自己定义一个outputFormat

自定义一个OutputFormat，用于输出<Text, MapWritable>格式的数据

MapWritable的内容是 [Text: LongWritable]

输出格式 [url    url2:times2,url3:times3,...]

 

参考TextOutputFormat，修改简化之

 

Java代码  

1. public class TextAndMapWritableOutputFormat extends  
2.         FileOutputFormat<Text, MapWritable> {  
3.   
4.     @Override  
5.     public RecordWriter<Text, MapWritable> getRecordWriter(FileSystem ignored,  
6.             JobConf job, String name, Progressable progress) throws IOException {  
7.         Path file = FileOutputFormat.getTaskOutputPath(job, name);  
8.         FileSystem fs = file.getFileSystem(job);  
9.         FSDataOutputStream fileOut = fs.create(file, progress);  
10.   
11.         return new TextAndMapWritableRecordWriter(fileOut);  
12.     }  
13.   
14.     protected static class TextAndMapWritableRecordWriter implements RecordWriter<Text, MapWritable> {  
15.   
16.         private static final String utf8 = "UTF-8";  
17.         private static final byte[] newline;  
18.         private static final byte[] keyValueSeparator;  
19.         private static final byte[] colon;  
20.         private static final byte[] comma;  
21.         static {  
22.             try {  
23.                 newline = "\n".getBytes(utf8);  
24.                 keyValueSeparator = "\t".getBytes(utf8);  
25.                 colon = ":".getBytes(utf8);  
26.                 comma = ",".getBytes(utf8);  
27.             } catch (UnsupportedEncodingException uee) {  
28.                 throw new IllegalArgumentException("can't find " + utf8  
29.                         + " encoding");  
30.             }  
31.         }  
32.   
33.         protected DataOutputStream out;  
34.   
35.         public TextAndMapWritableRecordWriter(DataOutputStream out) {  
36.             this.out = out;  
37.         }  
38.   
39.         @Override  
40.         public synchronized void write(Text key, MapWritable value)  
41.                 throws IOException {  
42.             out.write(key.getBytes(), 0, key.getLength());  
43.             out.write(keyValueSeparator);  
44.   
45.             Iterator<Writable> it = value.keySet().iterator();  
46.             while (it.hasNext()) {  
47.                 Writable k = it.next();  
48.                 LongWritable v = (LongWritable) value.get(k);  
49.   
50.                 out.write(((Text) k).getBytes());  
51.                 out.write(colon);  
52.                 out.write(v.toString().getBytes(utf8));  
53.                 out.write(comma);  
54.             }  
55.   
56.             out.write(newline);  
57.         }  
58.   
59.         @Override  
60.         public synchronized void close(Reporter reporter) throws IOException {  
61.             out.close();  
62.         }  
63.   
64.     }  
65.   
66. } 
    

转自：http://roserouge.iteye.com/blog/347857