2010-03-29 19:17:55 -04:00
|
|
|
package spark
|
|
|
|
|
|
|
|
import java.io._
|
2010-11-09 16:46:30 -05:00
|
|
|
import java.net.InetAddress
|
2010-10-16 19:14:13 -04:00
|
|
|
import java.util.UUID
|
2011-04-28 01:13:01 -04:00
|
|
|
import java.util.concurrent.{Executors, ThreadFactory, ThreadPoolExecutor}
|
2010-03-29 19:17:55 -04:00
|
|
|
|
2010-10-04 15:01:05 -04:00
|
|
|
import scala.collection.mutable.ArrayBuffer
|
2010-11-04 01:45:44 -04:00
|
|
|
import scala.util.Random
|
2010-10-04 15:01:05 -04:00
|
|
|
|
2010-10-16 19:14:13 -04:00
|
|
|
/**
|
|
|
|
* Various utility methods used by Spark.
|
|
|
|
*/
|
2010-10-04 15:01:05 -04:00
|
|
|
object Utils {
|
2010-03-29 19:17:55 -04:00
|
|
|
|
2012-04-10 00:59:56 -04:00
|
|
|
// The serializer in this object is used by Spark to serialize closures.
|
|
|
|
val serializerClass = System.getProperty("spark.closure.serializer", "spark.JavaSerializer")
|
|
|
|
val ser = Class.forName(serializerClass).newInstance().asInstanceOf[Serializer]
|
|
|
|
|
|
|
|
def serialize[T](o: T): Array[Byte] = ser.newInstance().serialize[T](o)
|
|
|
|
|
|
|
|
def deserialize[T](bytes: Array[Byte]): T = ser.newInstance().deserialize[T](bytes)
|
2010-03-29 19:17:55 -04:00
|
|
|
|
|
|
|
def deserialize[T](bytes: Array[Byte], loader: ClassLoader): T = {
|
2012-04-10 00:59:56 -04:00
|
|
|
ser.newInstance().deserialize[T](bytes, loader)
|
2010-03-29 19:17:55 -04:00
|
|
|
}
|
2010-10-04 15:01:05 -04:00
|
|
|
|
2012-02-09 18:50:26 -05:00
|
|
|
def isAlpha(c: Char): Boolean = {
|
2010-10-04 15:01:05 -04:00
|
|
|
(c >= 'A' && c <= 'Z') || (c >= 'a' && c <= 'z')
|
|
|
|
}
|
|
|
|
|
|
|
|
def splitWords(s: String): Seq[String] = {
|
|
|
|
val buf = new ArrayBuffer[String]
|
|
|
|
var i = 0
|
|
|
|
while (i < s.length) {
|
|
|
|
var j = i
|
|
|
|
while (j < s.length && isAlpha(s.charAt(j))) {
|
|
|
|
j += 1
|
|
|
|
}
|
|
|
|
if (j > i) {
|
|
|
|
buf += s.substring(i, j);
|
|
|
|
}
|
|
|
|
i = j
|
|
|
|
while (i < s.length && !isAlpha(s.charAt(i))) {
|
|
|
|
i += 1
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return buf
|
|
|
|
}
|
2010-10-16 19:14:13 -04:00
|
|
|
|
|
|
|
// Create a temporary directory inside the given parent directory
|
2012-02-10 11:19:53 -05:00
|
|
|
def createTempDir(root: String = System.getProperty("java.io.tmpdir")): File = {
|
2010-10-16 19:14:13 -04:00
|
|
|
var attempts = 0
|
|
|
|
val maxAttempts = 10
|
|
|
|
var dir: File = null
|
|
|
|
while (dir == null) {
|
|
|
|
attempts += 1
|
|
|
|
if (attempts > maxAttempts) {
|
2012-02-10 11:19:53 -05:00
|
|
|
throw new IOException("Failed to create a temp directory after " + maxAttempts +
|
|
|
|
" attempts!")
|
2010-10-16 19:14:13 -04:00
|
|
|
}
|
|
|
|
try {
|
|
|
|
dir = new File(root, "spark-" + UUID.randomUUID.toString)
|
|
|
|
if (dir.exists() || !dir.mkdirs()) {
|
|
|
|
dir = null
|
|
|
|
}
|
|
|
|
} catch { case e: IOException => ; }
|
|
|
|
}
|
|
|
|
return dir
|
|
|
|
}
|
|
|
|
|
|
|
|
// Copy all data from an InputStream to an OutputStream
|
|
|
|
def copyStream(in: InputStream,
|
|
|
|
out: OutputStream,
|
|
|
|
closeStreams: Boolean = false)
|
|
|
|
{
|
|
|
|
val buf = new Array[Byte](8192)
|
|
|
|
var n = 0
|
|
|
|
while (n != -1) {
|
|
|
|
n = in.read(buf)
|
|
|
|
if (n != -1) {
|
|
|
|
out.write(buf, 0, n)
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (closeStreams) {
|
|
|
|
in.close()
|
|
|
|
out.close()
|
|
|
|
}
|
|
|
|
}
|
2010-11-04 01:45:44 -04:00
|
|
|
|
|
|
|
// Shuffle the elements of a collection into a random order, returning the
|
|
|
|
// result in a new collection. Unlike scala.util.Random.shuffle, this method
|
|
|
|
// uses a local random number generator, avoiding inter-thread contention.
|
2011-05-20 03:19:53 -04:00
|
|
|
def randomize[T](seq: TraversableOnce[T]): Seq[T] = {
|
2010-11-08 03:45:02 -05:00
|
|
|
val buf = new ArrayBuffer[T]()
|
|
|
|
buf ++= seq
|
2010-11-04 01:45:44 -04:00
|
|
|
val rand = new Random()
|
|
|
|
for (i <- (buf.size - 1) to 1 by -1) {
|
|
|
|
val j = rand.nextInt(i)
|
|
|
|
val tmp = buf(j)
|
|
|
|
buf(j) = buf(i)
|
|
|
|
buf(i) = tmp
|
|
|
|
}
|
|
|
|
buf
|
|
|
|
}
|
2010-11-09 16:46:30 -05:00
|
|
|
|
|
|
|
/**
|
2012-02-10 01:58:24 -05:00
|
|
|
* Get the local host's IP address in dotted-quad format (e.g. 1.2.3.4).
|
2010-11-09 16:46:30 -05:00
|
|
|
*/
|
2011-04-28 01:13:01 -04:00
|
|
|
def localIpAddress(): String = InetAddress.getLocalHost.getHostAddress
|
|
|
|
|
|
|
|
/**
|
2012-02-10 01:58:24 -05:00
|
|
|
* Returns a standard ThreadFactory except all threads are daemons.
|
2011-04-28 01:13:01 -04:00
|
|
|
*/
|
|
|
|
private def newDaemonThreadFactory: ThreadFactory = {
|
|
|
|
new ThreadFactory {
|
|
|
|
def newThread(r: Runnable): Thread = {
|
|
|
|
var t = Executors.defaultThreadFactory.newThread (r)
|
|
|
|
t.setDaemon (true)
|
|
|
|
return t
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2012-02-10 01:58:24 -05:00
|
|
|
* Wrapper over newCachedThreadPool.
|
2011-04-28 01:13:01 -04:00
|
|
|
*/
|
|
|
|
def newDaemonCachedThreadPool(): ThreadPoolExecutor = {
|
2012-02-10 11:19:53 -05:00
|
|
|
var threadPool = Executors.newCachedThreadPool.asInstanceOf[ThreadPoolExecutor]
|
2011-04-28 01:13:01 -04:00
|
|
|
|
|
|
|
threadPool.setThreadFactory (newDaemonThreadFactory)
|
|
|
|
|
|
|
|
return threadPool
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2012-02-10 01:58:24 -05:00
|
|
|
* Wrapper over newFixedThreadPool.
|
2011-04-28 01:13:01 -04:00
|
|
|
*/
|
|
|
|
def newDaemonFixedThreadPool(nThreads: Int): ThreadPoolExecutor = {
|
2012-02-10 11:19:53 -05:00
|
|
|
var threadPool = Executors.newFixedThreadPool(nThreads).asInstanceOf[ThreadPoolExecutor]
|
2011-04-28 01:13:01 -04:00
|
|
|
|
|
|
|
threadPool.setThreadFactory(newDaemonThreadFactory)
|
|
|
|
|
|
|
|
return threadPool
|
2010-11-09 16:46:30 -05:00
|
|
|
}
|
2011-02-27 22:15:52 -05:00
|
|
|
|
|
|
|
/**
|
2012-02-10 01:58:24 -05:00
|
|
|
* Get the local machine's hostname.
|
2011-02-27 22:15:52 -05:00
|
|
|
*/
|
|
|
|
def localHostName(): String = {
|
|
|
|
return InetAddress.getLocalHost().getHostName
|
|
|
|
}
|
2012-02-10 01:58:24 -05:00
|
|
|
|
|
|
|
/**
|
|
|
|
* Delete a file or directory and its contents recursively.
|
|
|
|
*/
|
|
|
|
def deleteRecursively(file: File) {
|
|
|
|
if (file.isDirectory) {
|
|
|
|
for (child <- file.listFiles()) {
|
|
|
|
deleteRecursively(child)
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (!file.delete()) {
|
|
|
|
throw new IOException("Failed to delete: " + file)
|
|
|
|
}
|
|
|
|
}
|
2010-03-29 19:17:55 -04:00
|
|
|
}
|