Project

General

Profile

1
package eu.dnetlib.pace.distance;
2

    
3
import java.util.HashSet;
4
import java.util.Map;
5
import java.util.Set;
6

    
7
import org.apache.commons.logging.Log;
8
import org.apache.commons.logging.LogFactory;
9

    
10
import com.google.common.base.Function;
11
import com.google.common.base.Predicates;
12
import com.google.common.collect.Iterables;
13
import com.google.common.collect.Sets;
14
import com.google.common.collect.Sets.SetView;
15

    
16
import eu.dnetlib.pace.model.Field;
17
import eu.dnetlib.pace.model.FieldListImpl;
18
import eu.dnetlib.pace.model.gt.CoAuthor;
19
import eu.dnetlib.pace.model.gt.CoAuthors;
20
import eu.dnetlib.pace.model.gt.GTAuthor;
21

    
22
public class PersonCoAuthorSurnamesDistance extends ConfigurableDistanceAlgo implements DistanceAlgo {
23

    
24
	/**
25
	 * logger.
26
	 */
27
	private static final Log log = LogFactory.getLog(PersonCoAuthorSurnamesDistance.class); // NOPMD by marko on 11/24/08 5:02 PM
28

    
29
	public PersonCoAuthorSurnamesDistance(final Map<String, String> params, final double weight) {
30
		super(params, weight);
31
	}
32

    
33
	@Override
34
	public double distance(final Field a, final Field b) {
35

    
36
		final GTAuthor ga = Iterables.getOnlyElement(GTAuthor.fromOafJson(((FieldListImpl) a).stringList()));
37
		final GTAuthor gb = Iterables.getOnlyElement(GTAuthor.fromOafJson(((FieldListImpl) b).stringList()));
38

    
39
		final Integer commonSurnames = Integer.parseInt(getParams().get("common.surnames"));
40
		// log.info("min common surnames: " + commonSurnames);
41
		if (isEnabled(commonSurnames) && (surnamesInCommon(ga, gb).size() >= commonSurnames)) return 1.0;
42

    
43
		return 0;
44
	}
45

    
46
	private boolean isEnabled(final Integer property) {
47
		return (property != null) && (property >= 0);
48
	}
49

    
50
	private Set<String> surnamesInCommon(final GTAuthor a, final GTAuthor b) {
51
		final SetView<String> intersection = Sets.intersection(getSurnames(a.getCoAuthors()), getSurnames(b.getCoAuthors()));
52
		// log.info(String.format("surnames intersection '%s' - '%s': %s", a.getAuthor().getFullname(), b.getAuthor().getFullname(),
53
		// intersection.size()));
54
		return intersection;
55
	}
56

    
57
	private Set<String> getSurnames(final CoAuthors ca) {
58
		if ((ca == null) || ca.isEmpty()) return new HashSet<String>();
59
		return Sets.newHashSet(Iterables.filter(Iterables.transform(ca, new Function<CoAuthor, String>() {
60

    
61
			@Override
62
			public String apply(final CoAuthor c) {
63
				return c.getSecondnames();
64
			}
65
		}), Predicates.notNull()));
66
	}
67

    
68
	@Override
69
	public double getWeight() {
70
		return getWeigth();
71
	}
72

    
73
}
(20-20/27)