antonis.lempesis
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
1
Fork 1
Code Issues Pull Requests Projects Releases Wiki Activity

apply limits (length, size) to pace Fields

This commit is contained in:
Claudio Atzori 2018-11-20 10:51:38 +01:00
parent c5979ffe18
commit f1c68d8ba3
3 changed files with 43 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

41
dnet-pace-core/src/main/java/eu/dnetlib/pace/distance/DistanceScorer.java
View File

@ -1,23 +1,28 @@
package eu.dnetlib.pace.distance;
import java.util.Collection;
import java.util.List;
import eu.dnetlib.pace.condition.ConditionAlgo;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.distance.eval.ConditionEvalMap;
import eu.dnetlib.pace.distance.eval.DistanceEval;
import eu.dnetlib.pace.distance.eval.DistanceEvalMap;
import eu.dnetlib.pace.distance.eval.ScoreResult;
import eu.dnetlib.pace.model.Document;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;
import eu.dnetlib.pace.model.*;
import eu.dnetlib.pace.util.PaceException;
import org.apache.commons.lang.StringUtils;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import java.util.Collection;
import java.util.List;
import java.util.stream.Collectors;
/**
* The distance between two documents is given by the weighted mean of the field distances
*/
public class DistanceScorer {
private static final Log log = LogFactory.getLog(DistanceScorer.class);
private Config config;
public DistanceScorer(final Config config) {
@ -79,7 +84,7 @@ public class DistanceScorer {
if (va.getType().equals(vb.getType())) {
de.setDistance(w * fd.distanceAlgo().distance(va, vb));
} else {
throw new IllegalArgumentException(String.format("Types are differents type: %s:%s - %s:%s", va, va.getType(), vb, vb.getType()));
throw new PaceException(String.format("Types are different: %s:%s - %s:%s", va, va.getType(), vb, vb.getType()));
}
}
return de;
@ -87,7 +92,27 @@ public class DistanceScorer {
}
private Field getValue(final Document d, final FieldDef fd) {
return d.values(fd.getName());
final Field v = d.values(fd.getName());
if (fd.getLength() > 0) {
if (v instanceof FieldValueImpl) {
((FieldValueImpl) v).setValue(StringUtils.substring(v.stringValue(), 0, fd.getLength()));
} else if (v instanceof FieldListImpl) {
List<String> strings = ((FieldListImpl) v).stringList();
strings = strings.stream()
.limit(fd.getSize() > 0 ? fd.getSize() : strings.size())
.map(s -> StringUtils.substring(s, 0, fd.getLength()))
.collect(Collectors.toList());
((FieldListImpl) v).clear();
((FieldListImpl) v).addAll(strings.stream()
.limit(fd.getSize() > 0 ? fd.getSize() : strings.size())
.map(s -> StringUtils.substring(s, 0, fd.getLength()))
.map(s -> new FieldValueImpl(v.getType(), v.getName(), s))
.collect(Collectors.toList()));
}
}
return v;
}
private double sumWeights(final Collection<FieldDef> fields) {

5
dnet-pace-core/src/main/java/eu/dnetlib/pace/distance/algo/LevensteinTitle.java
View File

@ -2,13 +2,18 @@ package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.DistanceScorer;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import java.util.Map;
@DistanceClass("LevensteinTitle")
public class LevensteinTitle extends SecondStringDistanceAlgo {
private static final Log log = LogFactory.getLog(LevensteinTitle.class);
public LevensteinTitle(Map<String,Number> params){
super(params, new com.wcohen.ss.Levenstein());
}

6
dnet-pace-core/src/main/java/eu/dnetlib/pace/util/PaceException.java
View File

@ -2,8 +2,12 @@ package eu.dnetlib.pace.util;
public class PaceException extends RuntimeException {
public PaceException(String s, Throwable e){
public PaceException(String s, Throwable e) {
super(s, e);
}
public PaceException(String s) {
super(s);
}
}