HADOOP-6346. Add support for specifying unpack pattern regex to RunJar.unJar. Contributed by Todd Lipcon.

git-svn-id: https://svn.apache.org/repos/asf/hadoop/common/trunk@889018 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Thomas White 2009-12-09 23:25:51 +00:00
parent 2e5fd1c0ab
commit a473f37733
5 changed files with 243 additions and 36 deletions

View File

@ -43,6 +43,9 @@ Trunk (unreleased changes)
HADOOP-6400. Log errors getting Unix UGI. (Todd Lipcon via tomwhite)
HADOOP-6346. Add support for specifying unpack pattern regex to
RunJar.unJar. (Todd Lipcon via tomwhite)
OPTIMIZATIONS
BUG FIXES

View File

@ -47,6 +47,7 @@ import java.util.WeakHashMap;
import java.util.concurrent.CopyOnWriteArrayList;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
import java.util.regex.PatternSyntaxException;
import javax.xml.parsers.DocumentBuilder;
import javax.xml.parsers.DocumentBuilderFactory;
@ -831,6 +832,45 @@ public class Configuration implements Iterable<Map.Entry<String,String>>,
: Enum.valueOf(defaultValue.getDeclaringClass(), val);
}
/**
* Get the value of the <code>name</code> property as a <code>Pattern</code>.
* If no such property is specified, or if the specified value is not a valid
* <code>Pattern</code>, then <code>DefaultValue</code> is returned.
*
* @param name property name
* @param defaultValue default value
* @return property value as a compiled Pattern, or defaultValue
*/
public Pattern getPattern(String name, Pattern defaultValue) {
String valString = get(name);
if (null == valString || "".equals(valString)) {
return defaultValue;
}
try {
return Pattern.compile(valString);
} catch (PatternSyntaxException pse) {
LOG.warn("Regular expression '" + valString + "' for property '" +
name + "' not valid. Using default", pse);
return defaultValue;
}
}
/**
* Set the given property to <code>Pattern</code>.
* If the pattern is passed as null, sets the empty pattern which results in
* further calls to getPattern(...) returning the default value.
*
* @param name property name
* @param pattern new value
*/
public void setPattern(String name, Pattern pattern) {
if (null == pattern) {
set(name, null);
} else {
set(name, pattern.pattern());
}
}
/**
* A class that represents a set of positive integer ranges. It parses
* strings of the form: "2-3,5,7-" where ranges are separated by comma and

View File

@ -18,43 +18,66 @@
package org.apache.hadoop.util;
import java.util.jar.*;
import java.lang.reflect.*;
import java.lang.reflect.Array;
import java.lang.reflect.Method;
import java.lang.reflect.InvocationTargetException;
import java.net.URL;
import java.net.URLClassLoader;
import java.io.*;
import java.util.*;
import java.io.FileOutputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
import java.io.File;
import java.util.regex.Pattern;
import java.util.Arrays;
import java.util.ArrayList;
import java.util.Enumeration;
import java.util.jar.JarFile;
import java.util.jar.JarEntry;
import java.util.jar.Manifest;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileUtil;
import org.apache.hadoop.io.IOUtils;
/** Run a Hadoop job jar. */
public class RunJar {
/** Unpack a jar file into a directory. */
/** Pattern that matches any string */
public static final Pattern MATCH_ANY = Pattern.compile(".*");
/**
* Unpack a jar file into a directory.
*
* This version unpacks all files inside the jar regardless of filename.
*/
public static void unJar(File jarFile, File toDir) throws IOException {
unJar(jarFile, toDir, MATCH_ANY);
}
/**
* Unpack matching files from a jar. Entries inside the jar that do
* not match the given pattern will be skipped.
*
* @param jarFile the .jar file to unpack
* @param toDir the destination directory into which to unpack the jar
* @param unpackRegex the pattern to match jar entries against
*/
public static void unJar(File jarFile, File toDir, Pattern unpackRegex)
throws IOException {
JarFile jar = new JarFile(jarFile);
try {
Enumeration entries = jar.entries();
Enumeration<JarEntry> entries = jar.entries();
while (entries.hasMoreElements()) {
JarEntry entry = (JarEntry)entries.nextElement();
if (!entry.isDirectory()) {
if (!entry.isDirectory() &&
unpackRegex.matcher(entry.getName()).matches()) {
InputStream in = jar.getInputStream(entry);
try {
File file = new File(toDir, entry.getName());
if (!file.getParentFile().mkdirs()) {
if (!file.getParentFile().isDirectory()) {
throw new IOException("Mkdirs failed to create " +
file.getParentFile().toString());
}
}
ensureDirectory(file.getParentFile());
OutputStream out = new FileOutputStream(file);
try {
byte[] buffer = new byte[8192];
int i;
while ((i = in.read(buffer)) != -1) {
out.write(buffer, 0, i);
}
IOUtils.copyBytes(in, out, 8192);
} finally {
out.close();
}
@ -68,6 +91,18 @@ public class RunJar {
}
}
/**
* Ensure the existence of a given directory.
*
* @throws IOException if it cannot be created and does not already exist
*/
private static void ensureDirectory(File dir) throws IOException {
if (!dir.mkdirs() && !dir.isDirectory()) {
throw new IOException("Mkdirs failed to create " +
dir.toString());
}
}
/** Run a Hadoop job jar. If the main class is not in the jar's manifest,
* then it must be provided on the command line. */
public static void main(String[] args) throws Throwable {
@ -107,22 +142,14 @@ public class RunJar {
mainClassName = mainClassName.replaceAll("/", ".");
File tmpDir = new File(new Configuration().get("hadoop.tmp.dir"));
boolean b = tmpDir.mkdirs();
if (!b && !tmpDir.isDirectory()) {
System.err.println("Mkdirs failed to create " + tmpDir);
System.exit(-1);
}
ensureDirectory(tmpDir);
final File workDir = File.createTempFile("hadoop-unjar", "", tmpDir);
b = workDir.delete();
if (!b) {
if (!workDir.delete()) {
System.err.println("Delete failed for " + workDir);
System.exit(-1);
}
b = workDir.mkdirs();
if (!b && !workDir.isDirectory()) {
System.err.println("Mkdirs failed to create " + workDir);
System.exit(-1);
}
ensureDirectory(workDir);
Runtime.getRuntime().addShutdownHook(new Thread() {
public void run() {
@ -134,15 +161,15 @@ public class RunJar {
});
unJar(file, workDir);
ArrayList<URL> classPath = new ArrayList<URL>();
classPath.add(new File(workDir+"/").toURL());
classPath.add(file.toURL());
classPath.add(new File(workDir, "classes/").toURL());
classPath.add(new File(workDir+"/").toURI().toURL());
classPath.add(file.toURI().toURL());
classPath.add(new File(workDir, "classes/").toURI().toURL());
File[] libs = new File(workDir, "lib").listFiles();
if (libs != null) {
for (int i = 0; i < libs.length; i++) {
classPath.add(libs[i].toURL());
classPath.add(libs[i].toURI().toURL());
}
}

View File

@ -25,6 +25,7 @@ import java.io.StringWriter;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.Random;
import java.util.regex.Pattern;
import junit.framework.TestCase;
@ -364,6 +365,33 @@ public class TestConfiguration extends TestCase {
assertTrue(fail);
}
public void testPattern() throws IOException {
out = new BufferedWriter(new FileWriter(CONFIG));
startConfig();
appendProperty("test.pattern1", "");
appendProperty("test.pattern2", "(");
appendProperty("test.pattern3", "a+b");
endConfig();
Path fileResource = new Path(CONFIG);
conf.addResource(fileResource);
Pattern defaultPattern = Pattern.compile("x+");
// Return default if missing
assertEquals(defaultPattern.pattern(),
conf.getPattern("xxxxx", defaultPattern).pattern());
// Return null if empty and default is null
assertNull(conf.getPattern("test.pattern1", null));
// Return default for empty
assertEquals(defaultPattern.pattern(),
conf.getPattern("test.pattern1", defaultPattern).pattern());
// Return default for malformed
assertEquals(defaultPattern.pattern(),
conf.getPattern("test.pattern2", defaultPattern).pattern());
// Works for correct patterns
assertEquals("a+b",
conf.getPattern("test.pattern3", defaultPattern).pattern());
}
public void testReload() throws IOException {
out=new BufferedWriter(new FileWriter(CONFIG));
startConfig();

View File

@ -0,0 +1,109 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.util;
import junit.framework.TestCase;
import java.io.File;
import java.io.FileOutputStream;
import java.io.IOException;
import java.util.jar.JarOutputStream;
import java.util.regex.Pattern;
import java.util.zip.ZipEntry;
import org.junit.After;
import org.junit.Before;
import org.junit.Test;
import org.apache.hadoop.fs.FileUtil;
public class TestRunJar extends TestCase {
private File TEST_ROOT_DIR;
private static final String TEST_JAR_NAME="test-runjar.jar";
@Before
protected void setUp()
throws Exception {
TEST_ROOT_DIR =
new File(System.getProperty("test.build.data", "/tmp"), getClass()
.getSimpleName());
if (!TEST_ROOT_DIR.exists()) {
TEST_ROOT_DIR.mkdirs();
}
makeTestJar();
}
@After
protected void tearDown()
throws Exception {
FileUtil.fullyDelete(TEST_ROOT_DIR);
}
/**
* Construct a jar with two files in it in our
* test dir.
*/
private void makeTestJar() throws IOException {
File jarFile = new File(TEST_ROOT_DIR, TEST_JAR_NAME);
JarOutputStream jstream =
new JarOutputStream(new FileOutputStream(jarFile));
jstream.putNextEntry(new ZipEntry("foobar.txt"));
jstream.closeEntry();
jstream.putNextEntry(new ZipEntry("foobaz.txt"));
jstream.closeEntry();
jstream.close();
}
/**
* Test default unjarring behavior - unpack everything
*/
@Test
public void testUnJar() throws Exception {
File unjarDir = new File(TEST_ROOT_DIR, "unjar-all");
assertFalse("unjar dir shouldn't exist at test start",
new File(unjarDir, "foobar.txt").exists());
// Unjar everything
RunJar.unJar(new File(TEST_ROOT_DIR, TEST_JAR_NAME),
unjarDir);
assertTrue("foobar unpacked",
new File(unjarDir, "foobar.txt").exists());
assertTrue("foobaz unpacked",
new File(unjarDir, "foobaz.txt").exists());
}
/**
* Test unjarring a specific regex
*/
public void testUnJarWithPattern() throws Exception {
File unjarDir = new File(TEST_ROOT_DIR, "unjar-pattern");
assertFalse("unjar dir shouldn't exist at test start",
new File(unjarDir, "foobar.txt").exists());
// Unjar only a regex
RunJar.unJar(new File(TEST_ROOT_DIR, TEST_JAR_NAME),
unjarDir,
Pattern.compile(".*baz.*"));
assertFalse("foobar not unpacked",
new File(unjarDir, "foobar.txt").exists());
assertTrue("foobaz unpacked",
new File(unjarDir, "foobaz.txt").exists());
}
}