1
|
package eu.dnetlib.pace.clustering;
|
2
|
|
3
|
import java.util.Collection;
|
4
|
import java.util.List;
|
5
|
import java.util.Map;
|
6
|
|
7
|
import com.google.common.collect.Lists;
|
8
|
import com.google.common.collect.Sets;
|
9
|
import eu.dnetlib.pace.model.Field;
|
10
|
import org.apache.commons.lang.StringUtils;
|
11
|
|
12
|
public class LowercaseClustering extends AbstractClusteringFunction {
|
13
|
|
14
|
public LowercaseClustering(final Map<String, Integer> params) {
|
15
|
super(params);
|
16
|
}
|
17
|
|
18
|
@Override
|
19
|
public Collection<String> apply(List<Field> fields) {
|
20
|
Collection<String> c = Sets.newLinkedHashSet();
|
21
|
for(Field f : fields) {
|
22
|
c.addAll(doApply(f.stringValue()));
|
23
|
}
|
24
|
return c;
|
25
|
}
|
26
|
|
27
|
@Override
|
28
|
protected Collection<String> doApply(final String s) {
|
29
|
if(StringUtils.isBlank(s)) {
|
30
|
return Lists.newArrayList();
|
31
|
}
|
32
|
return Lists.newArrayList(s.toLowerCase().trim());
|
33
|
}
|
34
|
}
|