1
|
package eu.dnetlib.pace.clustering;
|
2
|
|
3
|
import eu.dnetlib.pace.common.AbstractPaceFunctions;
|
4
|
import eu.dnetlib.pace.model.Field;
|
5
|
|
6
|
import java.net.MalformedURLException;
|
7
|
import java.net.URL;
|
8
|
import java.util.Collection;
|
9
|
import java.util.HashSet;
|
10
|
import java.util.List;
|
11
|
import java.util.Map;
|
12
|
import java.util.stream.Collectors;
|
13
|
|
14
|
public class UrlClustering extends AbstractPaceFunctions implements ClusteringFunction {
|
15
|
|
16
|
protected Map<String, Integer> params;
|
17
|
|
18
|
public UrlClustering(final Map<String, Integer> params) {
|
19
|
this.params = params;
|
20
|
}
|
21
|
|
22
|
@Override
|
23
|
public Collection<String> apply(List<Field> fields) {
|
24
|
return fields.stream()
|
25
|
.filter(f -> !f.isEmpty())
|
26
|
.map(Field::stringValue)
|
27
|
.map(this::asUrl)
|
28
|
.map(URL::getHost)
|
29
|
.collect(Collectors.toCollection(HashSet::new));
|
30
|
}
|
31
|
|
32
|
@Override
|
33
|
public Map<String, Integer> getParams() {
|
34
|
return null;
|
35
|
}
|
36
|
|
37
|
private URL asUrl(final String value) {
|
38
|
try {
|
39
|
return new URL(value);
|
40
|
} catch (MalformedURLException e) {
|
41
|
// should not happen as checked by pace typing
|
42
|
throw new IllegalStateException("invalid URL: " + value);
|
43
|
}
|
44
|
}
|
45
|
|
46
|
}
|