@Override
public void open(int taskNumber, int numTasks) throws IOException {
super.open(taskNumber, numTasks);
DatumWriter<E> datumWriter;
Schema schema;
if (org.apache.avro.specific.SpecificRecordBase.class.isAssignableFrom(avroValueType)) {
datumWriter = new SpecificDatumWriter<E>(avroValueType);
try {
schema =
((org.apache.avro.specific.SpecificRecordBase) avroValueType.newInstance()).getSchema();
} catch (InstantiationException e) {
throw new RuntimeException(e.getMessage());
} catch (IllegalAccessException e) {
throw new RuntimeException(e.getMessage());
}
} else {
datumWriter = new ReflectDatumWriter<E>(avroValueType);
schema = ReflectData.get().getSchema(avroValueType);
}
dataFileWriter = new DataFileWriter<E>(datumWriter);
if (userDefinedSchema == null) {
dataFileWriter.create(schema, stream);
} else {
dataFileWriter.create(userDefinedSchema, stream);
}
}
/**
* Builds the target file path as <datasource directory>/<month>.avro. If the target file already
* exists, then it is open for appending, otherwise it is created
*/
private DataFileWriter<GenericRecord> openHDFSFile(DatasourceMonth datasourceMonth)
throws IOException {
DataFileWriter<GenericRecord> writer =
new DataFileWriter<GenericRecord>(new GenericDatumWriter<GenericRecord>(AVRO_SCHEMA));
writer.setSyncInterval(FILEWRITER_SYNC_INTERVAL);
// writer.setCodec(CodecFactory.snappyCodec()); // omit for now
Path targetPath = buildTargetPath(datasourceMonth);
// just for logging
String fullTargetPath = this.hdfs.getWorkingDirectory() + "/" + targetPath;
// Append to an existing file, or create a new file is file otherwise
if (this.hdfs.exists(targetPath)) {
// appending to an existing file
// based on
// http://technicaltidbit.blogspot.com.es/2013/02/avro-can-append-in-hdfs-after-all.html
if (debugMode) {
this.hdfs.setReplication(targetPath, (short) 1);
}
LOGGER.info("Appending to existing file {}", fullTargetPath);
OutputStream outputStream = this.hdfs.append(targetPath);
writer.appendTo(new FsInput(targetPath, this.hadoopConf), outputStream);
} else {
// creating a new file
LOGGER.info(
"Creating new file " + fullTargetPath + " for datasource {} and month {}",
datasourceMonth.datasource(),
datasourceMonth.month());
OutputStream outputStream = this.hdfs.create(targetPath);
writer.create(AVRO_SCHEMA, outputStream);
}
return writer;
}
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);
}
}
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 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;
}
@Override
public void prepare(Map conf, TridentOperationContext context) {
try {
String path = (String) conf.get("DOCUMENT_PATH");
schema = Schema.parse(PersistDocumentFunction.class.getResourceAsStream("/document.avsc"));
File file = new File(path);
DatumWriter<GenericRecord> datumWriter = new GenericDatumWriter<GenericRecord>(schema);
dataFileWriter = new DataFileWriter<GenericRecord>(datumWriter);
if (file.exists()) dataFileWriter.appendTo(file);
else dataFileWriter.create(schema, file);
} catch (IOException e) {
throw new RuntimeException(e);
}
}
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 {
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();
}
}
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();
}
@SuppressWarnings("deprecation") // uses internal test functionality.
@Override
protected void prepareWrite(WritableByteChannel channel) throws Exception {
dataFileWriter = new DataFileWriter<>(coder.createDatumWriter());
dataFileWriter.create(coder.getSchema(), Channels.newOutputStream(channel));
}