteNumber(256);
// Leave favorite color null
// Alternate constructor
User user2 = new User("Ben", 7, "red");
// Construct via builder
User user3 =
User.newBuilder()
.setName("Charlie")
.setFavoriteColor("blue")
.setFavoriteNumber(null)
.build();
// Serialize user1 and user2 to disk
File file = new File("users.avro");
DatumWriter<User> userDatumWriter = new SpecificDatumWriter<User>(User.class);
DataFileWriter<User> dataFileWriter = new DataFileWriter<User>(userDatumWriter);
dataFileWriter.create(user1.getSchema(), file);
dataFileWriter.append(user1);
dataFileWriter.append(user2);
dataFileWriter.append(user3);
dataFileWriter.close();
// Deserialize Users from disk
DatumReader<User> userDatumReader = new SpecificDatumReader<User>(User.class);
DataFileReader<User> dataFileReader = new DataFileReader<User>(file, userDatumReader);
try {
User user = null;
while (dataFileReader.hasNext()) {
// Reuse user object by passing it to next(). This saves us from
// allocating and garbage collecting many objects for files with
// many items.
user = dataFileReader.next(user);
System.out.println(user);
}
} finally {
dataFileReader.close();
}
}
/** Writes an avro file of generic records with a 'key', 'blah', and 'value' field. */
private Path writeGenericRecordAvroFile() throws IOException {
// Open a writer.
final File file = new File(getLocalTempDir(), "generic-kv.avro");
final Schema writerSchema = Schema.createRecord("record", null, null, false);
writerSchema.setFields(
Lists.newArrayList(
new Schema.Field("key", Schema.create(Schema.Type.INT), null, null),
new Schema.Field("blah", Schema.create(Schema.Type.STRING), null, null),
new Schema.Field("value", Schema.create(Schema.Type.STRING), null, null)));
final DataFileWriter<GenericRecord> fileWriter =
new DataFileWriter<GenericRecord>(new GenericDatumWriter<GenericRecord>(writerSchema))
.create(writerSchema, file);
try {
// Write a record.
GenericData.Record record = new GenericData.Record(writerSchema);
record.put("key", 1);
record.put("blah", "blah");
record.put("value", "one");
fileWriter.append(record);
// Write another record.
record = new GenericData.Record(writerSchema);
record.put("key", 2);
record.put("blah", "blah");
record.put("value", "two");
fileWriter.append(record);
// Write a duplicate record with the same key field value.
record = new GenericData.Record(writerSchema);
record.put("key", 2);
record.put("blah", "blah");
record.put("value", "deux");
fileWriter.append(record);
// Close it and return the path.
} finally {
fileWriter.close();
}
return new Path(file.getPath());
}
public void execute(TridentTuple tuple, TridentCollector collector) {
GenericRecord docEntry = new GenericData.Record(schema);
docEntry.put("docid", tuple.getStringByField("documentId"));
docEntry.put("time", Time.currentTimeMillis());
docEntry.put("line", tuple.getStringByField("document"));
try {
dataFileWriter.append(docEntry);
dataFileWriter.flush();
} catch (IOException e) {
LOG.error("Error writing to document record: " + e);
throw new RuntimeException(e);
}
}
@Override
public void execute(Tuple inputTuple) {
/* Processing tuples of the shape
(DATASOURCE_ID, TIMESTAMP_FIELD, CONTENT_FIELD) */
// get datasource
String datasource = inputTuple.getStringByField(RestIngestionSpout.DATASOURCE_ID);
// compute month
long timestamp = inputTuple.getLongByField(RestIngestionSpout.TIMESTAMP_FIELD);
// this computation is completely stateless
String month = timestampToMonth(timestamp);
// now get the DataFileWriter
DataFileWriter<GenericRecord> writer = null;
try {
writer = this.writersCache.get(DatasourceMonth.create(datasource, month));
} catch (ExecutionException ee) {
LOGGER.error(
"Error getting DataFileWriter for tuple for datasource "
+ datasource
+ " and timestamp "
+ timestamp
+ " : "
+ ee.getMessage());
this.collector.fail(inputTuple);
return;
}
// create and write a new record
GenericRecord newDataRecord = new GenericData.Record(AVRO_SCHEMA);
newDataRecord.put(AVRO_TIMESTAMP_FIELD, new Long(timestamp));
newDataRecord.put(
AVRO_CONTENT_FIELD, inputTuple.getStringByField(RestIngestionSpout.CONTENT_FIELD));
try {
writer.append(newDataRecord);
} catch (IOException ioe) {
LOGGER.error(
"Error writing Avro record for datasource "
+ datasource
+ " and timestamp "
+ timestamp
+ " : "
+ ioe.getMessage());
this.collector.fail(inputTuple);
return;
}
// ACK processing for this tupe as ok
this.collector.ack(inputTuple);
}
private void populateGenericFile(List<GenericRecord> genericRecords) throws IOException {
FileOutputStream outputStream = new FileOutputStream(this.avroFile);
GenericDatumWriter<GenericRecord> genericDatumWriter =
new GenericDatumWriter<GenericRecord>(schema);
DataFileWriter<GenericRecord> dataFileWriter =
new DataFileWriter<GenericRecord>(genericDatumWriter);
dataFileWriter.create(schema, outputStream);
for (GenericRecord record : genericRecords) {
dataFileWriter.append(record);
}
dataFileWriter.close();
outputStream.close();
}
/**
* Generates an input Avro file containing the given records in the temporary directory and
* returns the full path of the file.
*/
private <T> String generateTestFile(
String filename,
List<T> elems,
SyncBehavior syncBehavior,
int syncInterval,
AvroCoder<T> coder,
String codec)
throws IOException {
Random random = new Random(0);
File tmpFile = tmpFolder.newFile(filename);
String path = tmpFile.toString();
FileOutputStream os = new FileOutputStream(tmpFile);
DatumWriter<T> datumWriter = coder.createDatumWriter();
try (DataFileWriter<T> writer = new DataFileWriter<>(datumWriter)) {
writer.setCodec(CodecFactory.fromString(codec));
writer.create(coder.getSchema(), os);
int recordIndex = 0;
int syncIndex = syncBehavior == SyncBehavior.SYNC_RANDOM ? random.nextInt(syncInterval) : 0;
for (T elem : elems) {
writer.append(elem);
recordIndex++;
switch (syncBehavior) {
case SYNC_REGULAR:
if (recordIndex == syncInterval) {
recordIndex = 0;
writer.sync();
}
break;
case SYNC_RANDOM:
if (recordIndex == syncIndex) {
recordIndex = 0;
writer.sync();
syncIndex = random.nextInt(syncInterval);
}
break;
case SYNC_DEFAULT:
default:
}
}
}
return path;
}
public void testWrite() throws IOException {
URL url = this.getClass().getClassLoader().getResource("input/Company.avsc");
assertNotNull(url);
Schema schema = new Schema.Parser().parse(new File(url.getFile()));
assertNotNull(schema);
DatumReader<GenericRecord> datumReader = new GenericDatumReader<GenericRecord>(schema);
// Another way of loading a file
File file = new File("src/test/resources/input/companies.avro");
DataFileReader<GenericRecord> dataFileReader =
new DataFileReader<GenericRecord>(file, datumReader);
File fileOut = new File("target/companies2.avro");
Schema schemaOut =
new Schema.Parser().parse(new File("src/test/resources/input/Company2.avsc"));
DatumWriter<GenericRecord> datumWriter = new GenericDatumWriter<GenericRecord>(schemaOut);
DataFileWriter<GenericRecord> dataFileWriter = new DataFileWriter<GenericRecord>(datumWriter);
GenericRecord company = null;
int count = 0;
while (dataFileReader.hasNext()) {
company = dataFileReader.next(company);
if (company.get("name").toString().equals("aol")) {
dataFileWriter.create(schemaOut, fileOut);
GenericRecord recordOut = new GenericData.Record(schemaOut);
recordOut.put("id", company.get("id"));
recordOut.put("name", company.get("name"));
assertTrue(recordOut.getSchema().getField("address") != null);
assertTrue(recordOut.getSchema().getField("employeeCount") == null);
// address is of complex type
GenericRecord address =
new GenericData.Record((GenericData.Record) company.get("address"), true);
recordOut.put("address", address);
dataFileWriter.append(recordOut);
count++;
}
}
assertTrue(count > 0);
dataFileWriter.close();
}
public static void main(String[] args) throws IOException {
DatumWriter<ArchivePlace> datumWriter =
new SpecificDatumWriter<ArchivePlace>(ArchivePlace.class);
DataFileWriter<ArchivePlace> falloutDatafileWriter =
new DataFileWriter<ArchivePlace>(datumWriter);
FileOutputStream falloutOutputStream =
new FileOutputStream("src/test/resources/archive-places/input.avro", true);
falloutDatafileWriter.create(ArchivePlace.SCHEMA$, falloutOutputStream);
List<ArchivePlace> places =
SerializationUtil.loadFromJsons(
ArchivePlace.SCHEMA$, "src/test/resources/archive-places/input.json");
for (ArchivePlace place : places) {
falloutDatafileWriter.append(place);
falloutDatafileWriter.flush();
}
falloutDatafileWriter.close();
falloutOutputStream.close();
}
private List<Map.Entry> previewAvroBatch(FileStatus fileStatus, int batchSize)
throws IOException, InterruptedException {
SeekableInput input = new FsInput(fileStatus.getPath(), hadoopConf);
DatumReader<GenericRecord> reader = new GenericDatumReader<>();
FileReader<GenericRecord> fileReader = DataFileReader.openReader(input, reader);
List<Map.Entry> batch = new ArrayList<>();
int count = 0;
while (fileReader.hasNext() && batch.size() < batchSize) {
GenericRecord datum = fileReader.next();
ByteArrayOutputStream out = new ByteArrayOutputStream();
DataFileWriter<GenericRecord> dataFileWriter =
new DataFileWriter<GenericRecord>(
new GenericDatumWriter<GenericRecord>(datum.getSchema()));
dataFileWriter.create(datum.getSchema(), out);
dataFileWriter.append(datum);
dataFileWriter.close();
out.close();
batch.add(new Pair(fileStatus.getPath().toUri().getPath() + "::" + count, out.toByteArray()));
count++;
}
return batch;
}
public void serializeSpecific() throws IOException {
// Create a datum to serialize.
MyPair datum = new MyPair();
datum.left = new Utf8("dog");
datum.right = new Utf8("cat");
File tmpFile = File.createTempFile("myPairAvroExample", ".avro");
// Serialize it.
DataFileWriter<MyPair> writer =
new DataFileWriter<MyPair>(new SpecificDatumWriter<MyPair>(MyPair.class));
writer.create(MyPair.SCHEMA$, tmpFile);
writer.append(datum);
writer.close();
System.out.println("Serialization: " + tmpFile);
// Deserialize it.
FileReader<MyPair> reader =
DataFileReader.openReader(tmpFile, new SpecificDatumReader<MyPair>(MyPair.class));
while (reader.hasNext()) {
MyPair result = reader.next();
System.out.printf("Left: %s, Right: %s\n", result.left, result.right);
}
reader.close();
}
@Override
public void writeRecord(E record) throws IOException {
dataFileWriter.append(record);
}
@Override
public void write(T value) throws Exception {
dataFileWriter.append(value);
}