/**
* 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;
}
@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);
}
}