HDFSTouch.java

package org.archive.hadoop.io;

import java.text.SimpleDateFormat;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FSDataOutputStream;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;

public class HDFSTouch implements Tool {
	
	protected Configuration conf;
	
	public final static String TOOL_NAME = "hdfs-touch";
	
	public final static String TMP_FILENAME = ".tmp_touch_latest";
	
	public final static String FORMAT_STR = "yyyy-MM-dd HH:mm:ss";
	public final static String HTTP_FORMAT = "EEE, dd MMM yyyy HH:mm:ss zzz";
	
	protected SimpleDateFormat formats[] = {
			new SimpleDateFormat(FORMAT_STR),
			new SimpleDateFormat(HTTP_FORMAT)
	};
	
	
	public static void main(String[] args) throws Exception {
		int res = ToolRunner.run(new Configuration(), new HDFSTouch(), args);
		System.exit(res);
	}

	public void setConf(Configuration conf) {
		this.conf = conf;
	}
	
	public Configuration getConf() {
		return conf;
	}
	
	public static void print_usage() {
		System.err.println("Usage: " + TOOL_NAME + " [ -d ] HDFS_URL [ TIMESTAMP ]");
		System.err.println("Updates the mtime and atime of HDFS_URL to current time, or optional TIMESTAMP");
		System.err.println("Accepts following TIMESTAMP formats:");
		System.err.println("  " + FORMAT_STR);
		System.err.println("  " + HTTP_FORMAT);
	}
	
	protected long parseFormats(String arg) {
		long mtime = -1;
		
		for (SimpleDateFormat format : formats) {
			try {
				mtime = format.parse(arg).getTime();
				return mtime;
				
			} catch (Exception exc) {
				
			}
		}
		
		if (mtime == -1) {
			System.err.println("Error parsing timestamp: " + arg + "\nExpected format: " + FORMAT_STR + " or " + HTTP_FORMAT);
			mtime = System.currentTimeMillis();
		}
		
		return mtime;
	}

	public int run(String[] args) throws Exception {
		
		if (args.length < 1) {
			print_usage();
			return 1;
		}
		
		String filePath = null;
		boolean updateDir = false;
		
		if (args.length == 1) {
			filePath = args[0];
		} else {
			updateDir = (args[0].equals("-d"));
			filePath = args[1];
		}
		
		Path path = new Path(filePath);
		FileSystem fs = path.getFileSystem(conf);
		
		if (fs.getFileStatus(path).isDir()) {
			System.err.println("Can't touch directories in this version\n" +
					"This is a directory: " + path);
			return 1;
		}
		
		long mtime = System.currentTimeMillis();
		
		if (args.length >= 3) {
			mtime = parseFormats(args[2]);
		}
		
		long atime = mtime;		
		fs.setTimes(path, mtime, atime);
		
		if (updateDir) {			
			Path tempFilePath = new Path(path.getParent(), TMP_FILENAME);
			
			// Create empty output stream
			FSDataOutputStream out = fs.create(tempFilePath);
			out.close();
			// Delete tmp file
			fs.delete(tempFilePath, false);
		}
						
		return 0;
	}
}