Java FileSystem.getFileStatus方法代码示例

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
@SuppressWarnings("deprecation")
public void testBlockSize() throws Exception {
  FileSystem fs = getFileSystem();
  long defaultBlockSize = fs.getDefaultBlockSize();
  assertEquals("incorrect blocksize",
      S3AFileSystem.DEFAULT_BLOCKSIZE, defaultBlockSize);
  long newBlockSize = defaultBlockSize * 2;
  fs.getConf().setLong(Constants.FS_S3A_BLOCK_SIZE, newBlockSize);

  Path dir = path("testBlockSize");
  Path file = new Path(dir, "file");
  createFile(fs, file, true, dataset(1024, 'a', 'z' - 'a'));
  FileStatus fileStatus = fs.getFileStatus(file);
  assertEquals("Double default block size in stat(): " + fileStatus,
      newBlockSize,
      fileStatus.getBlockSize());

  // check the listing  & assert that the block size is picked up by
  // this route too.
  boolean found = false;
  FileStatus[] listing = fs.listStatus(dir);
  for (FileStatus stat : listing) {
    LOG.info("entry: {}", stat);
    if (file.equals(stat.getPath())) {
      found = true;
      assertEquals("Double default block size in ls(): " + stat,
          newBlockSize,
          stat.getBlockSize());
    }
  }
  assertTrue("Did not find " + fileStatsToString(listing, ", "), found);
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
public void addResource(FileSystem fs, Configuration conf, Path destPath,
    Map<String, LocalResource> localResources, LocalResourceType resourceType, String link,
    Map<URI, FileStatus> statCache, boolean appMasterOnly) throws IOException {

  FileStatus destStatus = fs.getFileStatus(destPath);
  LocalResource amJarRsrc = Records.newRecord(LocalResource.class);
  amJarRsrc.setType(resourceType);

  LocalResourceVisibility visibility = getVisibility(conf, destPath.toUri(), statCache);
  amJarRsrc.setVisibility(visibility);
  amJarRsrc.setResource(ConverterUtils.getYarnUrlFromPath(destPath));
  amJarRsrc.setTimestamp(destStatus.getModificationTime());
  amJarRsrc.setSize(destStatus.getLen());

  if (link == null || link.isEmpty())
    throw new IOException("You must specify a valid link name");

  localResources.put(link, amJarRsrc);
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private static BlockLocation[] getBlockLocationsOutput(int fileSize,
    int blockSize, long start, long len, String blockLocationHost)
    throws Exception {
  Configuration conf = new Configuration();
  conf.set(NativeAzureFileSystem.AZURE_BLOCK_SIZE_PROPERTY_NAME, ""
      + blockSize);
  if (blockLocationHost != null) {
    conf.set(NativeAzureFileSystem.AZURE_BLOCK_LOCATION_HOST_PROPERTY_NAME,
        blockLocationHost);
  }
  AzureBlobStorageTestAccount testAccount = AzureBlobStorageTestAccount
      .createMock(conf);
  FileSystem fs = testAccount.getFileSystem();
  Path testFile = createTestFile(fs, fileSize);
  FileStatus stat = fs.getFileStatus(testFile);
  BlockLocation[] locations = fs.getFileBlockLocations(stat, start, len);
  testAccount.cleanup();
  return locations;
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private void testCreate(Path path, boolean override) throws Exception {
  FileSystem fs = getHttpFSFileSystem();
  FsPermission permission = new FsPermission(FsAction.READ_WRITE, FsAction.NONE, FsAction.NONE);
  OutputStream os = fs.create(new Path(path.toUri().getPath()), permission, override, 1024,
                              (short) 2, 100 * 1024 * 1024, null);
  os.write(1);
  os.close();
  fs.close();

  fs = FileSystem.get(getProxiedFSConf());
  FileStatus status = fs.getFileStatus(path);
  if (!isLocalFS()) {
    Assert.assertEquals(status.getReplication(), 2);
    Assert.assertEquals(status.getBlockSize(), 100 * 1024 * 1024);
  }
  Assert.assertEquals(status.getPermission(), permission);
  InputStream is = fs.open(path);
  Assert.assertEquals(is.read(), 1);
  is.close();
  fs.close();
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Override
public List<InputSplit> getSplits(JobContext jobCtxt) throws IOException {
  final JobConf jobConf = new JobConf(jobCtxt.getConfiguration());
  final JobClient client = new JobClient(jobConf);
  ClusterStatus stat = client.getClusterStatus(true);
  int numTrackers = stat.getTaskTrackers();
  final int fileCount = jobConf.getInt(GRIDMIX_DISTCACHE_FILE_COUNT, -1);

  // Total size of distributed cache files to be generated
  final long totalSize = jobConf.getLong(GRIDMIX_DISTCACHE_BYTE_COUNT, -1);
  // Get the path of the special file
  String distCacheFileList = jobConf.get(GRIDMIX_DISTCACHE_FILE_LIST);
  if (fileCount < 0 || totalSize < 0 || distCacheFileList == null) {
    throw new RuntimeException("Invalid metadata: #files (" + fileCount
        + "), total_size (" + totalSize + "), filelisturi ("
        + distCacheFileList + ")");
  }

  Path sequenceFile = new Path(distCacheFileList);
  FileSystem fs = sequenceFile.getFileSystem(jobConf);
  FileStatus srcst = fs.getFileStatus(sequenceFile);
  // Consider the number of TTs * mapSlotsPerTracker as number of mappers.
  int numMapSlotsPerTracker = jobConf.getInt(TTConfig.TT_MAP_SLOTS, 2);
  int numSplits = numTrackers * numMapSlotsPerTracker;

  List<InputSplit> splits = new ArrayList<InputSplit>(numSplits);
  LongWritable key = new LongWritable();
  BytesWritable value = new BytesWritable();

  // Average size of data to be generated by each map task
  final long targetSize = Math.max(totalSize / numSplits,
                            DistributedCacheEmulator.AVG_BYTES_PER_MAP);
  long splitStartPosition = 0L;
  long splitEndPosition = 0L;
  long acc = 0L;
  long bytesRemaining = srcst.getLen();
  SequenceFile.Reader reader = null;
  try {
    reader = new SequenceFile.Reader(fs, sequenceFile, jobConf);
    while (reader.next(key, value)) {

      // If adding this file would put this split past the target size,
      // cut the last split and put this file in the next split.
      if (acc + key.get() > targetSize && acc != 0) {
        long splitSize = splitEndPosition - splitStartPosition;
        splits.add(new FileSplit(
            sequenceFile, splitStartPosition, splitSize, (String[])null));
        bytesRemaining -= splitSize;
        splitStartPosition = splitEndPosition;
        acc = 0L;
      }
      acc += key.get();
      splitEndPosition = reader.getPosition();
    }
  } finally {
    if (reader != null) {
      reader.close();
    }
  }
  if (bytesRemaining != 0) {
    splits.add(new FileSplit(
        sequenceFile, splitStartPosition, bytesRemaining, (String[])null));
  }

  return splits;
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private Path copy(Path sCopy, Path dstdir) throws IOException {
  FileSystem sourceFs = sCopy.getFileSystem(conf);
  Path dCopy = new Path(dstdir, "tmp_"+sCopy.getName());
  FileStatus sStat = sourceFs.getFileStatus(sCopy);
  if (sStat.getModificationTime() != resource.getTimestamp()) {
    throw new IOException("Resource " + sCopy +
        " changed on src filesystem (expected " + resource.getTimestamp() +
        ", was " + sStat.getModificationTime());
  }
  if (resource.getVisibility() == LocalResourceVisibility.PUBLIC) {
    if (!isPublic(sourceFs, sCopy, sStat, statCache)) {
      throw new IOException("Resource " + sCopy +
          " is not publicly accessable and as such cannot be part of the" +
          " public cache.");
    }
  }

  FileUtil.copy(sourceFs, sStat, FileSystem.getLocal(conf), dCopy, false,
      true, conf);
  return dCopy;
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private void verifyCopy(FileSystem fs, boolean preserveChecksum)
    throws Exception {
  for (Path path : pathList) {
    final Path targetPath = new Path(path.toString().replaceAll(SOURCE_PATH,
        TARGET_PATH));
    Assert.assertTrue(fs.exists(targetPath));
    Assert.assertTrue(fs.isFile(targetPath) == fs.isFile(path));
    FileStatus sourceStatus = fs.getFileStatus(path);
    FileStatus targetStatus = fs.getFileStatus(targetPath);
    Assert.assertEquals(sourceStatus.getReplication(),
        targetStatus.getReplication());
    if (preserveChecksum) {
      Assert.assertEquals(sourceStatus.getBlockSize(),
          targetStatus.getBlockSize());
    }
    Assert.assertTrue(!fs.isFile(targetPath)
        || fs.getFileChecksum(targetPath).equals(fs.getFileChecksum(path)));
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
/**
 * Load from a Hadoop filesystem
 * @param fs filesystem
 * @param path path
 * @return a loaded CD
 * @throws IOException IO problems
 * @throws EOFException if not enough bytes were read in
 * @throws JsonParseException parse problems
 * @throws JsonMappingException O/J mapping problems
 */
public T load(FileSystem fs, Path path)
    throws IOException, JsonParseException, JsonMappingException {
  FileStatus status = fs.getFileStatus(path);
  long len = status.getLen();
  byte[] b = new byte[(int) len];
  FSDataInputStream dataInputStream = fs.open(path);
  int count = dataInputStream.read(b);
  if (count != len) {
    throw new EOFException(path.toString() + ": read finished prematurely");
  }
  return fromBytes(path.toString(), b);
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testPreserveGroupOnFile() throws IOException {
  FileSystem fs = FileSystem.get(config);
  EnumSet<FileAttribute> attributes = EnumSet.of(FileAttribute.GROUP);

  Path dst = new Path("/tmp/dest2");
  Path src = new Path("/tmp/src2");

  createFile(fs, src);
  createFile(fs, dst);

  fs.setPermission(src, fullPerm);
  fs.setOwner(src, "somebody", "somebody-group");
  fs.setTimes(src, 0, 0);
  fs.setReplication(src, (short) 1);

  fs.setPermission(dst, noPerm);
  fs.setOwner(dst, "nobody", "nobody-group");
  fs.setTimes(dst, 100, 100);
  fs.setReplication(dst, (short) 2);

  CopyListingFileStatus srcStatus = new CopyListingFileStatus(fs.getFileStatus(src));

  DistCpUtils.preserve(fs, dst, srcStatus, attributes, false);

  CopyListingFileStatus dstStatus = new CopyListingFileStatus(fs.getFileStatus(dst));

  // FileStatus.equals only compares path field, must explicitly compare all fields
  Assert.assertFalse(srcStatus.getPermission().equals(dstStatus.getPermission()));
  Assert.assertFalse(srcStatus.getOwner().equals(dstStatus.getOwner()));
  Assert.assertTrue(srcStatus.getGroup().equals(dstStatus.getGroup()));
  Assert.assertFalse(srcStatus.getAccessTime() == dstStatus.getAccessTime());
  Assert.assertFalse(srcStatus.getModificationTime() == dstStatus.getModificationTime());
  Assert.assertFalse(srcStatus.getReplication() == dstStatus.getReplication());
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test(expected = FileNotFoundException.class)
public void testGetFileStatusWithInvalidPath() throws Exception {
  {
    setupRPC(
        DFS.RpcType.GET_FILE_STATUS_REQUEST, DFS.GetFileStatusRequest.newBuilder().setPath("/foo/bar").build(),
        DFS.GetFileStatusResponse.class, newRPCException(LOCAL_ENDPOINT, new FileNotFoundException("File not found")));
  }

  FileSystem fs = newRemoteNodeFileSystem();

  Path path = new Path("/foo/bar");
  fs.getFileStatus(path);
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private static void touchFile(String path, boolean createMultipleBlocks,
    ChecksumOpt checksumOpt) throws Exception {
  FileSystem fs;
  DataOutputStream outputStream = null;
  try {
    fs = cluster.getFileSystem();
    final Path qualifiedPath = new Path(path).makeQualified(fs.getUri(),
        fs.getWorkingDirectory());
    final long blockSize = createMultipleBlocks ? NON_DEFAULT_BLOCK_SIZE : fs
        .getDefaultBlockSize(qualifiedPath) * 2;
    FsPermission permission = FsPermission.getFileDefault().applyUMask(
        FsPermission.getUMask(fs.getConf()));
    outputStream = fs.create(qualifiedPath, permission,
        EnumSet.of(CreateFlag.CREATE, CreateFlag.OVERWRITE), 0,
        (short) (fs.getDefaultReplication(qualifiedPath) * 2), blockSize,
        null, checksumOpt);
    byte[] bytes = new byte[DEFAULT_FILE_SIZE];
    outputStream.write(bytes);
    long fileSize = DEFAULT_FILE_SIZE;
    if (createMultipleBlocks) {
      while (fileSize < 2*blockSize) {
        outputStream.write(bytes);
        outputStream.flush();
        fileSize += DEFAULT_FILE_SIZE;
      }
    }
    pathList.add(qualifiedPath);
    ++nFiles;

    FileStatus fileStatus = fs.getFileStatus(qualifiedPath);
    System.out.println(fileStatus.getBlockSize());
    System.out.println(fileStatus.getReplication());
  }
  finally {
    IOUtils.cleanup(null, outputStream);
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private boolean checkExists(FileSystem fs, Path path, FsPermission fsPerm)
    throws IOException {
  boolean exists = true;
  try {
    FileStatus appDirStatus = fs.getFileStatus(path);
    if (!APP_DIR_PERMISSIONS.equals(appDirStatus.getPermission())) {
      fs.setPermission(path, APP_DIR_PERMISSIONS);
    }
  } catch (FileNotFoundException fnfe) {
    exists = false;
  }
  return exists;
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private void testListStatus() throws Exception {
  FileSystem fs = FileSystem.get(getProxiedFSConf());
  Path path = new Path(getProxiedFSTestDir(), "foo.txt");
  OutputStream os = fs.create(path);
  os.write(1);
  os.close();
  FileStatus status1 = fs.getFileStatus(path);
  fs.close();

  fs = getHttpFSFileSystem();
  FileStatus status2 = fs.getFileStatus(new Path(path.toUri().getPath()));
  fs.close();

  Assert.assertEquals(status2.getPermission(), status1.getPermission());
  Assert.assertEquals(status2.getPath().toUri().getPath(), status1.getPath().toUri().getPath());
  Assert.assertEquals(status2.getReplication(), status1.getReplication());
  Assert.assertEquals(status2.getBlockSize(), status1.getBlockSize());
  Assert.assertEquals(status2.getAccessTime(), status1.getAccessTime());
  Assert.assertEquals(status2.getModificationTime(), status1.getModificationTime());
  Assert.assertEquals(status2.getOwner(), status1.getOwner());
  Assert.assertEquals(status2.getGroup(), status1.getGroup());
  Assert.assertEquals(status2.getLen(), status1.getLen());

  FileStatus[] stati = fs.listStatus(path.getParent());
  Assert.assertEquals(stati.length, 1);
  Assert.assertEquals(stati[0].getPath().getName(), path.getName());
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testPreserveReplicationOnFile() throws IOException {
  FileSystem fs = FileSystem.get(config);
  EnumSet<FileAttribute> attributes = EnumSet.of(FileAttribute.REPLICATION);

  Path dst = new Path("/tmp/dest2");
  Path src = new Path("/tmp/src2");

  createFile(fs, src);
  createFile(fs, dst);

  fs.setPermission(src, fullPerm);
  fs.setOwner(src, "somebody", "somebody-group");
  fs.setTimes(src, 0, 0);
  fs.setReplication(src, (short) 1);

  fs.setPermission(dst, noPerm);
  fs.setOwner(dst, "nobody", "nobody-group");
  fs.setTimes(dst, 100, 100);
  fs.setReplication(dst, (short) 2);

  CopyListingFileStatus srcStatus = new CopyListingFileStatus(fs.getFileStatus(src));

  DistCpUtils.preserve(fs, dst, srcStatus, attributes, false);

  CopyListingFileStatus dstStatus = new CopyListingFileStatus(fs.getFileStatus(dst));

  // FileStatus.equals only compares path field, must explicitly compare all fields
  Assert.assertFalse(srcStatus.getPermission().equals(dstStatus.getPermission()));
  Assert.assertFalse(srcStatus.getOwner().equals(dstStatus.getOwner()));
  Assert.assertFalse(srcStatus.getGroup().equals(dstStatus.getGroup()));
  Assert.assertFalse(srcStatus.getAccessTime() == dstStatus.getAccessTime());
  Assert.assertFalse(srcStatus.getModificationTime() == dstStatus.getModificationTime());
  Assert.assertTrue(srcStatus.getReplication() == dstStatus.getReplication());
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testRootDirPermission() throws IOException {
  FileSystem fs = FileSystem.getLocal(new YarnConfiguration());
  FileStatus file = fs.getFileStatus(
      new Path(fsPath.getAbsolutePath(), LeveldbTimelineStore.FILENAME));
  assertNotNull(file);
  assertEquals(LeveldbTimelineStore.LEVELDB_DIR_UMASK, file.getPermission());
}