HDFS常用的文件API操作

1、常用文件API操作

package cn.luxh.app.util;

import java.io.IOException;
import java.text.SimpleDateFormat;
import java.util.Date;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.BlockLocation;
import org.apache.hadoop.fs.FSDataOutputStream;
import org.apache.hadoop.fs.FileStatus;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.hdfs.DistributedFileSystem;
import org.apache.hadoop.hdfs.protocol.DatanodeInfo;

public class HDFSUtil {
    
    /**
     * 从本地复制文件到HDFS
     * @param srcFilePath
     * @param dstFilePath
     * @throws IOException
     */
    public static void copyFile2HDFS(String srcFilePath,String dstFilePath) throws IOException{
        Configuration conf = new Configuration();

        Path src = new Path(srcFilePath);
        Path dst = new Path(dstFilePath);
        
        FileSystem hdfs = dst.getFileSystem(conf);
        
        hdfs.copyFromLocalFile(src, dst);
        
        FileStatus[] files = hdfs.listStatus(dst);
        if(files!=null) {
            for(int i=0;i<files.length;i++) {
                System.out.println("the file is:"+files[i].getPath().getName());
            }
        }else {
            System.out.println("no files");
        }
        hdfs.close();
        
        
    }
    
    /**
     * 在HDFS上创建文件
     * @param content
     * @param dstFile
     * @throws IOException 
     */
    public static void createFileInHDFS(String content,String dstFile) throws IOException {
        Configuration conf = new Configuration();
        Path dst = new Path(dstFile);
        FileSystem hdfs = null;
        FSDataOutputStream out = null;
        try {
            hdfs = dst.getFileSystem(conf);
            out = hdfs.create(dst);
            out.writeBytes(content);
            out.flush();
        } catch (IOException e) {
            e.printStackTrace();
            throw new IOException(e);
        }finally {
            if(hdfs != null) {
                hdfs.close();
            }
            if(out != null) {
                try {
                    out.close();
                } catch (IOException e) {
                    e.printStackTrace();
                }
            }
            
        }
        
    }
    
    /**
     * 重命名文件
     * @param originalFile
     * @param newFile
     * @throws IOException
     */
    public static void renameFileInHDFS(String originalFile,String newFile) throws IOException {
        Configuration conf = new Configuration();
        Path originalPath = new Path(originalFile);
        Path newPath = new Path(newFile);
        FileSystem hdfs = newPath.getFileSystem(conf);
        hdfs.rename(originalPath, newPath);
        hdfs.close();
        
        
    }
    
    /**
     * 获得文件的最后修改时间
     * @param dstFile
     * @throws IOException
     */
    public static void getFileLastModifyTime(String dstFile) throws IOException{
        Configuration conf = new Configuration();
        Path dstPath = new Path(dstFile);
        FileSystem hdfs =  dstPath.getFileSystem(conf);
        FileStatus file = hdfs.getFileStatus(dstPath);
        long time = file.getModificationTime();
        hdfs.close();
        System.out.println("the last modify time is : "+new SimpleDateFormat("yyyy-MM-dd HH:mm:ss").format(new Date(time)));
        
    }
    
    /**
     * 检查文件是否存在
     * @param dstFile
     * @throws IOException
     */
    public static void checkFileIsExists(String dstFile) throws IOException {
        Configuration conf = new Configuration();
        Path dstPath = new Path(dstFile);
        FileSystem hdfs =  dstPath.getFileSystem(conf);
        boolean flag = hdfs.exists(dstPath);
        hdfs.close();
        System.out.println("is the file exists:"+flag);
    }
    
    /**
     * 获得文件的存放位置
     * @param dstFile
     * @throws IOException
     */
    public static void getFileLocations(String dstFile) throws IOException {
        Configuration conf = new Configuration();
        Path dstPath = new Path(dstFile);
        FileSystem hdfs =  dstPath.getFileSystem(conf);
        FileStatus file = hdfs.getFileStatus(dstPath);
        BlockLocation[] blkLocations = hdfs.getFileBlockLocations(file, 0, file.getLen());
        if(blkLocations != null) {
            int len = blkLocations.length;
            for(int i=0;i<len;i++) {
                String[] hosts = blkLocations[i].getHosts();
                for(String host:hosts) {
                    System.out.println("the location'host is : "+host);
                }
            }
        }
        
        hdfs.close();
    }
    
    /**
     * 删除文件
     * @throws IOException
     */
    public static void deleteFile(String dstFile) throws IOException {
        Configuration conf = new Configuration();
        Path dstPath = new Path(dstFile);
        FileSystem hdfs =  dstPath.getFileSystem(conf);
        boolean flag = hdfs.delete(dstPath, true);
        
        System.out.println("is deleted : "+flag);
    }
}

2、测试

package cn.luxh.app.test;

import java.io.IOException;

import org.junit.Test;

import cn.luxh.app.util.HDFSUtil;

public class HDFSTester {
    
    @Test
    public void testCopyFile2HDFS() throws IOException {
        String srcFilePath = "/home/coder/words.txt";
        String dstFilePath = "hdfs://h1:9000/user/coder/in";
        HDFSUtil.copyFile2HDFS(srcFilePath, dstFilePath);
    }
    
    @Test
    public void testCreateFileInHDFS() throws IOException {
        String content = "hey,Hadoop.";
        String dstFile = "hdfs://h1:9000/user/coder/in/hey";
        HDFSUtil.createFileInHDFS(content, dstFile);
    }
    
    @Test
    public void testRenameFileInHDFS() throws IOException {
        String originalFile = "hdfs://h1:9000/user/coder/in/hey";
        String newFile = "hdfs://h1:9000/user/coder/in/hey_hadoop";
        HDFSUtil.renameFileInHDFS(originalFile, newFile);
    }
    
    @Test
    public void testGetFileLastModifyTimme() throws IOException {
        String dstFile = "hdfs://h1:9000/user/coder/in/hey_hadoop";
        HDFSUtil.getFileLastModifyTime(dstFile);
    }
    
    @Test
    public void testCheckFileIsExists() throws IOException {
        String dstFile = "hdfs://h1:9000/user/coder/in/hey_hadoop";
        HDFSUtil.checkFileIsExists(dstFile);
    }
    
    @Test
    public void testGetFileLocations() throws IOException {
        String dstFile = "hdfs://h1:9000/user/coder/in/hey_hadoop";
        HDFSUtil.getFileLocations(dstFile);
    }
    
    @Test
    public void testDeleteFile() throws IOException {
        String dstFile = "hdfs://h1:9000/user/coder/output";
        HDFSUtil.deleteFile(dstFile);
    }
}