【转自】自定义InputFormat、OutputFormat
转自：http://www.cnblogs.com/xiaolong1032/p/4529534.html
一:自定义实现InputFormat
*数据源来自于内存
*1.InputFormat是用于处理各种数据源的,下面是实现InputFormat,数据源是来自于内存.
*1.1 在程序的job.setInputFormatClass(MyselfmemoryInputFormat.class);
*1.2 实现InputFormat,extends InputFormat< , >,实现其中的两个方法,分别是getSplits(..),createRecordReader(..).
*1.3 getSplits(..)返回的是一个java.util.List<T>,List中的每个元素是InputSplit.每个InputSplit对应一个mappper任务.
*1.4 InputSplit是对原始海量数据源的划分,因为我们处理的是海量数据,不划分不行.InputSplit数据的大小完全是我们自己来定的.本例中是在内存中产生数据,然后封装到InputSplit.
*1.5 InputSplit封装的是hadoop数据类型,实现Writable接口.
*1.6 RecordReader读取每个InputSplit中的数据.解析成一个个<k,v>,供map处理.
*1.7 RecordReader有4个核心方法,分别是initalize(..).nextKeyValue(),getCurrentKey(),getCurrentValue().
*1.8 initalize重要性在于是拿到InputSplit和定义临时变量.
*1.9 nexKeyValue(..)该方法的每次调用,可以获得key和value值.
*1.10 当nextKeyValue(..)调用后,紧接着调用getCurrentKey(),getCurrentValue().
* mapper方法中的run方法调用.
public class MyselInputFormatApp {
private static final String OUT_PATH = "hdfs://hadoop1:9000/out";// 输出路径,reduce作业输出的结果是一个目录
public static void main(String[] args) {
Configuration conf = new Configuration();// 配置对象
try {
FileSystem fileSystem = FileSystem.get(new URI(OUT_PATH), conf);
fileSystem.delete(new Path(OUT_PATH), true);
Job job = new Job(conf, WordCountApp.class.getSimpleName());// jobName:作业名称
job.setJarByClass(WordCountApp.class);
job.setInputFormatClass(MyselfMemoryInputFormat.class);
job.setMapperClass(MyMapper.class);// 指定自定义map类
job.setMapOutputKeyClass(Text.class);// 指定map输出key的类型
job.setMapOutputValueClass(LongWritable.class);// 指定map输出value的类型
job.setReducerClass(MyReducer.class);// 指定自定义Reduce类
job.setOutputKeyClass(Text.class);// 设置Reduce输出key的类型
job.setOutputValueClass(LongWritable.class);// 设置Reduce输出的value类型
FileOutputFormat.setOutputPath(job, new Path(OUT_PATH));// Reduce输出完之后,就会产生一个最终的输出,指定最终输出的位置
job.waitForCompletion(true);// 提交给jobTracker并等待结束
} catch (Exception e) {
e.printStackTrace();
}
}
public static class MyMapper extends
Mapper<NullWritable, Text, Text, LongWritable> {
@Override
protected void map(NullWritable key, Text value, Context context)
throws IOException, InterruptedException {
String line = value.toString();
String[] splited = line.split("\t");
for (String word : splited) {
context.write(new Text(word), new LongWritable(1));// 把每个单词出现的次数1写出去.
}
}
}
public static class MyReducer extends
Reducer<Text, LongWritable, Text, LongWritable> {
@Override
protected void reduce(Text key, Iterable<LongWritable> values,
Context context) throws IOException, InterruptedException {
long count = 0L;
for (LongWritable times : values) {
count += times.get();
}
context.write(key, new LongWritable(count));
}
}
/**
* 从内存中产生数据,然后解析成一个个的键值对
*
*/
public static class MyselfMemoryInputFormat extends InputFormat<NullWritable,Text>{
@Override
public List<InputSplit> getSplits(JobContext context)
throws IOException, InterruptedException {
ArrayList<InputSplit> result = new ArrayList<InputSplit>();
result.add(new MemoryInputSplit());
result.add(new MemoryInputSplit());
result.add(new MemoryInputSplit());
return result;
}
@Override
public RecordReader<NullWritable, Text> createRecordReader(
InputSplit split, TaskAttemptContext context)
throws IOException, InterruptedException {
return new MemoryRecordReader();
}
}
public static class MemoryInputSplit extends InputSplit implements Writable{
int SIZE = 10;
//java中的数组在hadoop中不被支持,所以这里使用hadoop的数组
//在hadoop中使用的是这种数据结构,不能使用java中的数组表示.
ArrayWritable arrayWritable = new ArrayWritable(Text.class);
/**
* 先创建一个java数组类型,然后转化为hadoop的数据类型.
* @throws FileNotFoundException
*/
public MemoryInputSplit() throws FileNotFoundException {
//一个inputSplit供一个map使用,map函数如果要被调用多次的话,意味着InputSplit必须解析出多个键值对
Text[] array = new Text[SIZE];
Random random = new Random();
for(int i=0;i<SIZE;i++){
int nextInt = random.nextInt(999999);
Text text = new Text("Text"+nextInt);
array[i] = text ;
}
// FileInputStream fs = new FileInputStream(new File("\\etc\\profile"));//从文件中读取
// 将流中的数据解析出来放到数据结构中.
arrayWritable.set(array);
}
@Override
public long getLength() throws IOException, InterruptedException {
return SIZE;
}
@Override
public String[] getLocations() throws IOException,
InterruptedException {
return new String[]{};
}
public ArrayWritable getValues() {
return arrayWritable;
}
@Override
public void write(DataOutput out) throws IOException {
arrayWritable.write(out);
}
@Override
public void readFields(DataInput in) throws IOException {
arrayWritable.readFields(in);
}
}
public static class MemoryRecordReader extends RecordReader<NullWritable, Text>{
private Writable[] values = null ;
private Text value = null ;
private int i = 0;
@Override
public void initialize(InputSplit split, TaskAttemptContext context)
throws IOException, InterruptedException {
MemoryInputSplit inputSplit = (MemoryInputSplit)split;
ArrayWritable writables = inputSplit.getValues();
this.values = writables.get();
this.i = 0 ;
}
@Override
public boolean nextKeyValue() throws IOException,
InterruptedException {
if(i >= values.length){
return false ;
}
if(null == this.value){
value = new Text();
}
value.set((Text)values[i]);
i++ ;
return true;
}
@Override
public NullWritable getCurrentKey() throws IOException,
InterruptedException {
return NullWritable.get();
}
@Override
public Text getCurrentValue() throws IOException,
InterruptedException {
return value;
}
@Override
public float getProgress() throws IOException, InterruptedException {
// TODO Auto-generated method stub
return 0;
}
/**
* 程序结束的时候,关闭
*/
@Override
public void close() throws IOException {
}
}
}
二:自定义实现OutputFormat
常见的输出类型:TextInputFormat:默认输出格式,key和value中间用tab隔开.
DBOutputFormat:写出到数据库的.
SequenceFileFormat:将key,value以Sequence格式输出的.
SequenceFileAsOutputFormat:SequenceFile以原始二进制的格式输出.
MapFileOutputFormat:将key和value写入MapFile中.由于MapFile中key是有序的,所以写入的时候必须保证记录是按key值顺序入的.
MultipleOutputFormat:多文件的一个输出.默认情况下一个reducer产生一个输出,但是有些时候我们想一个reducer产生多个输出,MultipleOutputFormat和MultipleOutputs就可以实现这个功能.
MultipleOutputFormat:可以自定义输出文件的名称.
继承MultipleOutputFormat 需要实现
getBaseRecordWriter():
generateFileNameForKeyvalue():根据键值确定文件名.
/**
*自定义输出OutputFormat:用于处理各种输出目的地的.
*1.OutputFormat需要写出的键值对是来自于Reducer类.是通过RecordWriter获得的.
*2.RecordWriter(..)中write只有key和value,写到那里去哪?这要通过单独传入输出流来处理.write方法就是把k,v写入到outputStream中的.
*3.RecordWriter类是位于OutputFormat中的.因此,我们自定义OutputFormat必须继承OutputFormat类.那么流对象就必须在getRecordWriter(..)中获得.
*/
public class MySelfOutputFormatApp {
private static final String INPUT_PATH = "hdfs://hadoop1:9000/abd/hello";// 输入路径
private static final String OUT_PATH = "hdfs://hadoop1:9000/out";// 输出路径,reduce作业输出的结果是一个目录
private static final String OUT_FIE_NAME = "/abc";
public static void main(String[] args) {
Configuration conf = new Configuration();
try {
FileSystem fileSystem = FileSystem.get(new URI(OUT_PATH), conf);
fileSystem.delete(new Path(OUT_PATH), true);
Job job = new Job(conf, WordCountApp.class.getSimpleName());
job.setJarByClass(WordCountApp.class);
FileInputFormat.setInputPaths(job, INPUT_PATH);
job.setMapperClass(MyMapper.class);
job.setMapOutputKeyClass(Text.class);
job.setMapOutputValueClass(LongWritable.class);
job.setReducerClass(MyReducer.class);
job.setOutputKeyClass(Text.class);
job.setOutputValueClass(LongWritable.class);
job.setOutputFormatClass(MySelfTextOutputFormat.class);
job.waitForCompletion(true);
} catch (Exception e) {
e.printStackTrace();
}
} public static class MyMapper extends
Mapper<LongWritable, Text, Text, LongWritable> {
@Override
protected void map(LongWritable key, Text value, Context context)
throws IOException, InterruptedException {
String line = value.toString();
String[] splited = line.split("\t");
for (String word : splited) {
context.write(new Text(word), new LongWritable(1));// 把每个单词出现的次数1写出去.
}
}
} public static class MyReducer extends
Reducer<Text, LongWritable, Text, LongWritable> {
@Override
protected void reduce(Text key, Iterable<LongWritable> values,
Context context) throws IOException, InterruptedException {
long count = 0L;
for (LongWritable times : values) {
count += times.get();
}
context.write(key, new LongWritable(count));
}
}
/**
*自定义输出类型
*/
public static class MySelfTextOutputFormat extends OutputFormat<Text,LongWritable>{
FSDataOutputStream outputStream = null ;
@Override
public RecordWriter<Text, LongWritable> getRecordWriter(
TaskAttemptContext context) throws IOException,
InterruptedException {
try {
FileSystem fileSystem = FileSystem.get(new URI(MySelfOutputFormatApp.OUT_PATH), context.getConfiguration());
//指定的是输出文件的路径
String opath = MySelfOutputFormatApp.OUT_PATH+OUT_FIE_NAME;
outputStream = fileSystem.create(new Path(opath));
} catch (URISyntaxException e) {
e.printStackTrace();
}
return new MySelfRecordWriter(outputStream);
} @Override
public void checkOutputSpecs(JobContext context) throws IOException,
InterruptedException {
} /**
* OutputCommitter:在作业初始化的时候创建一些临时的输出目录,作业的输出目录,管理作业和任务的临时文件的.
* 作业运行过程中,会产生很多的Task,Task在处理的时候也会产生很多的输出.也会创建这个输出目录.
* 当我们的Task或者是作业都运行完成之后,输出目录由OutputCommitter给删了.所以程序在运行结束之后,我们根本看不见任何额外的输出.
* 在程序运行中会产生很多的临时文件,临时文件全交给OutputCommitter处理,真正的输出是RecordWriter(..),我们只需要关注最后的输出就可以了.中间的临时文件就是程序运行时产生的.
*/
@Override
public OutputCommitter getOutputCommitter(TaskAttemptContext context)
throws IOException, InterruptedException {
//提交任务的输出,包括初始化路径,包括在作业完成的时候清理作业,删除临时目录,包括作业和任务的临时目录.
//作业的输出路径应该是一个路径
return new FileOutputCommitter(new Path(MySelfOutputFormatApp.OUT_PATH), context);
}
}
public static class MySelfRecordWriter extends RecordWriter<Text, LongWritable>{
FSDataOutputStream outputStream = null ;
public MySelfRecordWriter(FSDataOutputStream outputStream) {
this.outputStream = outputStream ;
}
@Override
public void write(Text key, LongWritable value) throws IOException,
InterruptedException {
this.outputStream.writeBytes(key.toString());
this.outputStream.writeBytes("\t");
this.outputStream.writeLong(value.get());
}
@Override
public void close(TaskAttemptContext context) throws IOException,
InterruptedException {
this.outputStream.close();
}
}
}
三：输出到多个文件目录中去
/**
*输出到多个文件目录中去
*使用旧api
*/
public class MyMultipleOutputFormatApp {
private static final String INPUT_PATH = "hdfs://hadoop1:9000/abd";// 输入路径
private static final String OUT_PATH = "hdfs://hadoop1:9000/out";// 输出路径,reduce作业输出的结果是一个目录
public static void main(String[] args) {
Configuration conf = new Configuration();
try {
FileSystem fileSystem = FileSystem.get(new URI(OUT_PATH), conf);
fileSystem.delete(new Path(OUT_PATH), true);
JobConf job = new JobConf(conf, WordCountApp.class);
job.setJarByClass(WordCountApp.class);
FileInputFormat.setInputPaths(job, INPUT_PATH);
job.setMapperClass(MyMapper.class);
job.setMapOutputKeyClass(Text.class);
job.setMapOutputValueClass(LongWritable.class);
job.setReducerClass(MyReducer.class);
job.setOutputKeyClass(Text.class);
job.setOutputValueClass(LongWritable.class);
job.setOutputFormat(MyMutipleFilesTextOutputFormat.class);
FileOutputFormat.setOutputPath(job, new Path(OUT_PATH));
JobClient.runJob(job);
} catch (Exception e) {
e.printStackTrace();
}
} public static class MyMapper extends MapReduceBase implements Mapper<LongWritable, Text, Text, LongWritable> { @Override
public void map(LongWritable key, Text value,
OutputCollector<Text, LongWritable> output, Reporter reporter)
throws IOException {
String line = value.toString();
String[] splited = line.split("\t");
for (String word : splited) {
output.collect(new Text(word), new LongWritable(1));
}
}
} public static class MyReducer extends MapReduceBase implements Reducer<Text, LongWritable, Text, LongWritable> {
@Override
public void reduce(Text key, Iterator<LongWritable> values,
OutputCollector<Text, LongWritable> output, Reporter reporter)
throws IOException {
long count = 0L ;
while(values.hasNext()){
LongWritable times = values.next();
count += times.get();
}
output.collect(key, new LongWritable(count));
}
}
public static class MyMutipleFilesTextOutputFormat extends MultipleOutputFormat<Text,LongWritable>{ @Override
protected org.apache.hadoop.mapred.RecordWriter<Text, LongWritable> getBaseRecordWriter(
FileSystem fs, JobConf job, String name, Progressable progress)
throws IOException {
TextOutputFormat<Text, LongWritable> textOutputFormat = new TextOutputFormat<Text,LongWritable>();
return textOutputFormat.getRecordWriter(fs, job, name, progress);
} @Override
protected String generateFileNameForKeyValue(Text key,
LongWritable value, String name) {
String keyString = key.toString();
if(keyString.startsWith("hello")){
return "hello";
}else{
//输出的文件名就是k3的值
return keyString ;
}
} }
}
四：hadoop1.x api写单词计数的例子
/**
*hadoop1.x
*使用旧api写单词计数的例子
*/
public class WordCountApp {
private static final String INPUT_PATH = "hdfs://hadoop1:9000/abd/hello";
private static final String OUT_PATH = "hdfs://hadoop1:9000/out";
public static void main(String[] args) {
try {
Configuration conf = new Configuration();
FileSystem fs = FileSystem.get(conf);
fs.delete(new Path(OUT_PATH),true);
JobConf job = new JobConf(conf, WordCountApp.class);
job.setJarByClass(WordCountApp.class); FileInputFormat.setInputPaths(job, INPUT_PATH);
job.setMapperClass(MyMapper.class);
job.setMapOutputKeyClass(Text.class);
job.setMapOutputValueClass(LongWritable.class); job.setReducerClass(MyReducer.class);
job.setOutputKeyClass(Text.class);
job.setOutputValueClass(LongWritable.class);
FileOutputFormat.setOutputPath(job, new Path(OUT_PATH));
JobClient.runJob(job);
} catch (IOException e) {
e.printStackTrace();
}
}
public static class MyMapper extends MapReduceBase implements Mapper<LongWritable, Text, Text, LongWritable>{ @Override
public void map(LongWritable key, Text value,
OutputCollector<Text, LongWritable> output, Reporter reporter)
throws IOException {
String line = value.toString();
String[] splited = line.split("\t");
for (String word : splited) {
output.collect(new Text(word), new LongWritable(1L));
}
}
} public static class MyReducer extends MapReduceBase implements Reducer<Text, LongWritable, Text, LongWritable>{ @Override
public void reduce(Text key, Iterator<LongWritable> values,
OutputCollector<Text, LongWritable> output, Reporter reporter)
throws IOException {
long times = 0L ;
while (values.hasNext()) {
LongWritable longWritable = (LongWritable) values.next();
times += longWritable.get();
}
output.collect(key, new LongWritable(times));
} } }
五：运行时接收命令行参数
/**
*运行时会接收一些命令行的参数
*Tool接口:支持命令行的参数
*命令行执行:
* hadoop jar jar.jar cmd.WordCountApp hdfs://hadoop1:9000/abd/hello hdfs://hadoop1:9000/out
*/
public class WordCountApp extends Configured implements Tool {
private static String INPUT_PATH = null;// 输入路径
private static String OUT_PATH = null;// 输出路径,reduce作业输出的结果是一个目录
@Override
public int run(String[] args) throws Exception {
INPUT_PATH = args[0];
OUT_PATH = args[1];
Configuration conf = getConf();// 配置对象
try {
FileSystem fileSystem = FileSystem.get(new URI(OUT_PATH), conf);
fileSystem.delete(new Path(OUT_PATH), true);
Job job = new Job(conf, WordCountApp.class.getSimpleName());// jobName:作业名称
job.setJarByClass(WordCountApp.class);
FileInputFormat.setInputPaths(job, INPUT_PATH);// 指定数据的输入
job.setMapperClass(MyMapper.class);// 指定自定义map类
job.setMapOutputKeyClass(Text.class);// 指定map输出key的类型
job.setMapOutputValueClass(LongWritable.class);// 指定map输出value的类型
job.setReducerClass(MyReducer.class);// 指定自定义Reduce类
job.setOutputKeyClass(Text.class);// 设置Reduce输出key的类型
job.setOutputValueClass(LongWritable.class);// 设置Reduce输出的value类型
FileOutputFormat.setOutputPath(job, new Path(OUT_PATH));// Reduce输出完之后,就会产生一个最终的输出,指定最终输出的位置
job.waitForCompletion(true);// 提交给jobTracker并等待结束
} catch (Exception e) {
e.printStackTrace();
}
return 0;
}
public static void main(String[] args) {
try {
ToolRunner.run(new Configuration(), new WordCountApp(),args);
} catch (Exception e) {
e.printStackTrace();
}
}
public static class MyMapper extends
Mapper<LongWritable, Text, Text, LongWritable> {
@Override
protected void map(LongWritable key, Text value, Context context)
throws IOException, InterruptedException {
String line = value.toString();
String[] splited = line.split("\t");
for (String word : splited) {
context.write(new Text(word), new LongWritable(1));// 把每个单词出现的次数1写出去.
}
}
} public static class MyReducer extends
Reducer<Text, LongWritable, Text, LongWritable> {
@Override
protected void reduce(Text key, Iterable<LongWritable> values,
Context context) throws IOException, InterruptedException {
long count = 0L;
for (LongWritable times : values) {
count += times.get();
}
context.write(key, new LongWritable(count));
}
}
}
【转自】自定义InputFormat、OutputFormat的更多相关文章
- [Hadoop] - 自定义Mapreduce InputFormat&OutputFormat
在MR程序的开发过程中,经常会遇到输入数据不是HDFS或者数据输出目的地不是HDFS的,MapReduce的设计已经考虑到这种情况,它为我们提供了两个组建,只需要我们自定义适合的InputFormat ...
- 自定义InputFormat和OutputFormat案例
一.自定义InputFormat InputFormat是输入流,在前面的例子中使用的是文件输入输出流FileInputFormat和FileOutputFormat,而FileInputFormat ...
- MapReduce自定义InputFormat和OutputFormat
一.自定义InputFormat 需求:将多个小文件合并为SequenceFile(存储了多个小文件) 存储格式:文件路径+文件的内容 c:/a.txt I love Beijing c:/b.txt ...
- 自定义inputformat和outputformat
1. 自定义inputFormat 1.1 需求 无论hdfs还是mapreduce,对于小文件都有损效率,实践中,又难免面临处理大量小文件的场景,此时,就需要有相应解决方案 1.2 分析 小文件的优 ...
- 【Hadoop离线基础总结】MapReduce自定义InputFormat和OutputFormat案例
MapReduce自定义InputFormat和OutputFormat案例 自定义InputFormat 合并小文件 需求 无论hdfs还是mapreduce,存放小文件会占用元数据信息,白白浪费内 ...
- InputFormat,OutputFormat,InputSplit,RecordRead(一些常见面试题),使用yum安装64位Mysql
列举出hadoop常用的一些InputFormat InputFormat是用来对我们的输入数据进行格式化的.TextInputFormat是默认的. InputFormat有哪些类型? DBInpu ...
- MapReduce自定义InputFormat,RecordReader
MapReduce默认的InputFormat是TextInputFormat,且key是偏移量,value是文本,自定义InputFormat需要实现FileInputFormat,并重写creat ...
- commoncrawl 源码库是用于 Hadoop 的自定义 InputFormat 配送实现
commoncrawl 源码库是用于 Hadoop 的自定义 InputFormat 配送实现. Common Crawl 提供一个示例程序 BasicArcFileReaderSample.java ...
- Hadoop案例（六）小文件处理（自定义InputFormat）
小文件处理(自定义InputFormat) 1.需求分析 无论hdfs还是mapreduce,对于小文件都有损效率,实践中,又难免面临处理大量小文件的场景,此时,就需要有相应解决方案.将多个小文件合并 ...
随机推荐
- Web_0003：关于PHP上传文件大小的限制
相关设置如下: 1,file_uploads = on 是否允许通过HTTP上传文件的开关,默认为ON即是开 2,upload_max_filesize = 8m ; 即允许上传文件大小的最大值.默 ...
- apache poi下载流程
apache poi下载官网:http://poi.apache.org/apidocs/4.1/ apache poi下载流程:https://blog.csdn.net/qq_31065001/a ...
- css动画延迟好像有点怪
项目中需要使用到动画animate.css,在自定义的时候发现设置animation-delay 和 animation-duration 的总时间不对会导致 动画缺失. 比如 bounceInLef ...
- 微信小程序调试页面的坑
使用微信开发者工具切新的页面保存刷新无法在左侧直接预览必须在app.json文件配置页面(填写路径但是不用写后缀名),并且把想要预览的页面放在第一个位置.
- Laravel中使用QRcode自制二维码
一.配置 1.在项目根目录输入命令 composer require simplesoftwareio/simple-qrcode 1.3.* 2.在config/app.php 的 provider ...
- Android_侧滑菜单的实现
1.创建侧滑菜单Fragment package com.example.didida_corder; import android.os.Bundle; import android.view.La ...
- nginx 简单理解和配置
1.概念 Nginx是一个高性能的HTTP和反向代理的web服务器,同时也提供了IMAP/POP3/SMTP服务,Nginx是由伊戈尔·塞索耶夫为俄罗斯访问量第二的Rambler.ru站点开发的,第一 ...
- 2019牛客多校第一场H XOR 线性基模板
H XOR 题意 给出一组数,求所有满足异或和为0的子集的长度和 分析 n为1e5,所以枚举子集肯定是不可行的,这种时候我们通常要转化成求每一个数的贡献,对于一组数异或和为0.我们考虑使用线性基,对这 ...
- ubuntu系统定时运行 crontab
1,crontab是个啥? ubuntu系统自带cron工具,cron是一个系统上的定时工具,用它的好处在于,不同的程序可以用同一个计时器,这样就省得不同程序各自sleep了,另外它还支持比较多的个性 ...
- doctype的意思
<!DOCTYPE HTML>这句话在整个网页的最上头,意思是这个网页是一个用html5语法写的,因为还有html4和xhtml等语法. 为了兼容一些旧的页面,浏览器设置了两种解析模式:1 ...