@Override
public void map(LongWritable key, Text value, TupleMRContext context, Collector collector)
throws IOException, InterruptedException {
if (tuple == null) {
tuple = new Tuple(context.getTupleMRConfig().getIntermediateSchema(0));
}
StringTokenizer itr = new StringTokenizer(value.toString());
tuple.set(1, 1);
while (itr.hasMoreTokens()) {
tuple.set(0, itr.nextToken());
collector.write(tuple);
}
}
@Test
public void testCompare() {
ArrayList<Field> fields = new ArrayList<Field>();
fields.add(Field.create("int", Field.Type.INT));
Schema s = new Schema("schema", fields);
Criteria cWithCustom =
new Criteria(new OrderBy().add("int", Order.ASC, revIntComp).getElements());
Criteria c = new Criteria(new OrderBy().add("int", Order.ASC).getElements());
Tuple t1 = new Tuple(s);
Tuple t2 = new Tuple(s);
int index[] = new int[] {0};
t1.set("int", 1);
t2.set("int", 2);
SortComparator sortComparator = new SortComparator();
assertPositive(sortComparator.compare(s, cWithCustom, t1, index, t2, index));
assertNegative(sortComparator.compare(s, c, t1, index, t2, index));
}
@Override
public void map(LongWritable key, Text value, TupleMRContext context, Collector collector)
throws IOException, InterruptedException {
user.set("name", "Pere");
user.set("money", 100);
user.set("my_country", "ES");
collector.write(user);
user.set("name", "Iván");
user.set("my_country", "ES");
user.set("money", 50);
collector.write(user);
user.set("my_country", "FR");
user.set("money", 150);
user.set("name", "Eric");
collector.write(user);
country.set("country", "ES");
country.set("averageSalary", 1000);
collector.write(country);
country.set("country", "FR");
country.set("averageSalary", 1500);
collector.write(country);
}
@Test
public void test() throws Exception {
trash(OUTPUT);
Configuration conf = new Configuration();
createInput(INPUT, conf);
ToolRunner.run(getConf(), new TopicFingerprint(), new String[] {INPUT, OUTPUT, 2 + ""});
Path outPath = new Path(OUTPUT + "/part-r-00000");
FileSystem fs = FileSystem.get(outPath.toUri(), conf);
TupleFile.Reader reader = new TupleFile.Reader(fs, conf, outPath);
Tuple tuple = new Tuple(reader.getSchema());
// The order in the output file is deterministic (we have sorted by topic, count)
reader.next(tuple);
assertEquals(1, tuple.get("topic"));
assertEquals("a", tuple.get("word").toString());
reader.next(tuple);
assertEquals(1, tuple.get("topic"));
assertEquals("c", tuple.get("word").toString());
reader.next(tuple);
assertEquals(2, tuple.get("topic"));
assertEquals("a", tuple.get("word").toString());
reader.next(tuple);
assertEquals(2, tuple.get("topic"));
assertEquals("b", tuple.get("word").toString());
// Check the named output
reader.close();
outPath = new Path(OUTPUT + "/" + TopicFingerprint.OUTPUT_TOTALCOUNT + "/" + "part-r-00000");
reader = new TupleFile.Reader(fs, conf, outPath);
tuple = new Tuple(reader.getSchema());
reader.next(tuple);
assertEquals(1, tuple.get("topic"));
assertEquals(15, tuple.get("totalcount"));
reader.next(tuple);
assertEquals(2, tuple.get("topic"));
assertEquals(19, tuple.get("totalcount"));
reader.close();
trash(INPUT, OUTPUT);
}
