Java tutorial
/** * 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.tools; import java.io.ByteArrayOutputStream; import java.io.IOException; import java.io.PrintStream; import java.net.URI; import java.util.ArrayList; import java.util.Collections; import java.util.List; import java.util.StringTokenizer; import junit.framework.TestCase; import org.apache.commons.logging.LogFactory; import org.apache.commons.logging.impl.Log4JLogger; import org.apache.hadoop.conf.Configuration; import org.apache.hadoop.fs.FSDataOutputStream; import org.apache.hadoop.fs.FileSystem; import org.apache.hadoop.fs.FsShell; import org.apache.hadoop.fs.Path; import org.apache.hadoop.hdfs.MiniDFSCluster; import org.apache.hadoop.io.IOUtils; import org.apache.hadoop.mapred.MiniMRCluster; import org.apache.hadoop.util.ToolRunner; import org.apache.log4j.Level; /** * test {@link HadoopArchives} */ public class TestHadoopArchives extends TestCase { private static final String inputDir = "input"; private Path inputPath; private MiniDFSCluster dfscluster; private MiniMRCluster mapred; private FileSystem fs; private Path archivePath; static private Path createFile(Path dir, String filename, FileSystem fs) throws IOException { final Path f = new Path(dir, filename); final FSDataOutputStream out = fs.create(f); out.write(filename.getBytes()); out.close(); return f; } protected void setUp() throws Exception { super.setUp(); dfscluster = new MiniDFSCluster(new Configuration(), 2, true, null); fs = dfscluster.getFileSystem(); mapred = new MiniMRCluster(2, fs.getUri().toString(), 1); inputPath = new Path(fs.getHomeDirectory(), inputDir); archivePath = new Path(fs.getHomeDirectory(), "archive"); fs.mkdirs(inputPath); createFile(inputPath, "a", fs); createFile(inputPath, "b", fs); createFile(inputPath, "c", fs); } protected void tearDown() throws Exception { try { if (mapred != null) { mapred.shutdown(); } if (dfscluster != null) { dfscluster.shutdown(); } } catch (Exception e) { System.err.println(e); } super.tearDown(); } public void testPathWithSpaces() throws Exception { fs.delete(archivePath, true); //create files/directories with spaces createFile(inputPath, "c c", fs); final Path sub1 = new Path(inputPath, "sub 1"); fs.mkdirs(sub1); createFile(sub1, "file x y z", fs); createFile(sub1, "file", fs); createFile(sub1, "x", fs); createFile(sub1, "y", fs); createFile(sub1, "z", fs); final Path sub2 = new Path(inputPath, "sub 1 with suffix"); fs.mkdirs(sub2); createFile(sub2, "z", fs); final Configuration conf = mapred.createJobConf(); final FsShell shell = new FsShell(conf); final String inputPathStr = inputPath.toUri().getPath(); final List<String> originalPaths = lsr(shell, inputPathStr); final URI uri = fs.getUri(); final String prefix = "har://hdfs-" + uri.getHost() + ":" + uri.getPort() + archivePath.toUri().getPath() + Path.SEPARATOR; {//Enable space replacement final String harName = "foo.har"; final String[] args = { "-archiveName", harName, "-p", inputPathStr, "*", archivePath.toString() }; final HadoopArchives har = new HadoopArchives(mapred.createJobConf()); assertEquals(0, ToolRunner.run(har, args)); //compare results final List<String> harPaths = lsr(shell, prefix + harName); assertEquals(originalPaths, harPaths); } } private static List<String> lsr(final FsShell shell, String dir) throws Exception { final ByteArrayOutputStream bytes = new ByteArrayOutputStream(); final PrintStream out = new PrintStream(bytes); final PrintStream oldOut = System.out; final PrintStream oldErr = System.err; System.setOut(out); System.setErr(out); final String results; try { assertEquals(0, shell.run(new String[] { "-lsr", dir })); results = bytes.toString(); } finally { IOUtils.closeStream(out); System.setOut(oldOut); System.setErr(oldErr); } final String dirname = dir.substring(dir.lastIndexOf(Path.SEPARATOR)); final List<String> paths = new ArrayList<String>(); for (StringTokenizer t = new StringTokenizer(results, "\n"); t.hasMoreTokens();) { final String s = t.nextToken(); final int i = s.indexOf(dirname); if (i >= 0) { paths.add(s.substring(i + dirname.length())); } } Collections.sort(paths); return paths; } }