21、MapReduce读写SequenceFile、MapFile、ORCFile和ParquetFile文件

Hadoop系列文章目录

1、hadoop3.1.4简单介绍及部署、简单验证
2、HDFS操作 - shell客户端
3、HDFS的使用（读写、上传、下载、遍历、查找文件、整个目录拷贝、只拷贝文件、列出文件夹下文件、删除文件及目录、获取文件及文件夹属性等）-java
4、HDFS-java操作类HDFSUtil及junit测试（HDFS的常见操作以及HA环境的配置）
5、HDFS API的RESTful风格–WebHDFS
6、HDFS的HttpFS-代理服务
7、大数据中常见的文件存储格式以及hadoop中支持的压缩算法
8、HDFS内存存储策略支持和“冷热温”存储
9、hadoop高可用HA集群部署及三种方式验证
10、HDFS小文件解决方案–Archive
11、hadoop环境下的Sequence File的读写与合并
12、HDFS Trash垃圾桶回收介绍与示例
13、HDFS Snapshot快照
14、HDFS 透明加密KMS
15、MapReduce介绍及wordcount
16、MapReduce的基本用法示例-自定义序列化、排序、分区、分组和topN
17、MapReduce的分区Partition介绍
18、MapReduce的计数器与通过MapReduce读取/写入数据库示例
19、Join操作map side join 和 reduce side join
20、MapReduce 工作流介绍
21、MapReduce读写SequenceFile、MapFile、ORCFile和ParquetFile文件
22、MapReduce使用Gzip压缩、Snappy压缩和Lzo压缩算法写文件和读取相应的文件
23、hadoop集群中yarn运行mapreduce的内存、CPU分配调度计算与优化

---

---

本文介绍使用MapReduce读写文件，包括：读写SequenceFile、MapFile、ORCFile和ParquetFile文件。
本文前提：hadoop环境可正常使用。pom.xml文件内容参考本专栏中的其他文章内容。
本文分为四部分，即MapReduce读写SequenceFile、MapFile、ORCFile和ParquetFile文件。
下篇文章介绍压缩算法的使用。

关于本文的前置内容介绍，参考链接hdfs的文件系统与压缩算法

一、MapReduce读写SequenceFile

1、写SequenceFile

本示例的写入内容是根据读取的txt文件内容。
使用SequenceFileOutputFormat将结果保存为SequenceFile。
代码示例：

import java.io.IOException;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.SequenceFile;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.mapreduce.lib.output.SequenceFileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;

public class WriteSeqFile extends Configured implements Tool {
	static String in = "D:/workspace/bigdata-component/hadoop/test/in/seq";
	static String out = "D:/workspace/bigdata-component/hadoop/test/out/seq";

	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		int status = ToolRunner.run(conf, new WriteSeqFile(), args);
		System.exit(status);
	}

	/**
	 * 注意文件类型，确定mapper的keyin类型
	 * 如果使用mapper输出，则Mapper的输出keyOut类型需要是非null、text等类型，测试下来LongWritable可以
	 * Mapper
	 * 如果使用mapper-reducer输出，则Mapper输出keyOut类型好像都可以
	 * 
	 * @author alanchan
	 *
	 */
	static class WriteSeqFileMapper extends Mapper<LongWritable, Text, Text, NullWritable> {
		protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
			context.write(value, NullWritable.get());
		}
	}

	static class WriteSeqFileReducer extends Reducer<Text, NullWritable, Text, NullWritable> {
		protected void reduce(Text key, Iterable<NullWritable> values, Context context)
				throws IOException, InterruptedException {
			context.write(key, NullWritable.get());
		}
	}

	@Override
	public int run(String[] args) throws Exception {
		Job job = Job.getInstance(getConf(), this.getClass().getName());
		job.setJarByClass(this.getClass());

		job.setMapperClass(WriteSeqFileMapper.class);
		job.setMapOutputKeyClass(Text.class);
		job.setMapOutputValueClass(NullWritable.class);

		job.setReducerClass(WriteSeqFileReducer.class);
		job.setOutputKeyClass(Text.class);
		job.setOutputValueClass(NullWritable.class);

//		job.setNumReduceTasks(0);

		// 配置作业的输入数据路径
		FileInputFormat.addInputPath(job, new Path(in));

		// 设置作业的输出为SequenceFileOutputFormat
		job.setOutputFormatClass(SequenceFileOutputFormat.class);
		// 使用SequenceFile的块级别压缩
		SequenceFileOutputFormat.setOutputCompressionType(job, SequenceFile.CompressionType.BLOCK);

		Path outputDir = new Path(out);
		outputDir.getFileSystem(this.getConf()).delete(outputDir, true);
		FileOutputFormat.setOutputPath(job, outputDir);

		return job.waitForCompletion(true) ? 0 : 1;
	}

}

2、读SequenceFile

读取本示例中的Sequence文件，生成TextFile文件。
使用SequenceFileInputformat读取SequenceFile。
代码示例：

import java.io.IOException;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.SequenceFileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;

/**
 * @author alanchan
 * 读取SequenceFile文件
 */
public class ReadSeqFile extends Configured implements Tool {
	static String in = "D:/workspace/bigdata-component/hadoop/test/out/seq";;
	static String out = "D:/workspace/bigdata-component/hadoop/test/out/seqread";

	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		int status = ToolRunner.run(conf, new ReadSeqFile(), args);
		System.exit(status);
	}

	@Override
	public int run(String[] args) throws Exception {
		Job job = Job.getInstance(getConf(), this.getClass().getName());
		job.setJarByClass(this.getClass());

		job.setMapperClass(ReadSeqFileMapper.class);
		job.setMapOutputKeyClass(NullWritable.class);
		job.setMapOutputValueClass(Text.class);

		job.setReducerClass(ReadSeqFileReducer.class);
		job.setOutputKeyClass(Text.class);
		job.setOutputValueClass(NullWritable.class);

//		job.setNumReduceTasks(0);

		// 设置作业的输入为SequenceFileInputFormat（SequenceFile文本）
		job.setInputFormatClass(SequenceFileInputFormat.class);
		// 配置作业的输入数据路径
		SequenceFileInputFormat.addInputPath(job, new Path(in));

		Path outputDir = new Path(out);
		outputDir.getFileSystem(this.getConf()).delete(outputDir, true);
		FileOutputFormat.setOutputPath(job, outputDir);

		return job.waitForCompletion(true) ? 0 : 1;
	}

	/**
	 * 特别注意：mapper的输入key类型要根据文件类型来设定，否则会出现类型转换异常
	 * 
	 * @author alanchan
	 *
	 */
	static class ReadSeqFileMapper extends Mapper<NullWritable, Text, NullWritable, Text> {
		protected void map(NullWritable key, Text value, Context context) throws IOException, InterruptedException {
			context.write(NullWritable.get(), value);
		}
	}

	static class ReadSeqFileReducer extends Reducer<NullWritable, Text, Text, NullWritable> {
		protected void reduce(Text key, Iterable<NullWritable> values, Context context)
				throws IOException, InterruptedException {
			context.write(key, NullWritable.get());
		}
	}
}

3、使用SequenceFile合并小文件

将所有的小文件写入到一个Sequence File中，即将文件名作为key，文件内容作为value序列化到Sequence File大文件中。

import java.io.File;
import java.io.FileInputStream;
import java.nio.charset.Charset;
import java.util.ArrayList;
import java.util.List;

import org.apache.commons.codec.digest.DigestUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileUtil;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.BytesWritable;
import org.apache.hadoop.io.SequenceFile;
import org.apache.hadoop.io.SequenceFile.Reader;
import org.apache.hadoop.io.SequenceFile.Writer;
import org.apache.hadoop.io.Text;

public class MergeSmallFilesToSequenceFile {
	private static Configuration configuration = new Configuration();
	static String srcPath = "D:/workspace/bigdata-component/hadoop/test/in/sf";
	static String destPath = "D:/workspace/bigdata-component/hadoop/test/out/sf";

	public static void main(String[] args) throws Exception {
		MergeSmallFilesToSequenceFile msf = new MergeSmallFilesToSequenceFile();
		// 合并小文件
		List<String> fileList = msf.getFileListByPath(srcPath);

		msf.mergeFile(configuration, fileList, destPath);
		// 读取大文件
		msf.readMergedFile(configuration, destPath);
	}

	public List<String> getFileListByPath(String inputPath) throws Exception {
		List<String> smallFilePaths = new ArrayList<String>();
		File file = new File(inputPath);
		// 给定路径是文件夹，则遍历文件夹，将子文件夹中的文件都放入smallFilePaths
		// 给定路径是文件，则把文件的路径放入smallFilePaths
		if (file.isDirectory()) {
			File[] files = FileUtil.listFiles(file);
			for (File sFile : files) {
				smallFilePaths.add(sFile.getPath());
			}
		} else {
			smallFilePaths.add(file.getPath());
		}
		return smallFilePaths;
	}

	// 把smallFilePaths的小文件遍历读取，然后放入合并的sequencefile容器中
	public void mergeFile(Configuration configuration, List<String> smallFilePaths, String destPath) throws Exception {
		Writer.Option bigFile = Writer.file(new Path(destPath));
		Writer.Option keyClass = Writer.keyClass(Text.class);
		Writer.Option valueClass = Writer.valueClass(BytesWritable.class);
		// 构造writer
		Writer writer = SequenceFile.createWriter(configuration, bigFile, keyClass, valueClass);
		// 遍历读取小文件，逐个写入sequencefile
		Text key = new Text();
		for (String path : smallFilePaths) {
			File file = new File(path);
			long fileSize = file.length();// 获取文件的字节数大小
			byte[] fileContent = new byte[(int) fileSize];
			FileInputStream inputStream = new FileInputStream(file);
			inputStream.read(fileContent, 0, (int) fileSize);// 把文件的二进制流加载到fileContent字节数组中去
			String md5Str = DigestUtils.md5Hex(fileContent);
			System.out.println("merge小文件：" + path + ",md5:" + md5Str);
			key.set(path);
			// 把文件路径作为key，文件内容做为value，放入到sequencefile中
			writer.append(key, new BytesWritable(fileContent));
		}
		writer.hflush();
		writer.close();
	}

	// 读取大文件中的小文件
	public void readMergedFile(Configuration configuration, String srcPath) throws Exception {
		Reader.Option file = Reader.file(new Path(srcPath));
		Reader reader = new Reader(configuration, file);
		Text key = new Text();
		BytesWritable value = new BytesWritable();
		while (reader.next(key, value)) {
			byte[] bytes = value.copyBytes();
			String md5 = DigestUtils.md5Hex(bytes);
			String content = new String(bytes, Charset.forName("GBK"));
			System.out.println("读取到文件：" + key + ",md5:" + md5 + ",content:" + content);
		}
	}

}

运行日志输出

2022-09-22 19:16:55,192 WARN zlib.ZlibFactory: Failed to load/initialize native-zlib library
2022-09-22 19:16:55,193 INFO compress.CodecPool: Got brand-new compressor [.deflate]
merge小文件：D:\workspace\bigdata-component\hadoop\test\in\sf\java.txt,md5:b086a9d7084ccea407df5b3215085bd4
merge小文件：D:\workspace\bigdata-component\hadoop\test\in\sf\java1.txt,md5:b086a9d7084ccea407df5b3215085bd4
merge小文件：D:\workspace\bigdata-component\hadoop\test\in\sf\testhadoopclient_java.txt,md5:b086a9d7084ccea407df5b3215085bd4
2022-09-22 19:16:55,209 INFO compress.CodecPool: Got brand-new decompressor [.deflate]
读取到文件：D:\workspace\bigdata-component\hadoop\test\in\sf\java.txt,md5:b086a9d7084ccea407df5b3215085bd4,content:testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
读取到文件：D:\workspace\bigdata-component\hadoop\test\in\sf\java1.txt,md5:b086a9d7084ccea407df5b3215085bd4,content:testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
读取到文件：D:\workspace\bigdata-component\hadoop\test\in\sf\testhadoopclient_java.txt,md5:b086a9d7084ccea407df5b3215085bd4,content:testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt
testhadoopclient_java.txt

二、MapFile

可以理解为MapFile是排序后的SequenceFile，通过观察其结构可以看到MapFile由两部分组成。分别是data和index。data为存储数据的文件，index作为文件的数据索引，主要记录了每个Record的Key值，以及该Record在文件中的偏移位置

1、写MapFile

读取普通TextFile，生成MapFile文件

代码示例

import java.io.IOException;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.mapreduce.lib.output.MapFileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;

public class WriteMapFile extends Configured implements Tool {
	static String in = "D:/workspace/bigdata-component/hadoop/test/in/seq";
	static String out = "D:/workspace/bigdata-component/hadoop/test/out/mapfile";

	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		int status = ToolRunner.run(conf, new WriteMapFile(), args);
		System.exit(status);
	}

	@Override
	public int run(String[] args) throws Exception {
		Job job = Job.getInstance(getConf(), this.getClass().getName());
		job.setJarByClass(this.getClass());

		job.setMapperClass(WriteMapFileMapper.class);
		job.setMapOutputKeyClass(LongWritable.class);
		job.setMapOutputValueClass(Text.class);

		job.setNumReduceTasks(0);

		// 配置作业的输入数据路径
		FileInputFormat.addInputPath(job, new Path(in));

		// 设置作业的输出为MapFileOutputFormat
		job.setOutputFormatClass(MapFileOutputFormat.class);

		Path outputDir = new Path(out);
		outputDir.getFileSystem(this.getConf()).delete(outputDir, true);
		FileOutputFormat.setOutputPath(job, outputDir);

		return job.waitForCompletion(true) ? 0 : 1;
	}

	static class WriteMapFileMapper extends Mapper<LongWritable, Text, LongWritable, Text> {
		protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
			context.write(key, value);
		}
	}

}

运行结果

2、读MapFile

读取MapFile文件，生成普通TextFile文件

1）、实现说明

MapReduce中没有封装MapFile的读取输入类，工作中可根据情况选择以下方案来实现
方案一：自定义InputFormat，使用MapFileOutputFormat中的getReader方法获取读取对象
方案二：使用SequenceFileInputFormat对MapFile的数据进行解析

2）、实现

使用方案二示例

import java.io.IOException;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.input.SequenceFileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;

public class ReadMapFile extends Configured implements Tool {
	static String out = "D:/workspace/bigdata-component/hadoop/test/out/mapfileread";
	static String in = "D:/workspace/bigdata-component/hadoop/test/out/mapfile";

	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		int status = ToolRunner.run(conf, new ReadMapFile(), args);
		System.exit(status);
	}

	@Override
	public int run(String[] args) throws Exception {
		Job job = Job.getInstance(getConf(), this.getClass().getName());
		job.setJarByClass(this.getClass());

		job.setMapperClass(ReadMapFileMapper.class);
		job.setMapOutputKeyClass(NullWritable.class);
		job.setMapOutputValueClass(Text.class);

		job.setNumReduceTasks(0);

		FileInputFormat.addInputPath(job, new Path(in));

        // 设置作业的输入为SequenceFileInputFormat（Hadoop没有直接提供MapFileInput）
//		job.setInputFormatClass(MapFileInputFormat.class);
		job.setInputFormatClass(SequenceFileInputFormat.class);

		Path outputDir = new Path(out);
		outputDir.getFileSystem(this.getConf()).delete(outputDir, true);
		FileOutputFormat.setOutputPath(job, outputDir);

		return job.waitForCompletion(true) ? 0 : 1;
	}

	static class ReadMapFileMapper extends Mapper<LongWritable, Text, NullWritable, Text> {
		protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
			context.write(NullWritable.get(), value);
		}
	}
}

三、ORCFile

1、写ORCFile

读取普通TextFile，生成ORC文件

1）、pom.xml

需要在上文的基础上添加额外的orcfile支持内容

<dependency>
    <groupId>org.apache.orcgroupId>
    <artifactId>orc-shimsartifactId>
    <version>1.6.3version>
dependency>
<dependency>
    <groupId>org.apache.orcgroupId>
    <artifactId>orc-coreartifactId>
    <version>1.6.3version>
dependency>
<dependency>
    <groupId>org.apache.orcgroupId>
    <artifactId>orc-mapreduceartifactId>
    <version>1.6.3version>
dependency>

2）、实现

import java.io.IOException;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
import org.apache.orc.OrcConf;
import org.apache.orc.TypeDescription;
import org.apache.orc.mapred.OrcStruct;
import org.apache.orc.mapreduce.OrcOutputFormat;

/**
 * @author alanchan 
 * 读取普通文本文件转换为ORC文件
 */
public class WriteOrcFile extends Configured implements Tool {
	static String in = "D:/workspace/bigdata-component/hadoop/test/in/orc";
	static String out = "D:/workspace/bigdata-component/hadoop/test/out/orc";

	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		int status = ToolRunner.run(conf, new WriteOrcFile(), args);
		System.exit(status);
	}

	@Override
	public int run(String[] args) throws Exception {
		// 设置Schema
		OrcConf.MAPRED_OUTPUT_SCHEMA.setString(this.getConf(), SCHEMA);

		Job job = Job.getInstance(getConf(), this.getClass().getName());
		job.setJarByClass(this.getClass());

		job.setMapperClass(WriteOrcFileMapper.class);
		job.setMapOutputKeyClass(NullWritable.class);
		job.setMapOutputValueClass(OrcStruct.class);

		job.setNumReduceTasks(0);

		// 配置作业的输入数据路径
		FileInputFormat.addInputPath(job, new Path(in));

		// 设置作业的输出为MapFileOutputFormat
		job.setOutputFormatClass(OrcOutputFormat.class);

		Path outputDir = new Path(out);
		outputDir.getFileSystem(this.getConf()).delete(outputDir, true);
		FileOutputFormat.setOutputPath(job, outputDir);

		return job.waitForCompletion(true) ? 0 : 1;
	}

	// 定义数据的字段信息
//数据格式	
//	id                 ,type  ,orderid              ，bankcard,ctime              ,utime
//	2.0191130220014E+27,ALIPAY,191130-461197476510745,356886,,
//	2.01911302200141E+27,ALIPAY,191130-570038354832903,404118,2019/11/30 21:44,2019/12/16 14:24
//	2.01911302200143E+27,ALIPAY,191130-581296620431058,520083,2019/11/30 18:17,2019/12/4 20:26
//	2.0191201220014E+27,ALIPAY,191201-311567320052455,622688,2019/12/1 10:56,2019/12/16 11:54
	private static final String SCHEMA = "struct";

	static class WriteOrcFileMapper extends Mapper<LongWritable, Text, NullWritable, OrcStruct> {
		// 获取字段描述信息
		private TypeDescription schema = TypeDescription.fromString(SCHEMA);
		// 构建输出的Key
		private final NullWritable outputKey = NullWritable.get();
		// 构建输出的Value为ORCStruct类型
		private final OrcStruct outputValue = (OrcStruct) OrcStruct.createValue(schema);

		protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
			// 将读取到的每一行数据进行分割，得到所有字段
			String[] fields = value.toString().split(",", 6);
			// 将所有字段赋值给Value中的列
			outputValue.setFieldValue(0, new Text(fields[0]));
			outputValue.setFieldValue(1, new Text(fields[1]));
			outputValue.setFieldValue(2, new Text(fields[2]));
			outputValue.setFieldValue(3, new Text(fields[3]));
			outputValue.setFieldValue(4, new Text(fields[4]));
			outputValue.setFieldValue(5, new Text(fields[5]));

			context.write(outputKey, outputValue);
		}
	}

}

运行结果如下

2、读ORCFile

读取ORC文件，转换为普通文本文件
本示例就是读取上一个示例生成的文件。

import java.io.IOException;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
import org.apache.orc.mapred.OrcStruct;
import org.apache.orc.mapreduce.OrcInputFormat;

/**
 * @author alanchan 
 * 读取ORC文件进行解析还原成普通文本文件
 */
public class ReadOrcFile extends Configured implements Tool {
	static String out = "D:/workspace/bigdata-component/hadoop/test/out/orcread";
	static String in = "D:/workspace/bigdata-component/hadoop/test/out/orc";

	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		int status = ToolRunner.run(conf, new ReadOrcFile(), args);
		System.exit(status);
	}

	@Override
	public int run(String[] args) throws Exception {
		Job job = Job.getInstance(getConf(), this.getClass().getName());
		job.setJarByClass(this.getClass());

		job.setMapperClass(ReadOrcFileMapper.class);
		job.setMapOutputKeyClass(NullWritable.class);
		job.setMapOutputValueClass(Text.class);

		job.setNumReduceTasks(0);

		FileInputFormat.addInputPath(job, new Path(in));
		// 設置輸入文件類型
		job.setInputFormatClass(OrcInputFormat.class);

		Path outputDir = new Path(out);
		outputDir.getFileSystem(this.getConf()).delete(outputDir, true);
		FileOutputFormat.setOutputPath(job, outputDir);

		return job.waitForCompletion(true) ? 0 : 1;
	}

	static class ReadOrcFileMapper extends Mapper<NullWritable, OrcStruct, NullWritable, Text> {
		Text outValue = new Text();

		protected void map(NullWritable key, OrcStruct value, Context context)
				throws IOException, InterruptedException {
//			outValue.set(value.toString());
//			value.getFieldValue(0).toString()
			// 或者根據OrcStruct的格式進行獲取值，按照要求進行組裝輸出，本示例僅僅是轉為字符串輸出

			context.write(NullWritable.get(), new Text(value.toString()));
		}
	}

}

运行结果如下：

3、写ORCFile（读取数据库）

读取数据库，转换为ORC文件
pom.xml文件中需要增加mysql的驱动依赖。

源数据记录条数：12606948条
clickhouse系统存储文件大小：50.43 MB
逐条读出存成文本文件大小：1.07G(未压缩)
逐条读出存成ORC文件大小：105M(未压缩)

import java.io.IOException;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Counter;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.db.DBConfiguration;
import org.apache.hadoop.mapreduce.lib.db.DBInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
import org.apache.orc.OrcConf;
import org.apache.orc.TypeDescription;
import org.apache.orc.mapred.OrcStruct;
import org.apache.orc.mapreduce.OrcOutputFormat;
import org.hadoop.mr.db.User;

/**
 * @author alanchan 
 * 从mysql中读取数据，并写入到文件中
 *
 */

public class ReadFromMysqlToOrcFile extends Configured implements Tool {
	private static final String SCHEMA = "struct";
	static String out = "D:/workspace/bigdata-component/hadoop/test/out/mysql";

	static class ReadFromMysqlMapper extends Mapper<LongWritable, User, NullWritable, OrcStruct> {
		private TypeDescription schema = TypeDescription.fromString(SCHEMA);
		private final NullWritable outKey = NullWritable.get();
		private final OrcStruct outValue = (OrcStruct) OrcStruct.createValue(schema);

		protected void map(LongWritable key, User value, Context context) throws IOException, InterruptedException {
			Counter counter = context.getCounter("mysql_records_counters", "User Records");
			counter.increment(1);

			// 将所有字段赋值给Value中的列
			outValue.setFieldValue(0, new IntWritable(value.getId()));
			outValue.setFieldValue(1, new Text(value.getUserName()));
			outValue.setFieldValue(2, new Text(value.getPassword()));
			outValue.setFieldValue(3, new Text(value.getPhone()));
			outValue.setFieldValue(4, new Text(value.getEmail()));
			outValue.setFieldValue(5, new Text(value.getCreateDay()));

			context.write(outKey, outValue);
		}
	}

	@Override
	public int run(String[] args) throws Exception {
		OrcConf.MAPRED_OUTPUT_SCHEMA.setString(this.getConf(), SCHEMA);
		Configuration conf = getConf();

		DBConfiguration.configureDB(conf, "com.mysql.jdbc.Driver", "jdbc:mysql://192.168.10.44:3306/test", "root","root");

		Job job = Job.getInstance(conf, this.getClass().getSimpleName());
		job.setJarByClass(this.getClass());

		job.setInputFormatClass(DBInputFormat.class);
		DBInputFormat.setInput(job, User.class,
		"select id, user_Name,pass_word,phone,email,create_day from dx_user",
		// 12606948 条数据
		"select count(*) from dx_user ");

		Path outputDir = new Path(out);
		outputDir.getFileSystem(this.getConf()).delete(outputDir, true);
		FileOutputFormat.setOutputPath(job, outputDir);

		job.setMapperClass(ReadFromMysqlMapper.class);
		job.setMapOutputKeyClass(NullWritable.class);
		job.setMapOutputValueClass(OrcStruct.class);
		job.setOutputFormatClass(OrcOutputFormat.class);

		job.setNumReduceTasks(0);

		return job.waitForCompletion(true) ? 0 : 1;
	}

	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		int status = ToolRunner.run(conf, new ReadFromMysqlToOrcFile(), args);
		System.exit(status);
	}
}

4、读ORCFile（写入数据库）

读取ORC文件，写入mysql数据库

import java.io.IOException;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.db.DBConfiguration;
import org.apache.hadoop.mapreduce.lib.db.DBOutputFormat;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
import org.apache.orc.mapred.OrcStruct;
import org.apache.orc.mapreduce.OrcInputFormat;
import org.springframework.util.StopWatch;

public class WriteFromOrcFileToMysql extends Configured implements Tool {
	static String in = "D:/workspace/bigdata-component/hadoop/test/out/mysql";

	@Override
	public int run(String[] args) throws Exception {
		Configuration conf = getConf();
		DBConfiguration.configureDB(conf, "com.mysql.jdbc.Driver", "jdbc:mysql://192.168.10.44:3306/test", "root","root");
		Job job = Job.getInstance(conf, this.getClass().getSimpleName());
		job.setJarByClass(this.getClass());

		job.setMapperClass(WriteFromOrcFileToMysqlMapper.class);
		job.setMapOutputKeyClass(User.class);
		job.setMapOutputValueClass(NullWritable.class);

		FileInputFormat.addInputPath(job, new Path(in));
		job.setInputFormatClass(OrcInputFormat.class);
		job.setOutputFormatClass(DBOutputFormat.class);
		// id, user_Name,pass_word,phone,email,create_day
		DBOutputFormat.setOutput(job, "dx_user_copy", "id", "user_name", "pass_word", "phone", "email", "create_day");

//		job.setReducerClass(WriteFromOrcFileToMysqlReducer.class);
//		job.setOutputKeyClass(NullWritable.class);
//		job.setOutputValueClass(Text.class);

		job.setNumReduceTasks(0);

		return job.waitForCompletion(true) ? 0 : 1;
	}

	public static void main(String[] args) throws Exception {
		StopWatch clock = new StopWatch();
		clock.start(WriteFromOrcFileToMysql.class.getSimpleName());

		Configuration conf = new Configuration();

		int status = ToolRunner.run(conf, new WriteFromOrcFileToMysql(), args);

		clock.stop();
		System.out.println(clock.prettyPrint());

		System.exit(status);
	}

	static class WriteFromOrcFileToMysqlMapper extends Mapper<NullWritable, OrcStruct, User, NullWritable> {
		User outValue = new User();
		protected void map(NullWritable key, OrcStruct value, Context context)
				throws IOException, InterruptedException {
			// SCHEMA = "struct";
			outValue.setId(Integer.parseInt(value.getFieldValue("id").toString()));
			outValue.setUserName(value.getFieldValue("userName").toString());
			outValue.setPassword(value.getFieldValue("password").toString());
			outValue.setPhone(value.getFieldValue("phone").toString());
			outValue.setEmail(value.getFieldValue("email").toString());
			outValue.setCreateDay(value.getFieldValue("createDay").toString());
			context.write(outValue,NullWritable.get());
		}

	}
}

四、ParquetFile

1、pom.xml

读写需要增加额外的parquetfile支持的maven依赖

		<dependency>
			<groupId>org.apache.parquetgroupId>
			<artifactId>parquet-hadoopartifactId>
			<version>${parquet.version}version>
		dependency>
		<dependency>
			<groupId>org.apache.parquetgroupId>
			<artifactId>parquet-columnartifactId>
			<version>${parquet.version}version>
		dependency>
		<dependency>
			<groupId>org.apache.parquetgroupId>
			<artifactId>parquet-commonartifactId>
			<version>${parquet.version}version>
		dependency>
		<dependency>
			<groupId>org.apache.parquetgroupId>
			<artifactId>parquet-encodingartifactId>
			<version>${parquet.version}version>
		dependency>

2、写ParquetFile

读取textfile文件，写成parquetfile文件

import java.io.IOException;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.input.TextInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
import org.apache.parquet.example.data.Group;
import org.apache.parquet.example.data.simple.SimpleGroupFactory;
import org.apache.parquet.hadoop.ParquetOutputFormat;
import org.apache.parquet.hadoop.example.GroupWriteSupport;
import org.apache.parquet.hadoop.metadata.CompressionCodecName;
import org.apache.parquet.schema.MessageType;
import org.apache.parquet.schema.OriginalType;
import org.apache.parquet.schema.PrimitiveType.PrimitiveTypeName;
import org.apache.parquet.schema.Types;
import org.springframework.util.StopWatch;

/**
 * @author alanchan
 *
 */
public class WriteParquetFile extends Configured implements Tool {
	static String in = "D:/workspace/bigdata-component/hadoop/test/in/parquet";
	static String out = "D:/workspace/bigdata-component/hadoop/test/out/parquet";

	public static void main(String[] args) throws Exception {
		StopWatch clock = new StopWatch();
		clock.start(WriteParquetFile.class.getSimpleName());

		Configuration conf = new Configuration();
		int status = ToolRunner.run(conf, new WriteParquetFile(), args);
		System.exit(status);

		clock.stop();
		System.out.println(clock.prettyPrint());
	}

	@Override
	public int run(String[] args) throws Exception {
		Configuration conf = getConf();
		// 此demo 输入数据为2列 city ip
		//输入文件格式：https://www.win.com/233434,8283140
		//https://www.win.com/242288,8283139
		MessageType schema = Types.buildMessage().required(PrimitiveTypeName.BINARY).as(OriginalType.UTF8).named("city").required(PrimitiveTypeName.BINARY).as(OriginalType.UTF8)
				.named("ip").named("pair");

		System.out.println("[schema]==" + schema.toString());

		GroupWriteSupport.setSchema(schema, conf);

		Job job = Job.getInstance(conf, this.getClass().getName());
		job.setJarByClass(this.getClass());

		job.setMapperClass(WriteParquetFileMapper.class);
		job.setInputFormatClass(TextInputFormat.class);
		job.setMapOutputKeyClass(NullWritable.class);
		// 设置value是parquet的Group
		job.setMapOutputValueClass(Group.class);
		
		FileInputFormat.setInputPaths(job, in);

		// parquet输出
		job.setOutputFormatClass(ParquetOutputFormat.class);
		ParquetOutputFormat.setWriteSupportClass(job, GroupWriteSupport.class);

		Path outputDir = new Path(out);
		outputDir.getFileSystem(this.getConf()).delete(outputDir, true);
		FileOutputFormat.setOutputPath(job, new Path(out));
//        ParquetOutputFormat.setOutputPath(job, new Path(out));
		ParquetOutputFormat.setCompression(job, CompressionCodecName.SNAPPY);
		job.setNumReduceTasks(0);

		return job.waitForCompletion(true) ? 0 : 1;
	}

	public static class WriteParquetFileMapper extends Mapper<LongWritable, Text, NullWritable, Group> {
		SimpleGroupFactory factory = null;

		protected void setup(Context context) throws IOException, InterruptedException {
			factory = new SimpleGroupFactory(GroupWriteSupport.getSchema(context.getConfiguration()));
		};

		public void map(LongWritable _key, Text ivalue, Context context) throws IOException, InterruptedException {
			Group pair = factory.newGroup();
			//截取输入文件的一行，且是以逗号进行分割
			String[] strs = ivalue.toString().split(",");
			pair.append("city", strs[0]);
			pair.append("ip", strs[1]);
			context.write(null, pair);
		}
	}
}

3、读parquetfile

读取上示例的parquetFile，写成textfile文件

import java.io.IOException;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
import org.apache.parquet.example.data.Group;
import org.apache.parquet.hadoop.ParquetInputFormat;
import org.apache.parquet.hadoop.example.GroupReadSupport;
import org.apache.parquet.hadoop.example.GroupWriteSupport;
import org.apache.parquet.schema.MessageType;
import org.apache.parquet.schema.OriginalType;
import org.apache.parquet.schema.Types;
import org.apache.parquet.schema.PrimitiveType.PrimitiveTypeName;
import org.hadoop.mr.filetype.parquetfile.ParquetReaderAndWriteMRDemo.ParquetReadMapper;
import org.springframework.util.StopWatch;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Mapper.Context;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;

public class ReadParquetFile extends Configured implements Tool {
	static String in = "D:/workspace/bigdata-component/hadoop/test/out/parquet";
	static String out = "D:/workspace/bigdata-component/hadoop/test/out/parquet_read";

	public static void main(String[] args) throws Exception {
		StopWatch clock = new StopWatch();
		clock.start(ReadParquetFile.class.getSimpleName());

		Configuration conf = new Configuration();
		int status = ToolRunner.run(conf, new ReadParquetFile(), args);
		System.exit(status);

		clock.stop();
		System.out.println(clock.prettyPrint());
	}

	@Override
	public int run(String[] args) throws Exception {
		Configuration conf = new Configuration(this.getConf());
		// 此demo 输入数据为2列 city ip
		MessageType schema = Types.buildMessage().required(PrimitiveTypeName.BINARY).as(OriginalType.UTF8).named("city").required(PrimitiveTypeName.BINARY).as(OriginalType.UTF8)
				.named("ip").named("pair");

		System.out.println("[schema]==" + schema.toString());

		GroupWriteSupport.setSchema(schema, conf);
		Job job = Job.getInstance(conf, this.getClass().getName());
		job.setJarByClass(this.getClass());

		// parquet输入
		job.setMapperClass(ReadParquetFileMapper.class);
		job.setNumReduceTasks(0);
		job.setInputFormatClass(ParquetInputFormat.class);
		ParquetInputFormat.setReadSupportClass(job, GroupReadSupport.class);
		FileInputFormat.setInputPaths(job, in);

		job.setOutputKeyClass(NullWritable.class);
		job.setOutputValueClass(Text.class);
		Path outputDir = new Path(out);
		outputDir.getFileSystem(this.getConf()).delete(outputDir, true);
		FileOutputFormat.setOutputPath(job, new Path(out));

		job.setNumReduceTasks(0);

		return job.waitForCompletion(true) ? 0 : 1;
	}

	public static class ReadParquetFileMapper extends Mapper<NullWritable, Group, NullWritable, Text> {
		protected void map(NullWritable key, Group value, Context context) throws IOException, InterruptedException {
			
//			String city = value.getString(0, 0);
//			String ip = value.getString(1, 0);
//			context.write(NullWritable.get(), new Text(city + "," + ip));
			
			String city = value.getString("city", 0);
			String ip = value.getString("ip", 0);
			
			//输出文件格式：https://www.win.com/237516,8284068
			context.write(NullWritable.get(), new Text(value.getString(0, 0) + "," + value.getString(1, 0)));
			
			//输出文件格式：https://www.win.com/237516,8284068
			context.write(NullWritable.get(), new Text(city + "," + ip));
			
			//输出文件格式：
			//city: https://www.win.com/237516
			//ip: 8284068
			context.write(NullWritable.get(), new Text(value.toString()));
			
			context.write(NullWritable.get(), new Text("\n"));

		}
	}
}

至此，MapReduce读写SequenceFile、MapFile、ORCFile和ParquetFile文件操作完成，下片介绍压缩算法的使用。