scala 实现 hadoop 多重文件输出

448 查看

package com.timger.tools

/**
 * Created by timger on 15-1-26.
 */


import com.timger.etl.TokenizerMapper
import org.apache.hadoop.conf.Configuration
import org.apache.hadoop.fs.Path
import org.apache.hadoop.io.{LongWritable, NullWritable, IntWritable, Text}
//import org.apache.hadoop.mapred.lib.{MultipleOutputFormat, MultipleTextOutputFormat}
import org.apache.hadoop.mapreduce.lib.output.LazyOutputFormat
import org.apache.hadoop.mapreduce.lib.output.MultipleOutputs;
import org.apache.hadoop.mapreduce.{Job,InputFormat,OutputFormat}
import org.apache.hadoop.mapreduce.Mapper
import org.apache.hadoop.mapreduce.Reducer
import org.apache.hadoop.mapreduce.lib.input.{TextInputFormat, FileInputFormat}
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat
import org.apache.hadoop.util.GenericOptionsParser
import scala.collection.JavaConversions._
import org.apache.hadoop.mapreduce.lib.output;
import org.apache.hadoop.mapreduce.lib.output.{TextOutputFormat}

/**
 * https://hadoop.apache.org/docs/r2.2.0//api/org/apache/hadoop/mapred/lib/MultipleOutputs.html
 */

// This class performs the map operation, translating raw input into the key-value
// pairs we will feed into our reduce operation.
class SplitMapper extends Mapper[Object,Text,Text,Text] {
  //https://github.com/rystsov/learning-hadoop/blob/master/src/main/java/com/twitter/rystsov/mr/MultipulOutputExample.java

  //var multipleOutputs :MultipleOutputs = null;
  private var multipleOutputs: MultipleOutputs[Text, Text] = null

  @throws(classOf[java.io.IOException])
  @throws(classOf[java.lang.InterruptedException])
  override
  protected def setup(context:Mapper[Object, Text, Text, Text]#Context)  {
    //pattern = Pattern.compile("^http://([^/]+).+$");
   multipleOutputs = new MultipleOutputs(context);
    println("SetUp Ok ......................");
  }


  override
  def map(key:Object, value:Text, context:Mapper[Object,Text,Text,Text]#Context) = {
    //var conf :Configuration  = context.getConfiguration();
    //var splitkeys  = conf.get("splitkeys");

    var keys = "26/Jan/2015,27/Jan/2015".split(",");
    //var keys =
    val word = new Text;

    var line = value.toString();
    for (key <-  keys) {

      word.set(key)
      if (line.contains(key)) {
        var mkey :String = key.toString().replace("/",  "");
        context.write(word, value);
        multipleOutputs.write(mkey,
                              word,
                              value);
      }
    }
  }


  @throws(classOf[java.io.IOException])
  @throws(classOf[InterruptedException])
  override
  protected def cleanup( context:Mapper[Object,Text,Text,Text]#Context){
    multipleOutputs.close();
  }
}

/***
class MutiFIleReducer extends Reducer[Text,Text,Text,Text] {

  override
  def reduce(key:Text, values:java.lang.Iterable[Text], context:Reducer[Text,Text,Text,Text]#Context) = {
    for (value <- values) {
      context.write(key, value)
    }
  }
}
 ***/



// This class configures and runs the job with the map and reduce classes we've
// specified above.
object SplitFile {

  def main(args:Array[String]):Int = {
    val conf = new Configuration()
    val otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs
    var usage =
      """
        |Usage: FileSplit <inputdir> <outdir> <tmpdir> <keeyoriginfile>[<key1> <key2> <key3> key4>...]
        |inputdir is a file or dir
        |outputdir is a dir need not blank
        |tmpdir use for tmp
        |keeyoriginfile is a bool

"""
    print("\n\n");
    print(args.toString());
    print(otherArgs);
    println()
    if (args.length < 5) {
      print(usage)
      return 2
    }
    val job = new Job(conf, "com.timger.tools.SplitFile")
    job.setJarByClass(classOf[SplitMapper])
    job.setMapperClass(classOf[SplitMapper])

    //job.setCombinerClass(classOf[IntSumReducer])
    //job.setReducerClass(classOf[MutiFIleReducer])

    job.setOutputKeyClass(classOf[Text])
    job.setInputFormatClass(classOf[TextInputFormat])

    var inputdir = new Path(otherArgs(0));
    var outputdir = new Path(otherArgs(1));
    var tmpdir = new Path(otherArgs(2));
    var keeyoriginfile :String= otherArgs(3);
    var splitkeys :Array[String]= otherArgs.slice(4, otherArgs.length);
    var keys = splitkeys.mkString(",")
    var test_str :String =
      """
        |input: %s
        |output: %s
        |tmpdir: %s
        |keys: %s
      """.format(inputdir, outputdir, tmpdir, keys);
    print(test_str);
    //onf.set("splitkeys", keys);
    print(splitkeys);

   // val TextOutputFormatClass = classOf[TextOutputFormat].asInstanceOf[Class[T] forSome {type T <: OutputFormat[String, String]}]

    for (key <- splitkeys) {
      var mkey  :String = key.toString().replace("/",  "");
      print(mkey);


      MultipleOutputs.addNamedOutput(job, mkey,
        classOf[TextOutputFormat[Text,Text]],
        classOf[Text],
        classOf[Text]);
    }

    //job.setNumReduceTasks(splitkeys.length);
    job.setNumReduceTasks(0);

    job.setOutputValueClass(classOf[Text])
    FileInputFormat.addInputPath(job, inputdir)
    FileOutputFormat.setOutputPath(job, tmpdir)

    //job.setOutputFormatClass(classOf[MyMultipleOutputFormat]);

    //job.setOutputFormatClass(classOf[CustomMultipleTextOutputFormat]);


    //LazyOutputFormat.setOutputFormatClass(conf,  classOf[MyMultipleOutputFormat])
    if (job.waitForCompletion(true)) 0 else 1
  }

}

2018 - 知识虫 - 我的知识库 渝ICP备16002641号-2

渝公网安备 50010702501581号