hadoop: hdfs API示例

利用hdfs的api，可以实现向hdfs的文件、目录读写，利用这一套API可以设计一个简易的山寨版云盘，见下图：

为了方便操作,将常用的文件读写操作封装了一个工具类:

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.*;
import org.apache.hadoop.io.IOUtils;

import java.io.ByteArrayOutputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;

/**
 * HDFS工具类
 * Author: 菩提树下的杨过(http://yjmyzz.cnblogs.com)
 * Since: 2015-05-21
 */
public class HDFSUtil {


    private HDFSUtil() {

    }

    /**
     * 判断路径是否存在
     *
     * @param conf
     * @param path
     * @return
     * @throws IOException
     */
    public static boolean exits(Configuration conf, String path) throws IOException {
        FileSystem fs = FileSystem.get(conf);
        return fs.exists(new Path(path));
    }

    /**
     * 创建文件
     *
     * @param conf
     * @param filePath
     * @param contents
     * @throws IOException
     */
    public static void createFile(Configuration conf, String filePath, byte[] contents) throws IOException {
        FileSystem fs = FileSystem.get(conf);
        Path path = new Path(filePath);
        FSDataOutputStream outputStream = fs.create(path);
        outputStream.write(contents);
        outputStream.close();
        fs.close();
    }

    /**
     * 创建文件
     *
     * @param conf
     * @param filePath
     * @param fileContent
     * @throws IOException
     */
    public static void createFile(Configuration conf, String filePath, String fileContent) throws IOException {
        createFile(conf, filePath, fileContent.getBytes());
    }

    /**
     * @param conf
     * @param localFilePath
     * @param remoteFilePath
     * @throws IOException
     */
    public static void copyFromLocalFile(Configuration conf, String localFilePath, String remoteFilePath) throws IOException {
        FileSystem fs = FileSystem.get(conf);
        Path localPath = new Path(localFilePath);
        Path remotePath = new Path(remoteFilePath);
        fs.copyFromLocalFile(true, true, localPath, remotePath);
        fs.close();
    }

    /**
     * 删除目录或文件
     *
     * @param conf
     * @param remoteFilePath
     * @param recursive
     * @return
     * @throws IOException
     */
    public static boolean deleteFile(Configuration conf, String remoteFilePath, boolean recursive) throws IOException {
        FileSystem fs = FileSystem.get(conf);
        boolean result = fs.delete(new Path(remoteFilePath), recursive);
        fs.close();
        return result;
    }

    /**
     * 删除目录或文件(如果有子目录,则级联删除)
     *
     * @param conf
     * @param remoteFilePath
     * @return
     * @throws IOException
     */
    public static boolean deleteFile(Configuration conf, String remoteFilePath) throws IOException {
        return deleteFile(conf, remoteFilePath, true);
    }

    /**
     * 文件重命名
     *
     * @param conf
     * @param oldFileName
     * @param newFileName
     * @return
     * @throws IOException
     */
    public static boolean renameFile(Configuration conf, String oldFileName, String newFileName) throws IOException {
        FileSystem fs = FileSystem.get(conf);
        Path oldPath = new Path(oldFileName);
        Path newPath = new Path(newFileName);
        boolean result = fs.rename(oldPath, newPath);
        fs.close();
        return result;
    }

    /**
     * 创建目录
     *
     * @param conf
     * @param dirName
     * @return
     * @throws IOException
     */
    public static boolean createDirectory(Configuration conf, String dirName) throws IOException {
        FileSystem fs = FileSystem.get(conf);
        Path dir = new Path(dirName);
        boolean result = fs.mkdirs(dir);
        fs.close();
        return result;
    }

    /**
     * 列出指定路径下的所有文件(不包含目录)
     *
     * @param conf
     * @param basePath
     * @param recursive
     */
    public static RemoteIterator<LocatedFileStatus> listFiles(FileSystem fs, String basePath, boolean recursive) throws IOException {

        RemoteIterator<LocatedFileStatus> fileStatusRemoteIterator = fs.listFiles(new Path(basePath), recursive);

        return fileStatusRemoteIterator;
    }

    /**
     * 列出指定路径下的文件（非递归）
     *
     * @param conf
     * @param basePath
     * @return
     * @throws IOException
     */
    public static RemoteIterator<LocatedFileStatus> listFiles(Configuration conf, String basePath) throws IOException {
        FileSystem fs = FileSystem.get(conf);
        RemoteIterator<LocatedFileStatus> remoteIterator = fs.listFiles(new Path(basePath), false);
        fs.close();
        return remoteIterator;
    }

    /**
     * 列出指定目录下的文件子目录信息（非递归）
     *
     * @param conf
     * @param dirPath
     * @return
     * @throws IOException
     */
    public static FileStatus[] listStatus(Configuration conf, String dirPath) throws IOException {
        FileSystem fs = FileSystem.get(conf);
        FileStatus[] fileStatuses = fs.listStatus(new Path(dirPath));
        fs.close();
        return fileStatuses;
    }


    /**
     * 读取文件内容
     *
     * @param conf
     * @param filePath
     * @return
     * @throws IOException
     */
    public static String readFile(Configuration conf, String filePath) throws IOException {
        String fileContent = null;
        FileSystem fs = FileSystem.get(conf);
        Path path = new Path(filePath);
        InputStream inputStream = null;
        ByteArrayOutputStream outputStream = null;
        try {
            inputStream = fs.open(path);
            outputStream = new ByteArrayOutputStream(inputStream.available());
            IOUtils.copyBytes(inputStream, outputStream, conf);
            fileContent = outputStream.toString();
        } finally {
            IOUtils.closeStream(inputStream);
            IOUtils.closeStream(outputStream);
            fs.close();
        }
        return fileContent;
    }
}

简单的测试了一下:

    @Test
    public void test() throws IOException {
        Configuration conf = new Configuration();
        String newDir = "/test";
        //01.检测路径是否存在 测试
        if (HDFSUtil.exits(conf, newDir)) {
            System.out.println(newDir + " 已存在!");
        } else {
            //02.创建目录测试
            boolean result = HDFSUtil.createDirectory(conf, newDir);
            if (result) {
                System.out.println(newDir + " 创建成功!");
            } else {
                System.out.println(newDir + " 创建失败!");
            }
        }
        String fileContent = "Hi,hadoop. I love you";
        String newFileName = newDir + "/myfile.txt";

        //03.创建文件测试
        HDFSUtil.createFile(conf, newFileName, fileContent);
        System.out.println(newFileName + " 创建成功");

        //04.读取文件内容 测试
        System.out.println(newFileName + " 的内容为:
" + HDFSUtil.readFile(conf, newFileName));

        //05. 测试获取所有目录信息
        FileStatus[] dirs = HDFSUtil.listStatus(conf, "/");
        System.out.println("--根目录下的所有子目录---");
        for (FileStatus s : dirs) {
            System.out.println(s);
        }

        //06. 测试获取所有文件
        FileSystem fs = FileSystem.get(conf);
        RemoteIterator<LocatedFileStatus> files = HDFSUtil.listFiles(fs, "/", true);
        System.out.println("--根目录下的所有文件---");
        while (files.hasNext()) {
            System.out.println(files.next());
        }
        fs.close();

        //删除文件测试
        boolean isDeleted = HDFSUtil.deleteFile(conf, newDir);
        System.out.println(newDir + " 已被删除");

    }

注:测试时,不要忘记了在resources目录下放置core-site.xml文件,不然IDE环境下,代码不知道去连哪里的HDFS

输出结果：

/test 已存在!
/test/myfile.txt 创建成功
/test/myfile.txt 的内容为:
Hi,hadoop. I love you
--根目录下的所有子目录---
FileStatus{path=hdfs://172.28.20.102:9000/jimmy; isDirectory=true; modification_time=1432176691550; access_time=0; owner=hadoop; group=supergroup; permission=rwxrwxrwx; isSymlink=false}
FileStatus{path=hdfs://172.28.20.102:9000/test; isDirectory=true; modification_time=1432181331362; access_time=0; owner=jimmy; group=supergroup; permission=rwxr-xr-x; isSymlink=false}
FileStatus{path=hdfs://172.28.20.102:9000/user; isDirectory=true; modification_time=1431931797244; access_time=0; owner=hadoop; group=supergroup; permission=rwxr-xr-x; isSymlink=false}
--根目录下的所有文件---
LocatedFileStatus{path=hdfs://172.28.20.102:9000/jimmy/input/README.txt; isDirectory=false; length=1366; replication=1; blocksize=134217728; modification_time=1431922483851; access_time=1432174134018; owner=hadoop; group=supergroup; permission=rw-r--r--; isSymlink=false}
LocatedFileStatus{path=hdfs://172.28.20.102:9000/jimmy/output/_SUCCESS; isDirectory=false; length=0; replication=3; blocksize=134217728; modification_time=1432176692454; access_time=1432176692448; owner=jimmy; group=supergroup; permission=rw-r--r--; isSymlink=false}
LocatedFileStatus{path=hdfs://172.28.20.102:9000/jimmy/output/part-r-00000; isDirectory=false; length=1306; replication=3; blocksize=134217728; modification_time=1432176692338; access_time=1432176692182; owner=jimmy; group=supergroup; permission=rw-r--r--; isSymlink=false}
LocatedFileStatus{path=hdfs://172.28.20.102:9000/test/myfile.txt; isDirectory=false; length=21; replication=3; blocksize=134217728; modification_time=1432181331601; access_time=1432181331362; owner=jimmy; group=supergroup; permission=rw-r--r--; isSymlink=false}
/test 已被删除

用spring-mvc结合hdfs api仿造hadoop的文件浏览管理界面，做了一个山寨版：（只完成了文件列表功能）

源代码托管在taobao开源平台上了，有需要的可以参考下：

http://code.taobao.org/p/hdfs-web-client/src/trunk/