sabeel
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

New sources formatted by maven plugin

This commit is contained in:
Giambattista Bloisi 2023-07-06 10:28:53 +02:00
parent bd3fcf869a
commit 801da2fd4a
105 changed files with 5610 additions and 5267 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/AbstractClusteringFunction.java
View File

@ -1,9 +1,5 @@
package eu.dnetlib.pace.clustering;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import org.apache.commons.lang3.StringUtils;
package eu.dnetlib.pace.clustering;
import java.util.Collection;
import java.util.HashSet;
@ -11,6 +7,12 @@ import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;
import org.apache.commons.lang3.StringUtils;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
public abstract class AbstractClusteringFunction extends AbstractPaceFunctions implements ClusteringFunction {
protected Map<String, Integer> params;
@ -23,15 +25,17 @@ public abstract class AbstractClusteringFunction extends AbstractPaceFunctions i
@Override
public Collection<String> apply(Config conf, List<Field> fields) {
return fields.stream().filter(f -> !f.isEmpty())
.map(Field::stringValue)
.map(this::normalize)
.map(s -> filterAllStopWords(s))
.map(s -> doApply(conf, s))
.map(c -> filterBlacklisted(c, ngramBlacklist))
.flatMap(c -> c.stream())
.filter(StringUtils::isNotBlank)
.collect(Collectors.toCollection(HashSet::new));
return fields
.stream()
.filter(f -> !f.isEmpty())
.map(Field::stringValue)
.map(this::normalize)
.map(s -> filterAllStopWords(s))
.map(s -> doApply(conf, s))
.map(c -> filterBlacklisted(c, ngramBlacklist))
.flatMap(c -> c.stream())
.filter(StringUtils::isNotBlank)
.collect(Collectors.toCollection(HashSet::new));
}
public Map<String, Integer> getParams() {

2
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/Acronyms.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.Collection;
@ -6,6 +7,7 @@ import java.util.Set;
import java.util.StringTokenizer;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
@ClusteringClass("acronyms")

79
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/BlacklistAwareClusteringCombiner.java
View File

@ -1,11 +1,5 @@
package eu.dnetlib.pace.clustering;
import com.google.common.collect.Maps;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Document;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldListImpl;
import eu.dnetlib.pace.model.MapDocument;
package eu.dnetlib.pace.clustering;
import java.util.Collection;
import java.util.List;
@ -13,47 +7,54 @@ import java.util.Map;
import java.util.Map.Entry;
import java.util.regex.Pattern;
import com.google.common.collect.Maps;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Document;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldListImpl;
import eu.dnetlib.pace.model.MapDocument;
public class BlacklistAwareClusteringCombiner extends ClusteringCombiner {
public static Collection<String> filterAndCombine(final MapDocument a, final Config conf) {
Document filtered = filter(a, conf.blacklists());
return combine(filtered, conf);
}
public static Collection<String> filterAndCombine(final MapDocument a, final Config conf) {
Document filtered = filter(a, conf.blacklists());
return combine(filtered, conf);
}
private static MapDocument filter(final MapDocument a, final Map<String, List<Pattern>> blacklists) {
if (blacklists == null || blacklists.isEmpty()) {
return a;
}
private static MapDocument filter(final MapDocument a, final Map<String, List<Pattern>> blacklists) {
if (blacklists == null || blacklists.isEmpty()) {
return a;
}
final Map<String, Field> filtered = Maps.newHashMap(a.getFieldMap());
final Map<String, Field> filtered = Maps.newHashMap(a.getFieldMap());
for (final Entry<String, List<Pattern>> e : blacklists.entrySet()) {
Field fields = a.getFieldMap().get(e.getKey());
if (fields != null) {
final FieldListImpl fl = new FieldListImpl();
for (final Entry<String, List<Pattern>> e : blacklists.entrySet()) {
Field fields = a.getFieldMap().get(e.getKey());
if (fields != null) {
final FieldListImpl fl = new FieldListImpl();
for (Field f : fields) {
if (!isBlackListed(f.stringValue(), e.getValue())) {
fl.add(f);
}
}
for (Field f : fields) {
if (!isBlackListed(f.stringValue(), e.getValue())) {
fl.add(f);
}
}
filtered.put(e.getKey(), fl);
}
}
filtered.put(e.getKey(), fl);
}
}
return new MapDocument(a.getIdentifier(), filtered);
}
return new MapDocument(a.getIdentifier(), filtered);
}
private static boolean isBlackListed(String value, List<Pattern> blacklist) {
for (Pattern pattern : blacklist) {
if (pattern.matcher(value).matches()) {
return true;
}
}
private static boolean isBlackListed(String value, List<Pattern> blacklist) {
for (Pattern pattern : blacklist) {
if (pattern.matcher(value).matches()) {
return true;
}
}
return false;
}
return false;
}
}

3
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/ClusteringClass.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.lang.annotation.ElementType;
@ -9,5 +10,5 @@ import java.lang.annotation.Target;
@Target(ElementType.TYPE)
public @interface ClusteringClass {
public String value();
public String value();
}

40
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/ClusteringCombiner.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.ArrayList;
@ -5,6 +6,8 @@ import java.util.Collection;
import java.util.List;
import java.util.stream.Collectors;
import org.apache.commons.lang3.StringUtils;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
@ -12,12 +15,11 @@ import eu.dnetlib.pace.model.ClusteringDef;
import eu.dnetlib.pace.model.Document;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldValueImpl;
import org.apache.commons.lang3.StringUtils;
public class ClusteringCombiner {
private static String SEPARATOR = ":";
private static String COLLAPSE_ON= "collapseOn";
private static String COLLAPSE_ON = "collapseOn";
public static Collection<String> combine(final Document a, final Config conf) {
final Collection<String> res = Sets.newLinkedHashSet();
@ -30,31 +32,33 @@ public class ClusteringCombiner {
if (values instanceof FieldValueImpl) {
fields.add(values);
}
else {
} else {
fields.addAll((List<Field>) values);
}
res.addAll(
cd.clusteringFunction()
.apply(conf, fields)
.stream()
.map(k -> prefix + SEPARATOR +k)
.collect(Collectors.toList())
);
res
.addAll(
cd
.clusteringFunction()
.apply(conf, fields)
.stream()
.map(k -> prefix + SEPARATOR + k)
.collect(Collectors.toList()));
}
}
return res;
}
private static String getPrefix(ClusteringDef cd, String fieldName) {
return cd.getName()+ SEPARATOR +
cd.getParams().keySet()
.stream()
.filter(k -> k.contains(COLLAPSE_ON))
.findFirst()
.map(k -> StringUtils.substringAfter(k, SEPARATOR))
.orElse(fieldName);
return cd.getName() + SEPARATOR +
cd
.getParams()
.keySet()
.stream()
.filter(k -> k.contains(COLLAPSE_ON))
.findFirst()
.map(k -> StringUtils.substringAfter(k, SEPARATOR))
.orElse(fieldName);
}
}

1
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/ClusteringFunction.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.Collection;

2
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/ImmutableFieldValue.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.Collection;
@ -5,6 +6,7 @@ import java.util.List;
import java.util.Map;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.config.Config;
@ClusteringClass("immutablefieldvalue")

80
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/KeywordsClustering.java
View File

@ -1,53 +1,57 @@
package eu.dnetlib.pace.clustering;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import org.apache.commons.lang3.StringUtils;
package eu.dnetlib.pace.clustering;
import java.util.*;
import java.util.stream.Collectors;
import org.apache.commons.lang3.StringUtils;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
@ClusteringClass("keywordsclustering")
public class KeywordsClustering extends AbstractClusteringFunction {
public KeywordsClustering(Map<String, Integer> params) {
super(params);
}
public KeywordsClustering(Map<String, Integer> params) {
super(params);
}
@Override
protected Collection<String> doApply(final Config conf, String s) {
@Override
protected Collection<String> doApply(final Config conf, String s) {
//takes city codes and keywords codes without duplicates
Set<String> keywords = getKeywords(s, conf.translationMap(), params.getOrDefault("windowSize", 4));
Set<String> cities = getCities(s, params.getOrDefault("windowSize", 4));
// takes city codes and keywords codes without duplicates
Set<String> keywords = getKeywords(s, conf.translationMap(), params.getOrDefault("windowSize", 4));
Set<String> cities = getCities(s, params.getOrDefault("windowSize", 4));
//list of combination to return as result
final Collection<String> combinations = new LinkedHashSet<String>();
// list of combination to return as result
final Collection<String> combinations = new LinkedHashSet<String>();
for (String keyword: keywordsToCodes(keywords, conf.translationMap())){
for (String city: citiesToCodes(cities)) {
combinations.add(keyword+"-"+city);
if (combinations.size()>=params.getOrDefault("max", 2)) {
return combinations;
}
}
}
for (String keyword : keywordsToCodes(keywords, conf.translationMap())) {
for (String city : citiesToCodes(cities)) {
combinations.add(keyword + "-" + city);
if (combinations.size() >= params.getOrDefault("max", 2)) {
return combinations;
}
}
}
return combinations;
}
return combinations;
}
@Override
public Collection<String> apply(final Config conf, List<Field> fields) {
return fields.stream().filter(f -> !f.isEmpty())
.map(Field::stringValue)
.map(this::cleanup)
.map(this::normalize)
.map(s -> filterAllStopWords(s))
.map(s -> doApply(conf, s))
.map(c -> filterBlacklisted(c, ngramBlacklist))
.flatMap(c -> c.stream())
.filter(StringUtils::isNotBlank)
.collect(Collectors.toCollection(HashSet::new));
}
@Override
public Collection<String> apply(final Config conf, List<Field> fields) {
return fields
.stream()
.filter(f -> !f.isEmpty())
.map(Field::stringValue)
.map(this::cleanup)
.map(this::normalize)
.map(s -> filterAllStopWords(s))
.map(s -> doApply(conf, s))
.map(c -> filterBlacklisted(c, ngramBlacklist))
.flatMap(c -> c.stream())
.filter(StringUtils::isNotBlank)
.collect(Collectors.toCollection(HashSet::new));
}
}

122
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/LastNameFirstInitial.java
View File

@ -1,77 +1,81 @@
package eu.dnetlib.pace.clustering;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.Person;
import org.apache.commons.lang3.StringUtils;
package eu.dnetlib.pace.clustering;
import java.util.*;
import java.util.stream.Collectors;
import org.apache.commons.lang3.StringUtils;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.Person;
@ClusteringClass("lnfi")
public class LastNameFirstInitial extends AbstractClusteringFunction{
public class LastNameFirstInitial extends AbstractClusteringFunction {
private boolean DEFAULT_AGGRESSIVE = true;
private boolean DEFAULT_AGGRESSIVE = true;
public LastNameFirstInitial(final Map<String, Integer> params) {
super(params);
}
public LastNameFirstInitial(final Map<String, Integer> params) {
super(params);
}
@Override
public Collection<String> apply(Config conf, List<Field> fields) {
return fields.stream().filter(f -> !f.isEmpty())
.map(Field::stringValue)
.map(this::normalize)
.map(s -> doApply(conf, s))
.map(c -> filterBlacklisted(c, ngramBlacklist))
.flatMap(c -> c.stream())
.filter(StringUtils::isNotBlank)
.collect(Collectors.toCollection(HashSet::new));
}
@Override
public Collection<String> apply(Config conf, List<Field> fields) {
return fields
.stream()
.filter(f -> !f.isEmpty())
.map(Field::stringValue)
.map(this::normalize)
.map(s -> doApply(conf, s))
.map(c -> filterBlacklisted(c, ngramBlacklist))
.flatMap(c -> c.stream())
.filter(StringUtils::isNotBlank)
.collect(Collectors.toCollection(HashSet::new));
}
@Override
protected String normalize(final String s) {
return fixAliases(transliterate(nfd(unicodeNormalization(s))))
// do not compact the regexes in a single expression, would cause StackOverflowError in case of large input strings
.replaceAll("[^ \\w]+", "")
.replaceAll("(\\p{InCombiningDiacriticalMarks})+", "")
.replaceAll("(\\p{Punct})+", " ")
.replaceAll("(\\d)+", " ")
.replaceAll("(\\n)+", " ")
.trim();
}
@Override
protected String normalize(final String s) {
return fixAliases(transliterate(nfd(unicodeNormalization(s))))
// do not compact the regexes in a single expression, would cause StackOverflowError in case of large input
// strings
.replaceAll("[^ \\w]+", "")
.replaceAll("(\\p{InCombiningDiacriticalMarks})+", "")
.replaceAll("(\\p{Punct})+", " ")
.replaceAll("(\\d)+", " ")
.replaceAll("(\\n)+", " ")
.trim();
}
@Override
protected Collection<String> doApply(final Config conf, final String s) {
@Override
protected Collection<String> doApply(final Config conf, final String s) {
final List<String> res = Lists.newArrayList();
final List<String> res = Lists.newArrayList();
final boolean aggressive = (Boolean) (getParams().containsKey("aggressive") ? getParams().get("aggressive") : DEFAULT_AGGRESSIVE);
final boolean aggressive = (Boolean) (getParams().containsKey("aggressive") ? getParams().get("aggressive")
: DEFAULT_AGGRESSIVE);
Person p = new Person(s, aggressive);
Person p = new Person(s, aggressive);
if (p.isAccurate()) {
String lastName = p.getNormalisedSurname().toLowerCase();
String firstInitial = p.getNormalisedFirstName().toLowerCase().substring(0,1);
if (p.isAccurate()) {
String lastName = p.getNormalisedSurname().toLowerCase();
String firstInitial = p.getNormalisedFirstName().toLowerCase().substring(0, 1);
res.add(firstInitial.concat(lastName));
}
else { // is not accurate, meaning it has no defined name and surname
List<String> fullname = Arrays.asList(p.getNormalisedFullname().split(" "));
if (fullname.size() == 1) {
res.add(p.getNormalisedFullname().toLowerCase());
}
else if (fullname.size() == 2) {
res.add(fullname.get(0).substring(0,1).concat(fullname.get(1)).toLowerCase());
res.add(fullname.get(1).substring(0,1).concat(fullname.get(0)).toLowerCase());
}
else {
res.add(fullname.get(0).substring(0,1).concat(fullname.get(fullname.size()-1)).toLowerCase());
res.add(fullname.get(fullname.size()-1).substring(0,1).concat(fullname.get(0)).toLowerCase());
}
}
res.add(firstInitial.concat(lastName));
} else { // is not accurate, meaning it has no defined name and surname
List<String> fullname = Arrays.asList(p.getNormalisedFullname().split(" "));
if (fullname.size() == 1) {
res.add(p.getNormalisedFullname().toLowerCase());
} else if (fullname.size() == 2) {
res.add(fullname.get(0).substring(0, 1).concat(fullname.get(1)).toLowerCase());
res.add(fullname.get(1).substring(0, 1).concat(fullname.get(0)).toLowerCase());
} else {
res.add(fullname.get(0).substring(0, 1).concat(fullname.get(fullname.size() - 1)).toLowerCase());
res.add(fullname.get(fullname.size() - 1).substring(0, 1).concat(fullname.get(0)).toLowerCase());
}
}
return res;
}
return res;
}
}

9
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/LowercaseClustering.java
View File

@ -1,14 +1,17 @@
package eu.dnetlib.pace.clustering;
import java.util.Collection;
import java.util.List;
import java.util.Map;
import org.apache.commons.lang3.StringUtils;
import com.google.common.collect.Lists;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import org.apache.commons.lang3.StringUtils;
@ClusteringClass("lowercase")
public class LowercaseClustering extends AbstractClusteringFunction {
@ -20,7 +23,7 @@ public class LowercaseClustering extends AbstractClusteringFunction {
@Override
public Collection<String> apply(Config conf, List<Field> fields) {
Collection<String> c = Sets.newLinkedHashSet();
for(Field f : fields) {
for (Field f : fields) {
c.addAll(doApply(conf, f.stringValue()));
}
return c;
@ -28,7 +31,7 @@ public class LowercaseClustering extends AbstractClusteringFunction {
@Override
protected Collection<String> doApply(final Config conf, final String s) {
if(StringUtils.isBlank(s)) {
if (StringUtils.isBlank(s)) {
return Lists.newArrayList();
}
return Lists.newArrayList(s.toLowerCase().trim());

8
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/NGramUtils.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.Set;
@ -10,11 +11,14 @@ public class NGramUtils extends AbstractPaceFunctions {
private static final int SIZE = 100;
private static Set<String> stopwords = AbstractPaceFunctions.loadFromClasspath("/eu/dnetlib/pace/config/stopwords_en.txt");
private static Set<String> stopwords = AbstractPaceFunctions
.loadFromClasspath("/eu/dnetlib/pace/config/stopwords_en.txt");
public static String cleanupForOrdering(String s) {
NGramUtils utils = new NGramUtils();
return (utils.filterStopWords(utils.normalize(s), stopwords) + StringUtils.repeat(" ", SIZE)).substring(0, SIZE).replaceAll(" ", "");
return (utils.filterStopWords(utils.normalize(s), stopwords) + StringUtils.repeat(" ", SIZE))
.substring(0, SIZE)
.replaceAll(" ", "");
}
}

4
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/NgramPairs.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.Collection;
@ -6,6 +7,7 @@ import java.util.List;
import java.util.Map;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.config.Config;
@ClusteringClass("ngrampairs")
@ -28,7 +30,7 @@ public class NgramPairs extends Ngrams {
break;
}
res.add(ngrams.get(i) + ngrams.get(j));
//System.out.println("-- " + concatNgrams);
// System.out.println("-- " + concatNgrams);
}
return res;
}

7
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/Ngrams.java
View File

@ -1,9 +1,10 @@
package eu.dnetlib.pace.clustering;
import eu.dnetlib.pace.config.Config;
import java.util.*;
import eu.dnetlib.pace.config.Config;
@ClusteringClass("ngrams")
public class Ngrams extends AbstractClusteringFunction {
@ -36,7 +37,7 @@ public class Ngrams extends AbstractClusteringFunction {
}
}
}
//System.out.println(ngrams + " n: " + ngrams.size());
// System.out.println(ngrams + " n: " + ngrams.size());
return ngrams;
}

20
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/PersonClustering.java
View File

@ -1,17 +1,20 @@
package eu.dnetlib.pace.clustering;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.Person;
import org.apache.commons.lang3.StringUtils;
package eu.dnetlib.pace.clustering;
import java.util.Collection;
import java.util.List;
import java.util.Map;
import java.util.Set;
import org.apache.commons.lang3.StringUtils;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.Person;
@ClusteringClass("personClustering")
public class PersonClustering extends AbstractPaceFunctions implements ClusteringFunction {
@ -31,7 +34,8 @@ public class PersonClustering extends AbstractPaceFunctions implements Clusterin
final Person person = new Person(f.stringValue(), false);
if (StringUtils.isNotBlank(person.getNormalisedFirstName()) && StringUtils.isNotBlank(person.getNormalisedSurname())) {
if (StringUtils.isNotBlank(person.getNormalisedFirstName())
&& StringUtils.isNotBlank(person.getNormalisedSurname())) {
hashes.add(firstLC(person.getNormalisedFirstName()) + person.getNormalisedSurname().toLowerCase());
} else {
for (final String token1 : tokens(f.stringValue(), MAX_TOKENS)) {

4
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/PersonHash.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.Collection;
@ -22,7 +23,8 @@ public class PersonHash extends AbstractClusteringFunction {
protected Collection<String> doApply(final Config conf, final String s) {
final List<String> res = Lists.newArrayList();
final boolean aggressive = (Boolean) (getParams().containsKey("aggressive") ? getParams().get("aggressive") : DEFAULT_AGGRESSIVE);
final boolean aggressive = (Boolean) (getParams().containsKey("aggressive") ? getParams().get("aggressive")
: DEFAULT_AGGRESSIVE);
res.add(new Person(s, aggressive).hash());

5
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/RandomClusteringFunction.java
View File

@ -1,10 +1,11 @@
package eu.dnetlib.pace.clustering;
import eu.dnetlib.pace.config.Config;
package eu.dnetlib.pace.clustering;
import java.util.Collection;
import java.util.Map;
import eu.dnetlib.pace.config.Config;
public class RandomClusteringFunction extends AbstractClusteringFunction {
public RandomClusteringFunction(Map<String, Integer> params) {

6
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/SortedNgramPairs.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.*;
@ -5,6 +6,7 @@ import java.util.*;
import com.google.common.base.Joiner;
import com.google.common.base.Splitter;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.config.Config;
@ClusteringClass("sortedngrampairs")
@ -21,7 +23,9 @@ public class SortedNgramPairs extends NgramPairs {
Collections.sort(tokens);
return ngramPairs(Lists.newArrayList(getNgrams(Joiner.on(" ").join(tokens), param("ngramLen"), param("max") * 2, 1, 2)), param("max"));
return ngramPairs(
Lists.newArrayList(getNgrams(Joiner.on(" ").join(tokens), param("ngramLen"), param("max") * 2, 1, 2)),
param("max"));
}
}

9
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/SpaceTrimmingFieldValue.java
View File

@ -1,15 +1,17 @@
package eu.dnetlib.pace.clustering;
import java.util.Collection;
import java.util.List;
import java.util.Map;
import eu.dnetlib.pace.config.Config;
import org.apache.commons.lang3.RandomStringUtils;
import org.apache.commons.lang3.StringUtils;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.config.Config;
@ClusteringClass("spacetrimmingfieldvalue")
public class SpaceTrimmingFieldValue extends AbstractClusteringFunction {
@ -21,7 +23,10 @@ public class SpaceTrimmingFieldValue extends AbstractClusteringFunction {
protected Collection<String> doApply(final Config conf, final String s) {
final List<String> res = Lists.newArrayList();
res.add(StringUtils.isBlank(s) ? RandomStringUtils.random(getParams().get("randomLength")) : s.toLowerCase().replaceAll("\\s+", ""));
res
.add(
StringUtils.isBlank(s) ? RandomStringUtils.random(getParams().get("randomLength"))
: s.toLowerCase().replaceAll("\\s+", ""));
return res;
}

2
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/SuffixPrefix.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.Collection;
@ -5,6 +6,7 @@ import java.util.Map;
import java.util.Set;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
@ClusteringClass("suffixprefix")

70
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/UrlClustering.java
View File

@ -1,8 +1,5 @@
package eu.dnetlib.pace.clustering;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
package eu.dnetlib.pace.clustering;
import java.net.MalformedURLException;
import java.net.URL;
@ -12,43 +9,46 @@ import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
@ClusteringClass("urlclustering")
public class UrlClustering extends AbstractPaceFunctions implements ClusteringFunction {
protected Map<String, Integer> params;
protected Map<String, Integer> params;
public UrlClustering(final Map<String, Integer> params) {
this.params = params;
}
public UrlClustering(final Map<String, Integer> params) {
this.params = params;
}
@Override
public Collection<String> apply(final Config conf, List<Field> fields) {
try {
return fields.stream()
.filter(f -> !f.isEmpty())
.map(Field::stringValue)
.map(this::asUrl)
.map(URL::getHost)
.collect(Collectors.toCollection(HashSet::new));
}
catch (IllegalStateException e){
return new HashSet<>();
}
}
@Override
public Collection<String> apply(final Config conf, List<Field> fields) {
try {
return fields
.stream()
.filter(f -> !f.isEmpty())
.map(Field::stringValue)
.map(this::asUrl)
.map(URL::getHost)
.collect(Collectors.toCollection(HashSet::new));
} catch (IllegalStateException e) {
return new HashSet<>();
}
}
@Override
public Map<String, Integer> getParams() {
return null;
}
private URL asUrl(String value) {
try {
return new URL(value);
} catch (MalformedURLException e) {
// should not happen as checked by pace typing
throw new IllegalStateException("invalid URL: " + value);
}
}
@Override
public Map<String, Integer> getParams() {
return null;
}
private URL asUrl(String value) {
try {
return new URL(value);
} catch (MalformedURLException e) {
// should not happen as checked by pace typing
throw new IllegalStateException("invalid URL: " + value);
}
}
}

129
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/WordsStatsSuffixPrefixChain.java
View File

@ -1,90 +1,91 @@
package eu.dnetlib.pace.clustering;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
package eu.dnetlib.pace.clustering;
import java.util.*;
import java.util.stream.Collectors;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
@ClusteringClass("wordsStatsSuffixPrefixChain")
public class WordsStatsSuffixPrefixChain extends AbstractClusteringFunction {
public WordsStatsSuffixPrefixChain(Map<String, Integer> params) {
super(params);
}
public WordsStatsSuffixPrefixChain(Map<String, Integer> params) {
super(params);
}
@Override
protected Collection<String> doApply(Config conf, String s) {
return suffixPrefixChain(s, param("mod"));
}
@Override
protected Collection<String> doApply(Config conf, String s) {
return suffixPrefixChain(s, param("mod"));
}
private Collection<String> suffixPrefixChain(String s, int mod) {
private Collection<String> suffixPrefixChain(String s, int mod) {
//create the list of words from the string (remove short words)
List<String> wordsList =
Arrays.stream(s.split(" "))
.filter(si -> si.length() > 3)
.collect(Collectors.toList());
// create the list of words from the string (remove short words)
List<String> wordsList = Arrays
.stream(s.split(" "))
.filter(si -> si.length() > 3)
.collect(Collectors.toList());
final int words = wordsList.size();
final int letters = s.length();
final int words = wordsList.size();
final int letters = s.length();
//create the prefix: number of words + number of letters/mod
String prefix = words + "-" + letters/mod + "-";
// create the prefix: number of words + number of letters/mod
String prefix = words + "-" + letters / mod + "-";
return doSuffixPrefixChain(wordsList, prefix);
return doSuffixPrefixChain(wordsList, prefix);
}
}
private Collection<String> doSuffixPrefixChain(List<String> wordsList, String prefix) {
private Collection<String> doSuffixPrefixChain(List<String> wordsList, String prefix) {
Set<String> set = Sets.newLinkedHashSet();
switch(wordsList.size()){
case 0:
case 1:
break;
case 2:
set.add(
prefix +
suffix(wordsList.get(0), 3) +
prefix(wordsList.get(1), 3)
);
Set<String> set = Sets.newLinkedHashSet();
switch (wordsList.size()) {
case 0:
case 1:
break;
case 2:
set
.add(
prefix +
suffix(wordsList.get(0), 3) +
prefix(wordsList.get(1), 3));
set.add(
prefix +
prefix(wordsList.get(0), 3) +
suffix(wordsList.get(1), 3)
);
set
.add(
prefix +
prefix(wordsList.get(0), 3) +
suffix(wordsList.get(1), 3));
break;
default:
set.add(
prefix +
suffix(wordsList.get(0), 3) +
prefix(wordsList.get(1), 3) +
suffix(wordsList.get(2), 3)
);
break;
default:
set
.add(
prefix +
suffix(wordsList.get(0), 3) +
prefix(wordsList.get(1), 3) +
suffix(wordsList.get(2), 3));
set.add(
prefix +
prefix(wordsList.get(0), 3) +
suffix(wordsList.get(1), 3) +
prefix(wordsList.get(2), 3)
);
break;
}
set
.add(
prefix +
prefix(wordsList.get(0), 3) +
suffix(wordsList.get(1), 3) +
prefix(wordsList.get(2), 3));
break;
}
return set;
return set;
}
}
private String suffix(String s, int len) {
return s.substring(s.length() - len);
}
private String suffix(String s, int len) {
return s.substring(s.length()-len);
}
private String prefix(String s, int len) {
return s.substring(0, len);
}
private String prefix(String s, int len) {
return s.substring(0, len);
}
}

74
dhp-pace-core/src/main/java/eu/dnetlib/pace/clustering/WordsSuffixPrefix.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.clustering;
import java.util.Collection;
@ -5,53 +6,54 @@ import java.util.Map;
import java.util.Set;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
@ClusteringClass("wordssuffixprefix")
public class WordsSuffixPrefix extends AbstractClusteringFunction {
public WordsSuffixPrefix(Map<String, Integer> params) {
super(params);
}
public WordsSuffixPrefix(Map<String, Integer> params) {
super(params);
}
@Override
protected Collection<String> doApply(Config conf, String s) {
return suffixPrefix(s, param("len"), param("max"));
}
@Override
protected Collection<String> doApply(Config conf, String s) {
return suffixPrefix(s, param("len"), param("max"));
}
private Collection<String> suffixPrefix(String s, int len, int max) {
private Collection<String> suffixPrefix(String s, int len, int max) {
final int words = s.split(" ").length;
final int words = s.split(" ").length;
// adjust the token length according to the number of words
switch (words) {
case 1:
return Sets.newLinkedHashSet();
case 2:
return doSuffixPrefix(s, len+2, max, words);
case 3:
return doSuffixPrefix(s, len+1, max, words);
default:
return doSuffixPrefix(s, len, max, words);
}
}
// adjust the token length according to the number of words
switch (words) {
case 1:
return Sets.newLinkedHashSet();
case 2:
return doSuffixPrefix(s, len + 2, max, words);
case 3:
return doSuffixPrefix(s, len + 1, max, words);
default:
return doSuffixPrefix(s, len, max, words);
}
}
private Collection<String> doSuffixPrefix(String s, int len, int max, int words) {
final Set<String> bigrams = Sets.newLinkedHashSet();
int i = 0;
while (++i < s.length() && bigrams.size() < max) {
int j = s.indexOf(" ", i);
private Collection<String> doSuffixPrefix(String s, int len, int max, int words) {
final Set<String> bigrams = Sets.newLinkedHashSet();
int i = 0;
while (++i < s.length() && bigrams.size() < max) {
int j = s.indexOf(" ", i);
int offset = j + len + 1 < s.length() ? j + len + 1 : s.length();
int offset = j + len + 1 < s.length() ? j + len + 1 : s.length();
if (j - len > 0) {
String bigram = s.substring(j - len, offset).replaceAll(" ", "").trim();
if (bigram.length() >= 4) {
bigrams.add(words+bigram);
}
}
}
return bigrams;
}
if (j - len > 0) {
String bigram = s.substring(j - len, offset).replaceAll(" ", "").trim();
if (bigram.length() >= 4) {
bigrams.add(words + bigram);
}
}
}
return bigrams;
}
}

684
dhp-pace-core/src/main/java/eu/dnetlib/pace/common/AbstractPaceFunctions.java
View File

@ -1,16 +1,5 @@
package eu.dnetlib.pace.common;
import com.google.common.base.Joiner;
import com.google.common.base.Splitter;
import com.google.common.collect.Iterables;
import com.google.common.collect.Sets;
import com.ibm.icu.text.Transliterator;
import eu.dnetlib.pace.clustering.NGramUtils;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.model.FieldListImpl;
import org.apache.commons.io.IOUtils;
import org.apache.commons.lang3.StringUtils;
package eu.dnetlib.pace.common;
import java.io.IOException;
import java.io.StringWriter;
@ -21,6 +10,20 @@ import java.util.regex.Matcher;
import java.util.regex.Pattern;
import java.util.stream.Collectors;
import org.apache.commons.io.IOUtils;
import org.apache.commons.lang3.StringUtils;
import com.google.common.base.Joiner;
import com.google.common.base.Splitter;
import com.google.common.collect.Iterables;
import com.google.common.collect.Sets;
import com.ibm.icu.text.Transliterator;
import eu.dnetlib.pace.clustering.NGramUtils;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.model.FieldListImpl;
/**
* Set of common functions for the framework
*
@ -28,330 +31,337 @@ import java.util.stream.Collectors;
*/
public abstract class AbstractPaceFunctions {
//city map to be used when translating the city names into codes
private static Map<String, String> cityMap = AbstractPaceFunctions.loadMapFromClasspath("/eu/dnetlib/pace/config/city_map.csv");
//list of stopwords in different languages
protected static Set<String> stopwords_gr = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_gr.txt");
protected static Set<String> stopwords_en = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_en.txt");
protected static Set<String> stopwords_de = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_de.txt");
protected static Set<String> stopwords_es = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_es.txt");
protected static Set<String> stopwords_fr = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_fr.txt");
protected static Set<String> stopwords_it = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_it.txt");
protected static Set<String> stopwords_pt = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_pt.txt");
//transliterator
protected static Transliterator transliterator = Transliterator.getInstance("Any-Eng");
//blacklist of ngrams: to avoid generic keys
protected static Set<String> ngramBlacklist = loadFromClasspath("/eu/dnetlib/pace/config/ngram_blacklist.txt");
//html regex for normalization
public final String HTML_REGEX = "<[^>]*>";
private static final String alpha = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789 ";
private static final String aliases_from = "⁰¹²³⁴⁵⁶⁷⁸⁹⁺⁻⁼⁽⁾ⁿ₀₁₂₃₄₅₆₇₈₉₊₋₌₍₎àáâäæãåāèéêëēėęəîïíīįìôöòóœøōõûüùúūßśšłžźżçćčñń";
private static final String aliases_to = "0123456789+-=()n0123456789+-=()aaaaaaaaeeeeeeeeiiiiiioooooooouuuuussslzzzcccnn";
//doi prefix for normalization
public final String DOI_PREFIX = "(https?:\\/\\/dx\\.doi\\.org\\/)|(doi:)";
private Pattern numberPattern = Pattern.compile("-?\\d+(\\.\\d+)?");
private Pattern hexUnicodePattern = Pattern.compile("\\\\u(\\p{XDigit}{4})");
protected final static FieldList EMPTY_FIELD = new FieldListImpl();
protected String concat(final List<String> l) {
return Joiner.on(" ").skipNulls().join(l);
}
protected String cleanup(final String s) {
final String s1 = s.replaceAll(HTML_REGEX, "");
final String s2 = unicodeNormalization(s1.toLowerCase());
final String s3 = nfd(s2);
final String s4 = fixXML(s3);
final String s5 = s4.replaceAll("([0-9]+)", " $1 ");
final String s6 = transliterate(s5);
final String s7 = fixAliases(s6);
final String s8 = s7.replaceAll("[^\\p{ASCII}]", "");
final String s9 = s8.replaceAll("[\\p{Punct}]", " ");
final String s10 = s9.replaceAll("\\n", " ");
final String s11 = s10.replaceAll("(?m)\\s+", " ");
final String s12 = s11.trim();
return s12;
}
protected String fixXML(final String a){
return a.replaceAll("&ndash;", " ")
.replaceAll("&amp;", " ")
.replaceAll("&quot;", " ")
.replaceAll("&minus;", " ");
}
protected boolean checkNumbers(final String a, final String b) {
final String numbersA = getNumbers(a);
final String numbersB = getNumbers(b);
final String romansA = getRomans(a);
final String romansB = getRomans(b);
return !numbersA.equals(numbersB) || !romansA.equals(romansB);
}
protected String getRomans(final String s) {
final StringBuilder sb = new StringBuilder();
for (final String t : s.split(" ")) {
sb.append(isRoman(t) ? t : "");
}
return sb.toString();
}
protected boolean isRoman(final String s) {
return s.replaceAll("^M{0,4}(CM|CD|D?C{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})$", "qwertyuiop").equals("qwertyuiop");
}
protected String getNumbers(final String s) {
final StringBuilder sb = new StringBuilder();
for (final String t : s.split(" ")) {
sb.append(isNumber(t) ? t : "");
}
return sb.toString();
}
public boolean isNumber(String strNum) {
if (strNum == null) {
return false;
}
return numberPattern.matcher(strNum).matches();
}
protected static String fixAliases(final String s) {
final StringBuilder sb = new StringBuilder();
s.chars().forEach(ch -> {
final int i = StringUtils.indexOf(aliases_from, ch);
sb.append(i >= 0 ? aliases_to.charAt(i) : (char)ch);
});
return sb.toString();
}
protected static String transliterate(final String s) {
try {
return transliterator.transliterate(s);
}
catch(Exception e) {
return s;
}
}
protected String removeSymbols(final String s) {
final StringBuilder sb = new StringBuilder();
s.chars().forEach(ch -> {
sb.append(StringUtils.contains(alpha, ch) ? (char)ch : ' ');
});
return sb.toString().replaceAll("\\s+", " ");
}
protected String getFirstValue(final Field values) {
return (values != null) && !Iterables.isEmpty(values) ? Iterables.getFirst(values, EMPTY_FIELD).stringValue() : "";
}
protected boolean notNull(final String s) {
return s != null;
}
protected String normalize(final String s) {
return fixAliases(transliterate(nfd(unicodeNormalization(s))))
.toLowerCase()
// do not compact the regexes in a single expression, would cause StackOverflowError in case of large input strings
.replaceAll("[^ \\w]+", "")
.replaceAll("(\\p{InCombiningDiacriticalMarks})+", "")
.replaceAll("(\\p{Punct})+", " ")
.replaceAll("(\\d)+", " ")
.replaceAll("(\\n)+", " ")
.trim();
}
public String nfd(final String s) {
return Normalizer.normalize(s, Normalizer.Form.NFD);
}
public String utf8(final String s) {
byte[] bytes = s.getBytes(StandardCharsets.UTF_8);
return new String(bytes, StandardCharsets.UTF_8);
}
public String unicodeNormalization(final String s) {
Matcher m = hexUnicodePattern.matcher(s);
StringBuffer buf = new StringBuffer(s.length());
while (m.find()) {
String ch = String.valueOf((char) Integer.parseInt(m.group(1), 16));
m.appendReplacement(buf, Matcher.quoteReplacement(ch));
}
m.appendTail(buf);
return buf.toString();
}
protected String filterStopWords(final String s, final Set<String> stopwords) {
final StringTokenizer st = new StringTokenizer(s);
final StringBuilder sb = new StringBuilder();
while (st.hasMoreTokens()) {
final String token = st.nextToken();
if (!stopwords.contains(token)) {
sb.append(token);
sb.append(" ");
}
}
return sb.toString().trim();
}
public String filterAllStopWords(String s) {
s = filterStopWords(s, stopwords_en);
s = filterStopWords(s, stopwords_de);
s = filterStopWords(s, stopwords_it);
s = filterStopWords(s, stopwords_fr);
s = filterStopWords(s, stopwords_pt);
s = filterStopWords(s, stopwords_es);
s = filterStopWords(s, stopwords_gr);
return s;
}
protected Collection<String> filterBlacklisted(final Collection<String> set, final Set<String> ngramBlacklist) {
final Set<String> newset = Sets.newLinkedHashSet();
for (final String s : set) {
if (!ngramBlacklist.contains(s)) {
newset.add(s);
}
}
return newset;
}
public static Set<String> loadFromClasspath(final String classpath) {
Transliterator transliterator = Transliterator.getInstance("Any-Eng");
final Set<String> h = Sets.newHashSet();
try {
for (final String s : IOUtils.readLines(NGramUtils.class.getResourceAsStream(classpath), StandardCharsets.UTF_8)) {
h.add(fixAliases(transliterator.transliterate(s))); //transliteration of the stopwords
}
} catch (final Throwable e) {
return Sets.newHashSet();
}
return h;
}
public static Map<String, String> loadMapFromClasspath(final String classpath) {
Transliterator transliterator = Transliterator.getInstance("Any-Eng");
final Map<String, String> m = new HashMap<>();
try {
for (final String s : IOUtils.readLines(AbstractPaceFunctions.class.getResourceAsStream(classpath), StandardCharsets.UTF_8)) {
//string is like this: code;word1;word2;word3
String[] line = s.split(";");
String value = line[0];
for (int i = 1; i < line.length; i++) {
m.put(fixAliases(transliterator.transliterate(line[i].toLowerCase())), value);
}
}
} catch (final Throwable e) {
return new HashMap<>();
}
return m;
}
public String removeKeywords(String s, Set<String> keywords) {
s = " " + s + " ";
for (String k : keywords) {
s = s.replaceAll(k.toLowerCase(), "");
}
return s.trim();
}
public double commonElementsPercentage(Set<String> s1, Set<String> s2) {
double longer = Math.max(s1.size(), s2.size());
return (double) s1.stream().filter(s2::contains).count() / longer;
}
//convert the set of keywords to codes
public Set<String> toCodes(Set<String> keywords, Map<String, String> translationMap) {
return keywords.stream().map(s -> translationMap.get(s)).collect(Collectors.toSet());
}
public Set<String> keywordsToCodes(Set<String> keywords, Map<String, String> translationMap) {
return toCodes(keywords, translationMap);
}
public Set<String> citiesToCodes(Set<String> keywords) {
return toCodes(keywords, cityMap);
}
protected String firstLC(final String s) {
return StringUtils.substring(s, 0, 1).toLowerCase();
}
protected Iterable<String> tokens(final String s, final int maxTokens) {
return Iterables.limit(Splitter.on(" ").omitEmptyStrings().trimResults().split(s), maxTokens);
}
public String normalizePid(String pid) {
return pid.toLowerCase().replaceAll(DOI_PREFIX, "");
}
//get the list of keywords into the input string
public Set<String> getKeywords(String s1, Map<String, String> translationMap, int windowSize) {
String s = s1;
List<String> tokens = Arrays.asList(s.toLowerCase().split(" "));
Set<String> codes = new HashSet<>();
if (tokens.size() < windowSize)
windowSize = tokens.size();
int length = windowSize;
while (length != 0) {
for (int i = 0; i <= tokens.size() - length; i++) {
String candidate = concat(tokens.subList(i, i + length));
if (translationMap.containsKey(candidate)) {
codes.add(candidate);
s = s.replace(candidate, "").trim();
}
}
tokens = Arrays.asList(s.split(" "));
length -= 1;
}
return codes;
}
public Set<String> getCities(String s1, int windowSize) {
return getKeywords(s1, cityMap, windowSize);
}
public static <T> String readFromClasspath(final String filename, final Class<T> clazz) {
final StringWriter sw = new StringWriter();
try {
IOUtils.copy(clazz.getResourceAsStream(filename), sw, StandardCharsets.UTF_8);
return sw.toString();
} catch (final IOException e) {
throw new RuntimeException("cannot load resource from classpath: " + filename);
}
}
// city map to be used when translating the city names into codes
private static Map<String, String> cityMap = AbstractPaceFunctions
.loadMapFromClasspath("/eu/dnetlib/pace/config/city_map.csv");
// list of stopwords in different languages
protected static Set<String> stopwords_gr = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_gr.txt");
protected static Set<String> stopwords_en = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_en.txt");
protected static Set<String> stopwords_de = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_de.txt");
protected static Set<String> stopwords_es = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_es.txt");
protected static Set<String> stopwords_fr = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_fr.txt");
protected static Set<String> stopwords_it = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_it.txt");
protected static Set<String> stopwords_pt = loadFromClasspath("/eu/dnetlib/pace/config/stopwords_pt.txt");
// transliterator
protected static Transliterator transliterator = Transliterator.getInstance("Any-Eng");
// blacklist of ngrams: to avoid generic keys
protected static Set<String> ngramBlacklist = loadFromClasspath("/eu/dnetlib/pace/config/ngram_blacklist.txt");
// html regex for normalization
public final String HTML_REGEX = "<[^>]*>";
private static final String alpha = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789 ";
private static final String aliases_from = "⁰¹²³⁴⁵⁶⁷⁸⁹⁺⁻⁼⁽⁾ⁿ₀₁₂₃₄₅₆₇₈₉₊₋₌₍₎àáâäæãåāèéêëēėęəîïíīįìôöòóœøōõûüùúūßśšłžźżçćčñń";
private static final String aliases_to = "0123456789+-=()n0123456789+-=()aaaaaaaaeeeeeeeeiiiiiioooooooouuuuussslzzzcccnn";
// doi prefix for normalization
public final String DOI_PREFIX = "(https?:\\/\\/dx\\.doi\\.org\\/)|(doi:)";
private Pattern numberPattern = Pattern.compile("-?\\d+(\\.\\d+)?");
private Pattern hexUnicodePattern = Pattern.compile("\\\\u(\\p{XDigit}{4})");
protected final static FieldList EMPTY_FIELD = new FieldListImpl();
protected String concat(final List<String> l) {
return Joiner.on(" ").skipNulls().join(l);
}
protected String cleanup(final String s) {
final String s1 = s.replaceAll(HTML_REGEX, "");
final String s2 = unicodeNormalization(s1.toLowerCase());
final String s3 = nfd(s2);
final String s4 = fixXML(s3);
final String s5 = s4.replaceAll("([0-9]+)", " $1 ");
final String s6 = transliterate(s5);
final String s7 = fixAliases(s6);
final String s8 = s7.replaceAll("[^\\p{ASCII}]", "");
final String s9 = s8.replaceAll("[\\p{Punct}]", " ");
final String s10 = s9.replaceAll("\\n", " ");
final String s11 = s10.replaceAll("(?m)\\s+", " ");
final String s12 = s11.trim();
return s12;
}
protected String fixXML(final String a) {
return a
.replaceAll("&ndash;", " ")
.replaceAll("&amp;", " ")
.replaceAll("&quot;", " ")
.replaceAll("&minus;", " ");
}
protected boolean checkNumbers(final String a, final String b) {
final String numbersA = getNumbers(a);
final String numbersB = getNumbers(b);
final String romansA = getRomans(a);
final String romansB = getRomans(b);
return !numbersA.equals(numbersB) || !romansA.equals(romansB);
}
protected String getRomans(final String s) {
final StringBuilder sb = new StringBuilder();
for (final String t : s.split(" ")) {
sb.append(isRoman(t) ? t : "");
}
return sb.toString();
}
protected boolean isRoman(final String s) {
return s
.replaceAll("^M{0,4}(CM|CD|D?C{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})$", "qwertyuiop")
.equals("qwertyuiop");
}
protected String getNumbers(final String s) {
final StringBuilder sb = new StringBuilder();
for (final String t : s.split(" ")) {
sb.append(isNumber(t) ? t : "");
}
return sb.toString();
}
public boolean isNumber(String strNum) {
if (strNum == null) {
return false;
}
return numberPattern.matcher(strNum).matches();
}
protected static String fixAliases(final String s) {
final StringBuilder sb = new StringBuilder();
s.chars().forEach(ch -> {
final int i = StringUtils.indexOf(aliases_from, ch);
sb.append(i >= 0 ? aliases_to.charAt(i) : (char) ch);
});
return sb.toString();
}
protected static String transliterate(final String s) {
try {
return transliterator.transliterate(s);
} catch (Exception e) {
return s;
}
}
protected String removeSymbols(final String s) {
final StringBuilder sb = new StringBuilder();
s.chars().forEach(ch -> {
sb.append(StringUtils.contains(alpha, ch) ? (char) ch : ' ');
});
return sb.toString().replaceAll("\\s+", " ");
}
protected String getFirstValue(final Field values) {
return (values != null) && !Iterables.isEmpty(values) ? Iterables.getFirst(values, EMPTY_FIELD).stringValue()
: "";
}
protected boolean notNull(final String s) {
return s != null;
}
protected String normalize(final String s) {
return fixAliases(transliterate(nfd(unicodeNormalization(s))))
.toLowerCase()
// do not compact the regexes in a single expression, would cause StackOverflowError in case of large input
// strings
.replaceAll("[^ \\w]+", "")
.replaceAll("(\\p{InCombiningDiacriticalMarks})+", "")
.replaceAll("(\\p{Punct})+", " ")
.replaceAll("(\\d)+", " ")
.replaceAll("(\\n)+", " ")
.trim();
}
public String nfd(final String s) {
return Normalizer.normalize(s, Normalizer.Form.NFD);
}
public String utf8(final String s) {
byte[] bytes = s.getBytes(StandardCharsets.UTF_8);
return new String(bytes, StandardCharsets.UTF_8);
}
public String unicodeNormalization(final String s) {
Matcher m = hexUnicodePattern.matcher(s);
StringBuffer buf = new StringBuffer(s.length());
while (m.find()) {
String ch = String.valueOf((char) Integer.parseInt(m.group(1), 16));
m.appendReplacement(buf, Matcher.quoteReplacement(ch));
}
m.appendTail(buf);
return buf.toString();
}
protected String filterStopWords(final String s, final Set<String> stopwords) {
final StringTokenizer st = new StringTokenizer(s);
final StringBuilder sb = new StringBuilder();
while (st.hasMoreTokens()) {
final String token = st.nextToken();
if (!stopwords.contains(token)) {
sb.append(token);
sb.append(" ");
}
}
return sb.toString().trim();
}
public String filterAllStopWords(String s) {
s = filterStopWords(s, stopwords_en);
s = filterStopWords(s, stopwords_de);
s = filterStopWords(s, stopwords_it);
s = filterStopWords(s, stopwords_fr);
s = filterStopWords(s, stopwords_pt);
s = filterStopWords(s, stopwords_es);
s = filterStopWords(s, stopwords_gr);
return s;
}
protected Collection<String> filterBlacklisted(final Collection<String> set, final Set<String> ngramBlacklist) {
final Set<String> newset = Sets.newLinkedHashSet();
for (final String s : set) {
if (!ngramBlacklist.contains(s)) {
newset.add(s);
}
}
return newset;
}
public static Set<String> loadFromClasspath(final String classpath) {
Transliterator transliterator = Transliterator.getInstance("Any-Eng");
final Set<String> h = Sets.newHashSet();
try {
for (final String s : IOUtils
.readLines(NGramUtils.class.getResourceAsStream(classpath), StandardCharsets.UTF_8)) {
h.add(fixAliases(transliterator.transliterate(s))); // transliteration of the stopwords
}
} catch (final Throwable e) {
return Sets.newHashSet();
}
return h;
}
public static Map<String, String> loadMapFromClasspath(final String classpath) {
Transliterator transliterator = Transliterator.getInstance("Any-Eng");
final Map<String, String> m = new HashMap<>();
try {
for (final String s : IOUtils
.readLines(AbstractPaceFunctions.class.getResourceAsStream(classpath), StandardCharsets.UTF_8)) {
// string is like this: code;word1;word2;word3
String[] line = s.split(";");
String value = line[0];
for (int i = 1; i < line.length; i++) {
m.put(fixAliases(transliterator.transliterate(line[i].toLowerCase())), value);
}
}
} catch (final Throwable e) {
return new HashMap<>();
}
return m;
}
public String removeKeywords(String s, Set<String> keywords) {
s = " " + s + " ";
for (String k : keywords) {
s = s.replaceAll(k.toLowerCase(), "");
}
return s.trim();
}
public double commonElementsPercentage(Set<String> s1, Set<String> s2) {
double longer = Math.max(s1.size(), s2.size());
return (double) s1.stream().filter(s2::contains).count() / longer;
}
// convert the set of keywords to codes
public Set<String> toCodes(Set<String> keywords, Map<String, String> translationMap) {
return keywords.stream().map(s -> translationMap.get(s)).collect(Collectors.toSet());
}
public Set<String> keywordsToCodes(Set<String> keywords, Map<String, String> translationMap) {
return toCodes(keywords, translationMap);
}
public Set<String> citiesToCodes(Set<String> keywords) {
return toCodes(keywords, cityMap);
}
protected String firstLC(final String s) {
return StringUtils.substring(s, 0, 1).toLowerCase();
}
protected Iterable<String> tokens(final String s, final int maxTokens) {
return Iterables.limit(Splitter.on(" ").omitEmptyStrings().trimResults().split(s), maxTokens);
}
public String normalizePid(String pid) {
return pid.toLowerCase().replaceAll(DOI_PREFIX, "");
}
// get the list of keywords into the input string
public Set<String> getKeywords(String s1, Map<String, String> translationMap, int windowSize) {
String s = s1;
List<String> tokens = Arrays.asList(s.toLowerCase().split(" "));
Set<String> codes = new HashSet<>();
if (tokens.size() < windowSize)
windowSize = tokens.size();
int length = windowSize;
while (length != 0) {
for (int i = 0; i <= tokens.size() - length; i++) {
String candidate = concat(tokens.subList(i, i + length));
if (translationMap.containsKey(candidate)) {
codes.add(candidate);
s = s.replace(candidate, "").trim();
}
}
tokens = Arrays.asList(s.split(" "));
length -= 1;
}
return codes;
}
public Set<String> getCities(String s1, int windowSize) {
return getKeywords(s1, cityMap, windowSize);
}
public static <T> String readFromClasspath(final String filename, final Class<T> clazz) {
final StringWriter sw = new StringWriter();
try {
IOUtils.copy(clazz.getResourceAsStream(filename), sw, StandardCharsets.UTF_8);
return sw.toString();
} catch (final IOException e) {
throw new RuntimeException("cannot load resource from classpath: " + filename);
}
}
}

2
dhp-pace-core/src/main/java/eu/dnetlib/pace/config/Config.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.config;
import java.util.List;
@ -50,7 +51,6 @@ public interface Config {
*/
public Map<String, List<Pattern>> blacklists();
/**
* Translation map.
*

50
dhp-pace-core/src/main/java/eu/dnetlib/pace/config/DedupConfig.java
View File

@ -1,16 +1,5 @@
package eu.dnetlib.pace.config;
import com.fasterxml.jackson.annotation.JsonIgnore;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.common.collect.Maps;
import eu.dnetlib.pace.model.ClusteringDef;
import eu.dnetlib.pace.model.FieldDef;
import eu.dnetlib.pace.util.PaceException;
import org.antlr.stringtemplate.StringTemplate;
import org.apache.commons.io.IOUtils;
import org.apache.commons.lang3.StringUtils;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
package eu.dnetlib.pace.config;
import java.io.IOException;
import java.io.Serializable;
@ -22,9 +11,20 @@ import java.util.Map.Entry;
import java.util.regex.Pattern;
import java.util.stream.Collectors;
import org.antlr.stringtemplate.StringTemplate;
import org.apache.commons.io.IOUtils;
import org.apache.commons.lang3.StringUtils;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import com.fasterxml.jackson.annotation.JsonIgnore;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.common.collect.Maps;
import eu.dnetlib.pace.model.ClusteringDef;
import eu.dnetlib.pace.model.FieldDef;
import eu.dnetlib.pace.tree.support.TreeNodeDef;
import eu.dnetlib.pace.util.PaceException;
public class DedupConfig implements Config, Serializable {
@ -56,7 +56,8 @@ public class DedupConfig implements Config, Serializable {
defaults.put("idPath", "$.id");
}
public DedupConfig() {}
public DedupConfig() {
}
public static DedupConfig load(final String json) {
@ -66,10 +67,21 @@ public class DedupConfig implements Config, Serializable {
config.getPace().initModel();
config.getPace().initTranslationMap();
config.blacklists = config.getPace().getBlacklists().entrySet()
.stream()
.collect(Collectors.toMap(e -> e.getKey(),
e ->e.getValue().stream().filter(s -> !StringUtils.isBlank(s)).map(Pattern::compile).collect(Collectors.toList()) ));
config.blacklists = config
.getPace()
.getBlacklists()
.entrySet()
.stream()
.collect(
Collectors
.toMap(
e -> e.getKey(),
e -> e
.getValue()
.stream()
.filter(s -> !StringUtils.isBlank(s))
.map(Pattern::compile)
.collect(Collectors.toList())));
return config;
} catch (IOException e) {
@ -131,7 +143,7 @@ public class DedupConfig implements Config, Serializable {
}
@Override
public Map<String, TreeNodeDef> decisionTree(){
public Map<String, TreeNodeDef> decisionTree() {
return getPace().getDecisionTree();
}

23
dhp-pace-core/src/main/java/eu/dnetlib/pace/config/PaceConfig.java
View File

@ -1,19 +1,20 @@
package eu.dnetlib.pace.config;
import java.io.Serializable;
import java.util.List;
import java.util.Map;
import com.fasterxml.jackson.annotation.JsonIgnore;
import com.google.common.collect.Maps;
import com.ibm.icu.text.Transliterator;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.model.ClusteringDef;
import eu.dnetlib.pace.model.FieldDef;
import eu.dnetlib.pace.tree.support.TreeNodeDef;
import eu.dnetlib.pace.util.PaceResolver;
import java.io.Serializable;
import java.util.List;
import java.util.Map;
public class PaceConfig extends AbstractPaceFunctions implements Serializable {
private List<FieldDef> model;
@ -33,7 +34,8 @@ public class PaceConfig extends AbstractPaceFunctions implements Serializable {
@JsonIgnore
public static PaceResolver resolver = new PaceResolver();
public PaceConfig() {}
public PaceConfig() {
}
public void initModel() {
modelMap = Maps.newHashMap();
@ -42,20 +44,21 @@ public class PaceConfig extends AbstractPaceFunctions implements Serializable {
}
}
public void initTranslationMap(){
public void initTranslationMap() {
translationMap = Maps.newHashMap();
Transliterator transliterator = Transliterator.getInstance("Any-Eng");
for (String key : synonyms.keySet()) {
for (String term : synonyms.get(key)){
translationMap.put(
for (String term : synonyms.get(key)) {
translationMap
.put(
fixAliases(transliterator.transliterate(term.toLowerCase())),
key);
key);
}
}
}
public Map<String, String> translationMap(){
public Map<String, String> translationMap() {
return translationMap;
}

1
dhp-pace-core/src/main/java/eu/dnetlib/pace/config/Type.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.config;
public enum Type {

28
dhp-pace-core/src/main/java/eu/dnetlib/pace/config/WfConfig.java
View File

@ -1,10 +1,5 @@
package eu.dnetlib.pace.config;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.common.collect.Lists;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.util.PaceException;
import org.apache.commons.lang3.StringUtils;
package eu.dnetlib.pace.config;
import java.io.IOException;
import java.io.Serializable;
@ -12,6 +7,13 @@ import java.util.HashSet;
import java.util.List;
import java.util.Set;
import org.apache.commons.lang3.StringUtils;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.common.collect.Lists;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.util.PaceException;
public class WfConfig implements Serializable {
@ -76,7 +78,6 @@ public class WfConfig implements Serializable {
/** Maximum number of allowed children. */
private int maxChildren = MAX_CHILDREN;
/** Default maximum number of iterations. */
private final static int MAX_ITERATIONS = 20;
@ -84,9 +85,10 @@ public class WfConfig implements Serializable {
private int maxIterations = MAX_ITERATIONS;
/** The Jquery path to retrieve the identifier */
private String idPath = "$.id";
private String idPath = "$.id";
public WfConfig() {}
public WfConfig() {
}
/**
* Instantiates a new dedup config.
@ -114,8 +116,10 @@ public class WfConfig implements Serializable {
* @param idPath
* the path for the id of the entity
*/
public WfConfig(final String entityType, final String orderField, final List<String> rootBuilder, final String dedupRun,
final Set<String> skipList, final int queueMaxSize, final int groupMaxSize, final int slidingWindowSize, final boolean includeChildren, final int maxIterations, final String idPath) {
public WfConfig(final String entityType, final String orderField, final List<String> rootBuilder,
final String dedupRun,
final Set<String> skipList, final int queueMaxSize, final int groupMaxSize, final int slidingWindowSize,
final boolean includeChildren, final int maxIterations, final String idPath) {
super();
this.entityType = entityType;
this.orderField = orderField;
@ -257,7 +261,6 @@ public class WfConfig implements Serializable {
this.maxChildren = maxChildren;
}
public int getMaxIterations() {
return maxIterations;
}
@ -277,7 +280,6 @@ public class WfConfig implements Serializable {
/*
* (non-Javadoc)
*
* @see java.lang.Object#toString()
*/
@Override

8
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/AbstractField.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
import eu.dnetlib.pace.config.Type;
@ -16,7 +17,8 @@ public abstract class AbstractField implements Field {
/**
* Instantiates a new abstract field.
*/
protected AbstractField() {}
protected AbstractField() {
}
/**
* Instantiates a new abstract field.
@ -33,7 +35,6 @@ public abstract class AbstractField implements Field {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.Field#getName()
*/
@Override
@ -43,7 +44,6 @@ public abstract class AbstractField implements Field {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.Field#getType()
*/
@Override
@ -53,7 +53,6 @@ public abstract class AbstractField implements Field {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.Field#setName(java.lang.String)
*/
@Override
@ -63,7 +62,6 @@ public abstract class AbstractField implements Field {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.Field#setType(eu.dnetlib.pace.config.Type)
*/
@Override

14
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/ClusteringDef.java
View File

@ -1,15 +1,16 @@
package eu.dnetlib.pace.model;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.pace.clustering.ClusteringFunction;
import eu.dnetlib.pace.config.PaceConfig;
import eu.dnetlib.pace.util.PaceException;
package eu.dnetlib.pace.model;
import java.io.IOException;
import java.io.Serializable;
import java.util.List;
import java.util.Map;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.pace.clustering.ClusteringFunction;
import eu.dnetlib.pace.config.PaceConfig;
import eu.dnetlib.pace.util.PaceException;
public class ClusteringDef implements Serializable {
@ -19,7 +20,8 @@ public class ClusteringDef implements Serializable {
private Map<String, Integer> params;
public ClusteringDef() {}
public ClusteringDef() {
}
public String getName() {
return name;

1
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/Document.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
import java.util.Set;

5
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/Field.java
View File

@ -1,9 +1,10 @@
package eu.dnetlib.pace.model;
import eu.dnetlib.pace.config.Type;
import java.io.Serializable;
import eu.dnetlib.pace.config.Type;
/**
* The Interface Field.
*/

27
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/FieldDef.java
View File

@ -1,13 +1,15 @@
package eu.dnetlib.pace.model;
import java.io.Serializable;
import java.util.List;
import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.common.base.Splitter;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.config.Type;
import java.io.Serializable;
import java.util.List;
import eu.dnetlib.pace.config.Type;
/**
* The schema is composed by field definitions (FieldDef). Each field has a type, a name, and an associated compare algorithm.
@ -34,19 +36,20 @@ public class FieldDef implements Serializable {
*/
private int length = -1;
public FieldDef() {}
public FieldDef() {
}
// def apply(s: String): Field[A]
public Field apply(final Type type, final String s) {
switch (type) {
case Int:
return new FieldValueImpl(type, name, Integer.parseInt(s));
case String:
return new FieldValueImpl(type, name, s);
case List:
return new FieldListImpl(name, type);
default:
throw new IllegalArgumentException("Casting not implemented for type " + type);
case Int:
return new FieldValueImpl(type, name, Integer.parseInt(s));
case String:
return new FieldValueImpl(type, name, s);
case List:
return new FieldListImpl(name, type);
default:
throw new IllegalArgumentException("Casting not implemented for type " + type);
}
}

1
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/FieldList.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
import java.util.List;

69
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/FieldListImpl.java
View File

@ -1,17 +1,19 @@
package eu.dnetlib.pace.model;
import java.util.Collection;
import java.util.Iterator;
import java.util.List;
import java.util.ListIterator;
import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.common.base.Function;
import com.google.common.base.Joiner;
import com.google.common.collect.Iterables;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.config.Type;
import java.util.Collection;
import java.util.Iterator;
import java.util.List;
import java.util.ListIterator;
import eu.dnetlib.pace.config.Type;
/**
* The Class FieldListImpl.
@ -41,7 +43,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#add(java.lang.Object)
*/
@Override
@ -51,7 +52,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#add(int, java.lang.Object)
*/
@Override
@ -61,7 +61,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#addAll(java.util.Collection)
*/
@Override
@ -71,7 +70,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#addAll(int, java.util.Collection)
*/
@Override
@ -81,7 +79,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#clear()
*/
@Override
@ -91,7 +88,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#contains(java.lang.Object)
*/
@Override
@ -101,7 +97,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#containsAll(java.util.Collection)
*/
@Override
@ -111,7 +106,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#get(int)
*/
@Override
@ -121,7 +115,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#indexOf(java.lang.Object)
*/
@Override
@ -131,7 +124,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.Field#isEmpty()
*/
@Override
@ -141,7 +133,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.lang.Iterable#iterator()
*/
@Override
@ -151,7 +142,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#lastIndexOf(java.lang.Object)
*/
@Override
@ -161,7 +151,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#listIterator()
*/
@Override
@ -171,7 +160,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#listIterator(int)
*/
@Override
@ -181,7 +169,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#remove(java.lang.Object)
*/
@Override
@ -191,7 +178,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#remove(int)
*/
@Override
@ -201,7 +187,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#removeAll(java.util.Collection)
*/
@Override
@ -211,7 +196,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#retainAll(java.util.Collection)
*/
@Override
@ -221,7 +205,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#set(int, java.lang.Object)
*/
@Override
@ -231,7 +214,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#size()
*/
@Override
@ -241,7 +223,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#subList(int, int)
*/
@Override
@ -251,7 +232,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#toArray()
*/
@Override
@ -261,7 +241,6 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see java.util.List#toArray(java.lang.Object[])
*/
@Override
@ -271,33 +250,31 @@ public class FieldListImpl extends AbstractField implements FieldList {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.Field#stringValue()
*/
@Override
public String stringValue() {
switch (getType()) {
case List:
case Int:
case String:
return Joiner.on(" ").join(stringList());
case JSON:
String json;
try {
json = new ObjectMapper().writeValueAsString(this);
} catch (JsonProcessingException e) {
json = null;
}
return json;
default:
throw new IllegalArgumentException("Unknown type: " + getType().toString());
case List:
case Int:
case String:
return Joiner.on(" ").join(stringList());
case JSON:
String json;
try {
json = new ObjectMapper().writeValueAsString(this);
} catch (JsonProcessingException e) {
json = null;
}
return json;
default:
throw new IllegalArgumentException("Unknown type: " + getType().toString());
}
}
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.FieldList#stringList()
*/
@Override
@ -317,10 +294,10 @@ public class FieldListImpl extends AbstractField implements FieldList {
@Override
public double[] doubleArray() {
return Lists.newArrayList(Iterables.transform(fields, getDouble())).stream().mapToDouble(d-> d).toArray();
return Lists.newArrayList(Iterables.transform(fields, getDouble())).stream().mapToDouble(d -> d).toArray();
}
private Function<Field,Double> getDouble() {
private Function<Field, Double> getDouble() {
return new Function<Field, Double>() {
@Override

1
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/FieldValue.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
/**

25
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/FieldValueImpl.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
import java.net.MalformedURLException;
@ -6,9 +7,10 @@ import java.util.Collections;
import java.util.Iterator;
import java.util.List;
import eu.dnetlib.pace.config.Type;
import org.apache.commons.lang3.StringUtils;
import eu.dnetlib.pace.config.Type;
/**
* The Class FieldValueImpl.
*/
@ -20,7 +22,8 @@ public class FieldValueImpl extends AbstractField implements FieldValue {
/**
* Instantiates a new field value impl.
*/
public FieldValueImpl() {}
public FieldValueImpl() {
}
/**
* Instantiates a new field value impl.
@ -39,17 +42,17 @@ public class FieldValueImpl extends AbstractField implements FieldValue {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.Field#isEmpty()
*/
@Override
public boolean isEmpty() {
if (value == null) return false;
if (value == null)
return false;
switch (type) {
case String:
case JSON:
return value.toString().isEmpty();
return value.toString().isEmpty();
case List:
try {
List<?> list = (List<?>) value;
@ -61,9 +64,9 @@ public class FieldValueImpl extends AbstractField implements FieldValue {
String str = value.toString();
return StringUtils.isBlank(str) || !isValidURL(str);
case DoubleArray:
return doubleArrayValue().length==0;
default:
return true;
return doubleArrayValue().length == 0;
default:
return true;
}
}
@ -78,7 +81,6 @@ public class FieldValueImpl extends AbstractField implements FieldValue {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.FieldValue#getValue()
*/
@Override
@ -88,7 +90,6 @@ public class FieldValueImpl extends AbstractField implements FieldValue {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.FieldValue#setValue(java.lang.Object)
*/
@Override
@ -98,7 +99,6 @@ public class FieldValueImpl extends AbstractField implements FieldValue {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.Field#stringValue()
*/
@Override
@ -119,12 +119,11 @@ public class FieldValueImpl extends AbstractField implements FieldValue {
}
public double[] doubleArrayValue() {
return (double[])getValue();
return (double[]) getValue();
}
/*
* (non-Javadoc)
*
* @see java.lang.Iterable#iterator()
*/
@Override

6
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/MapDocument.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
import java.io.Serializable;
@ -57,7 +58,6 @@ public class MapDocument implements Document, Serializable {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.document.Document#fields()
*/
@Override
@ -67,7 +67,6 @@ public class MapDocument implements Document, Serializable {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.document.Document#values(java.lang.String)
*/
@Override
@ -77,7 +76,6 @@ public class MapDocument implements Document, Serializable {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.document.Document#fieldNames()
*/
@Override
@ -87,7 +85,6 @@ public class MapDocument implements Document, Serializable {
/*
* (non-Javadoc)
*
* @see java.lang.Object#toString()
*/
@Override
@ -107,7 +104,6 @@ public class MapDocument implements Document, Serializable {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.model.document.Document#getIdentifier()
*/
@Override

8
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/MapDocumentComparator.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
import java.util.Comparator;
@ -28,18 +29,19 @@ public class MapDocumentComparator implements Comparator<Document> {
/*
* (non-Javadoc)
*
* @see java.util.Comparator#compare(java.lang.Object, java.lang.Object)
*/
@Override
public int compare(final Document d1, final Document d2) {
if (d1.values(comparatorField).isEmpty() || d2.values(comparatorField).isEmpty()) return 0;
if (d1.values(comparatorField).isEmpty() || d2.values(comparatorField).isEmpty())
return 0;
final String o1 = Iterables.getFirst(d1.values(comparatorField), emptyField).stringValue();
final String o2 = Iterables.getFirst(d2.values(comparatorField), emptyField).stringValue();
if ((o1 == null) || (o2 == null)) return 0;
if ((o1 == null) || (o2 == null))
return 0;
final String to1 = NGramUtils.cleanupForOrdering(o1);
final String to2 = NGramUtils.cleanupForOrdering(o2);

4
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/MapDocumentSerializer.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
import java.lang.reflect.Type;
@ -33,7 +34,8 @@ public class MapDocumentSerializer implements InstanceCreator<MapDocument> {
gson.registerTypeAdapter(Field.class, new JsonDeserializer<Field>() {
@Override
public Field deserialize(final JsonElement json, final Type typeOfT, final JsonDeserializationContext context) throws JsonParseException {
public Field deserialize(final JsonElement json, final Type typeOfT,
final JsonDeserializationContext context) throws JsonParseException {
final FieldListImpl fl = new FieldListImpl();
if (json.isJsonObject()) {

3
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/Person.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
import java.nio.charset.Charset;
@ -43,7 +44,7 @@ public class Person {
// s = s.replaceAll("[\\W&&[^,-]]", "");
}
if (s.contains(",")) { //if the name contains a comma it is easy derivable the name and the surname
if (s.contains(",")) { // if the name contains a comma it is easy derivable the name and the surname
final String[] arr = s.split(",");
if (arr.length == 1) {
fullname = splitTerms(arr[0]);

5
dhp-pace-core/src/main/java/eu/dnetlib/pace/model/PersonComparatorUtils.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.model;
import java.util.ArrayList;
@ -57,7 +58,7 @@ public class PersonComparatorUtils {
private static boolean verifyNames(List<String> list1, List<String> list2) {
return verifySimilarity(extractExtendedNames(list1), extractExtendedNames(list2))
&& verifySimilarity(extractInitials(list1), extractInitials(list2));
&& verifySimilarity(extractInitials(list1), extractInitials(list2));
}
private static boolean verifySurnames(List<String> list1, List<String> list2) {
@ -76,7 +77,7 @@ public class PersonComparatorUtils {
Collections.sort(list1);
Collections.sort(list2);
return verifySimilarity(extractExtendedNames(list1), extractExtendedNames(list2))
&& verifySimilarity(extractInitials(list1), extractInitials(list2));
&& verifySimilarity(extractInitials(list1), extractInitials(list2));
}
private static List<String> extractExtendedNames(List<String> list) {

49
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/AlwaysMatch.java
View File

@ -1,42 +1,43 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
@ComparatorClass("alwaysMatch")
public class AlwaysMatch extends AbstractComparator {
public AlwaysMatch(final Map<String, String> params){
super(params, new com.wcohen.ss.JaroWinkler());
}
public AlwaysMatch(final Map<String, String> params) {
super(params, new com.wcohen.ss.JaroWinkler());
}
public AlwaysMatch(final double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}
public AlwaysMatch(final double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}
protected AlwaysMatch(final double weight, final AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
protected AlwaysMatch(final double weight, final AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
return 1.0;
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
return 1.0;
}
@Override
public double getWeight() {
return super.weight;
}
@Override
public double getWeight() {
return super.weight;
}
@Override
protected double normalize(final double d) {
return d;
}
@Override
protected double normalize(final double d) {
return d;
}
}

245
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/AuthorsMatch.java
View File

@ -1,13 +1,5 @@
package eu.dnetlib.pace.tree;
import com.google.common.collect.Iterables;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.model.Person;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import com.wcohen.ss.AbstractStringDistance;
package eu.dnetlib.pace.tree;
import java.util.Comparator;
import java.util.List;
@ -16,139 +8,164 @@ import java.util.function.Function;
import java.util.stream.Collectors;
import java.util.stream.Stream;
import com.google.common.collect.Iterables;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.model.Person;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("authorsMatch")
public class AuthorsMatch extends AbstractComparator {
Map<String, String> params;
Map<String, String> params;
private double SURNAME_THRESHOLD;
private double NAME_THRESHOLD;
private double FULLNAME_THRESHOLD;
private String MODE; //full or surname
private int SIZE_THRESHOLD;
private String TYPE; //count or percentage
private int common;
private double SURNAME_THRESHOLD;
private double NAME_THRESHOLD;
private double FULLNAME_THRESHOLD;
private String MODE; // full or surname
private int SIZE_THRESHOLD;
private String TYPE; // count or percentage
private int common;
public AuthorsMatch(Map<String, String> params){
super(params, new com.wcohen.ss.JaroWinkler());
this.params = params;
public AuthorsMatch(Map<String, String> params) {
super(params, new com.wcohen.ss.JaroWinkler());
this.params = params;
MODE = params.getOrDefault("mode", "full");
SURNAME_THRESHOLD = Double.parseDouble(params.getOrDefault("surname_th", "0.95"));
NAME_THRESHOLD = Double.parseDouble(params.getOrDefault("name_th", "0.95"));
FULLNAME_THRESHOLD = Double.parseDouble(params.getOrDefault("fullname_th", "0.9"));
SIZE_THRESHOLD = Integer.parseInt(params.getOrDefault("size_th", "20"));
TYPE = params.getOrDefault("type", "percentage");
common = 0;
}
MODE = params.getOrDefault("mode", "full");
SURNAME_THRESHOLD = Double.parseDouble(params.getOrDefault("surname_th", "0.95"));
NAME_THRESHOLD = Double.parseDouble(params.getOrDefault("name_th", "0.95"));
FULLNAME_THRESHOLD = Double.parseDouble(params.getOrDefault("fullname_th", "0.9"));
SIZE_THRESHOLD = Integer.parseInt(params.getOrDefault("size_th", "20"));
TYPE = params.getOrDefault("type", "percentage");
common = 0;
}
protected AuthorsMatch(double w, AbstractStringDistance ssalgo) {
super(w, ssalgo);
}
protected AuthorsMatch(double w, AbstractStringDistance ssalgo) {
super(w, ssalgo);
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
@Override
public double compare(final Field a, final Field b, final Config conf) {
if (a.isEmpty() || b.isEmpty())
return -1;
if (a.isEmpty() || b.isEmpty())
return -1;
if (((FieldList) a).size() > SIZE_THRESHOLD || ((FieldList) b).size() > SIZE_THRESHOLD)
return 1.0;
if (((FieldList) a).size() > SIZE_THRESHOLD || ((FieldList) b).size() > SIZE_THRESHOLD)
return 1.0;
List<Person> aList = ((FieldList) a).stringList().stream().map(author -> new Person(author, false)).collect(Collectors.toList());
List<Person> bList = ((FieldList) b).stringList().stream().map(author -> new Person(author, false)).collect(Collectors.toList());
List<Person> aList = ((FieldList) a)
.stringList()
.stream()
.map(author -> new Person(author, false))
.collect(Collectors.toList());
List<Person> bList = ((FieldList) b)
.stringList()
.stream()
.map(author -> new Person(author, false))
.collect(Collectors.toList());
common = 0;
//compare each element of List1 with each element of List2
for (Person p1 : aList)
common = 0;
// compare each element of List1 with each element of List2
for (Person p1 : aList)
for (Person p2 : bList) {
for (Person p2 : bList) {
//both persons are inaccurate
if (!p1.isAccurate() && !p2.isAccurate()) {
//compare just normalized fullnames
String fullname1 = normalization(p1.getNormalisedFullname().isEmpty()? p1.getOriginal() : p1.getNormalisedFullname());
String fullname2 = normalization(p2.getNormalisedFullname().isEmpty()? p2.getOriginal() : p2.getNormalisedFullname());
// both persons are inaccurate
if (!p1.isAccurate() && !p2.isAccurate()) {
// compare just normalized fullnames
String fullname1 = normalization(
p1.getNormalisedFullname().isEmpty() ? p1.getOriginal() : p1.getNormalisedFullname());
String fullname2 = normalization(
p2.getNormalisedFullname().isEmpty() ? p2.getOriginal() : p2.getNormalisedFullname());
if (ssalgo.score(fullname1, fullname2) > FULLNAME_THRESHOLD) {
common += 1;
break;
}
}
if (ssalgo.score(fullname1, fullname2) > FULLNAME_THRESHOLD) {
common += 1;
break;
}
}
//one person is inaccurate
if (p1.isAccurate() ^ p2.isAccurate()) {
//prepare data
//data for the accurate person
String name = normalization(p1.isAccurate()? p1.getNormalisedFirstName() : p2.getNormalisedFirstName());
String surname = normalization(p1.isAccurate()? p1.getNormalisedSurname() : p2.getNormalisedSurname());
// one person is inaccurate
if (p1.isAccurate() ^ p2.isAccurate()) {
// prepare data
// data for the accurate person
String name = normalization(
p1.isAccurate() ? p1.getNormalisedFirstName() : p2.getNormalisedFirstName());
String surname = normalization(
p1.isAccurate() ? p1.getNormalisedSurname() : p2.getNormalisedSurname());
//data for the inaccurate person
String fullname = normalization(
p1.isAccurate() ? ((p2.getNormalisedFullname().isEmpty()) ? p2.getOriginal() : p2.getNormalisedFullname()) : (p1.getNormalisedFullname().isEmpty() ? p1.getOriginal() : p1.getNormalisedFullname())
);
// data for the inaccurate person
String fullname = normalization(
p1.isAccurate()
? ((p2.getNormalisedFullname().isEmpty()) ? p2.getOriginal() : p2.getNormalisedFullname())
: (p1.getNormalisedFullname().isEmpty() ? p1.getOriginal() : p1.getNormalisedFullname()));
if (fullname.contains(surname)) {
if (MODE.equals("full")) {
if (fullname.contains(name)) {
common += 1;
break;
}
}
else { //MODE equals "surname"
common += 1;
break;
}
}
}
if (fullname.contains(surname)) {
if (MODE.equals("full")) {
if (fullname.contains(name)) {
common += 1;
break;
}
} else { // MODE equals "surname"
common += 1;
break;
}
}
}
//both persons are accurate
if (p1.isAccurate() && p2.isAccurate()) {
// both persons are accurate
if (p1.isAccurate() && p2.isAccurate()) {
if (compareSurname(p1, p2)) {
if (MODE.equals("full")) {
if(compareFirstname(p1, p2)) {
common += 1;
break;
}
}
else { //MODE equals "surname"
common += 1;
break;
}
}
if (compareSurname(p1, p2)) {
if (MODE.equals("full")) {
if (compareFirstname(p1, p2)) {
common += 1;
break;
}
} else { // MODE equals "surname"
common += 1;
break;
}
}
}
}
}
}
//normalization factor to compute the score
int normFactor = aList.size() == bList.size() ? aList.size() : (aList.size() + bList.size() - common);
// normalization factor to compute the score
int normFactor = aList.size() == bList.size() ? aList.size() : (aList.size() + bList.size() - common);
if(TYPE.equals("percentage")) {
return (double) common / normFactor;
}
else {
return (double) common;
}
}
if (TYPE.equals("percentage")) {
return (double) common / normFactor;
} else {
return (double) common;
}
}
public boolean compareSurname(Person p1, Person p2) {
return ssalgo.score(normalization(p1.getNormalisedSurname()), normalization(p2.getNormalisedSurname())) > SURNAME_THRESHOLD;
}
public boolean compareSurname(Person p1, Person p2) {
return ssalgo
.score(
normalization(p1.getNormalisedSurname()), normalization(p2.getNormalisedSurname())) > SURNAME_THRESHOLD;
}
public boolean compareFirstname(Person p1, Person p2) {
public boolean compareFirstname(Person p1, Person p2) {
if(p1.getNormalisedFirstName().length()<=2 || p2.getNormalisedFirstName().length()<=2) {
if (firstLC(p1.getNormalisedFirstName()).equals(firstLC(p2.getNormalisedFirstName())))
return true;
}
if (p1.getNormalisedFirstName().length() <= 2 || p2.getNormalisedFirstName().length() <= 2) {
if (firstLC(p1.getNormalisedFirstName()).equals(firstLC(p2.getNormalisedFirstName())))
return true;
}
return ssalgo.score(normalization(p1.getNormalisedFirstName()), normalization(p2.getNormalisedFirstName())) > NAME_THRESHOLD;
}
return ssalgo
.score(
normalization(p1.getNormalisedFirstName()),
normalization(p2.getNormalisedFirstName())) > NAME_THRESHOLD;
}
public String normalization(String s) {
return normalize(utf8(cleanup(s)));
}
public String normalization(String s) {
return normalize(utf8(cleanup(s)));
}
}

59
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/CityMatch.java
View File

@ -1,47 +1,48 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import java.util.Set;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
import java.util.Set;
@ComparatorClass("cityMatch")
public class CityMatch extends AbstractComparator {
private Map<String, String> params;
private Map<String, String> params;
public CityMatch(Map<String, String> params) {
super(params);
this.params = params;
}
public CityMatch(Map<String, String> params) {
super(params);
this.params = params;
}
@Override
public double distance(final String a, final String b, final Config conf) {
@Override
public double distance(final String a, final String b, final Config conf) {
String ca = cleanup(a);
String cb = cleanup(b);
String ca = cleanup(a);
String cb = cleanup(b);
ca = normalize(ca);
cb = normalize(cb);
ca = normalize(ca);
cb = normalize(cb);
ca = filterAllStopWords(ca);
cb = filterAllStopWords(cb);
ca = filterAllStopWords(ca);
cb = filterAllStopWords(cb);
Set<String> cities1 = getCities(ca, Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> cities2 = getCities(cb, Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> cities1 = getCities(ca, Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> cities2 = getCities(cb, Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> codes1 = citiesToCodes(cities1);
Set<String> codes2 = citiesToCodes(cities2);
Set<String> codes1 = citiesToCodes(cities1);
Set<String> codes2 = citiesToCodes(cities2);
//if no cities are detected, the comparator gives 1.0
if (codes1.isEmpty() && codes2.isEmpty())
return 1.0;
else {
if (codes1.isEmpty() ^ codes2.isEmpty())
return -1; //undefined if one of the two has no cities
return commonElementsPercentage(codes1, codes2);
}
}
// if no cities are detected, the comparator gives 1.0
if (codes1.isEmpty() && codes2.isEmpty())
return 1.0;
else {
if (codes1.isEmpty() ^ codes2.isEmpty())
return -1; // undefined if one of the two has no cities
return commonElementsPercentage(codes1, codes2);
}
}
}

60
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/CosineSimilarity.java
View File

@ -1,5 +1,11 @@
package eu.dnetlib.pace.tree;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
@ -8,46 +14,40 @@ import eu.dnetlib.pace.model.Person;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;
@ComparatorClass("cosineSimilarity")
public class CosineSimilarity extends AbstractComparator {
Map<String, String> params;
Map<String, String> params;
public CosineSimilarity(Map<String,String> params) {
super(params);
}
public CosineSimilarity(Map<String, String> params) {
super(params);
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
@Override
public double compare(final Field a, final Field b, final Config conf) {
if (a.isEmpty() || b.isEmpty())
return -1;
if (a.isEmpty() || b.isEmpty())
return -1;
double[] aVector = ((FieldValueImpl) a).doubleArrayValue();
double[] bVector = ((FieldValueImpl) b).doubleArrayValue();
double[] aVector = ((FieldValueImpl) a).doubleArrayValue();
double[] bVector = ((FieldValueImpl) b).doubleArrayValue();
return cosineSimilarity(aVector, bVector);
}
return cosineSimilarity(aVector, bVector);
}
double cosineSimilarity(double[] a, double[] b) {
double dotProduct = 0;
double normASum = 0;
double normBSum = 0;
double cosineSimilarity(double[] a, double[] b) {
double dotProduct = 0;
double normASum = 0;
double normBSum = 0;
for(int i = 0; i < a.length; i ++) {
dotProduct += a[i] * b[i];
normASum += a[i] * a[i];
normBSum += b[i] * b[i];
}
double eucledianDist = Math.sqrt(normASum) * Math.sqrt(normBSum);
return dotProduct / eucledianDist;
}
for (int i = 0; i < a.length; i++) {
dotProduct += a[i] * b[i];
normASum += a[i] * a[i];
normBSum += b[i] * b[i];
}
double eucledianDist = Math.sqrt(normASum) * Math.sqrt(normBSum);
return dotProduct / eucledianDist;
}
}

17
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/DoiExactMatch.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
@ -13,15 +14,15 @@ import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("doiExactMatch")
public class DoiExactMatch extends ExactMatchIgnoreCase {
public final String PREFIX = "(http:\\/\\/dx\\.doi\\.org\\/)|(doi:)";
public final String PREFIX = "(http:\\/\\/dx\\.doi\\.org\\/)|(doi:)";
public DoiExactMatch(final Map<String, String> params) {
super(params);
}
public DoiExactMatch(final Map<String, String> params) {
super(params);
}
@Override
protected String getValue(final Field f) {
return super.getValue(f).replaceAll(PREFIX, "");
}
@Override
protected String getValue(final Field f) {
return super.getValue(f).replaceAll(PREFIX, "");
}
}

35
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/DomainExactMatch.java
View File

@ -1,30 +1,31 @@
package eu.dnetlib.pace.tree;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.tree.support.ComparatorClass;
package eu.dnetlib.pace.tree;
import java.net.MalformedURLException;
import java.net.URL;
import java.util.Map;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("domainExactMatch")
public class DomainExactMatch extends ExactMatchIgnoreCase {
public DomainExactMatch(final Map<String, String> params) {
super(params);
}
public DomainExactMatch(final Map<String, String> params) {
super(params);
}
@Override
protected String getValue(final Field f) {
@Override
protected String getValue(final Field f) {
try {
return asUrl(super.getValue(f)).getHost();
} catch (MalformedURLException e) {
return "";
}
}
try {
return asUrl(super.getValue(f)).getHost();
} catch (MalformedURLException e) {
return "";
}
}
private URL asUrl(final String value) throws MalformedURLException {
return new URL(value);
}
private URL asUrl(final String value) throws MalformedURLException {
return new URL(value);
}
}

54
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/ExactMatch.java
View File

@ -1,42 +1,44 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
@ComparatorClass("exactMatch")
public class ExactMatch extends AbstractComparator {
public ExactMatch(Map<String, String> params){
super(params, new com.wcohen.ss.JaroWinkler());
}
public ExactMatch(Map<String, String> params) {
super(params, new com.wcohen.ss.JaroWinkler());
}
public ExactMatch(final double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}
public ExactMatch(final double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}
protected ExactMatch(final double weight, final AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
protected ExactMatch(final double weight, final AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
@Override
public double distance(final String a, final String b, final Config conf) {
if (a.isEmpty() || b.isEmpty()) {
return -1.0; //return -1 if a field is missing
}
return a.equals(b) ? 1.0 : 0;
}
@Override
public double distance(final String a, final String b, final Config conf) {
if (a.isEmpty() || b.isEmpty()) {
return -1.0; // return -1 if a field is missing
}
return a.equals(b) ? 1.0 : 0;
}
@Override
public double getWeight() {
return super.weight;
}
@Override
public double getWeight() {
return super.weight;
}
@Override
protected double normalize(final double d) {
return d;
}
@Override
protected double normalize(final double d) {
return d;
}
}

33
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/ExactMatchIgnoreCase.java
View File

@ -1,32 +1,33 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
@ComparatorClass("exactMatchIgnoreCase")
public class ExactMatchIgnoreCase extends AbstractComparator {
public ExactMatchIgnoreCase(Map<String, String> params) {
super(params);
}
public ExactMatchIgnoreCase(Map<String, String> params) {
super(params);
}
@Override
public double compare(Field a, Field b, final Config conf) {
@Override
public double compare(Field a, Field b, final Config conf) {
final String fa = getValue(a);
final String fb = getValue(b);
final String fa = getValue(a);
final String fb = getValue(b);
if (fa.isEmpty() || fb.isEmpty())
return -1;
if (fa.isEmpty() || fb.isEmpty())
return -1;
return fa.equalsIgnoreCase(fb) ? 1 : 0;
}
return fa.equalsIgnoreCase(fb) ? 1 : 0;
}
protected String getValue(final Field f) {
return getFirstValue(f);
}
protected String getValue(final Field f) {
return getFirstValue(f);
}
}

113
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/InstanceTypeMatch.java
View File

@ -1,11 +1,5 @@
package eu.dnetlib.pace.tree;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
package eu.dnetlib.pace.tree;
import java.util.HashMap;
import java.util.List;
@ -13,72 +7,79 @@ import java.util.Map;
import java.util.Set;
import java.util.stream.Collectors;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("instanceTypeMatch")
public class InstanceTypeMatch extends AbstractComparator {
final Map<String, String> translationMap = new HashMap<>();
final Map<String, String> translationMap = new HashMap<>();
public InstanceTypeMatch(Map<String, String> params){
super(params);
public InstanceTypeMatch(Map<String, String> params) {
super(params);
//jolly types
translationMap.put("Conference object", "*");
translationMap.put("Other literature type", "*");
translationMap.put("Unknown", "*");
// jolly types
translationMap.put("Conference object", "*");
translationMap.put("Other literature type", "*");
translationMap.put("Unknown", "*");
//article types
translationMap.put("Article", "Article");
translationMap.put("Data Paper", "Article");
translationMap.put("Software Paper", "Article");
translationMap.put("Preprint", "Article");
// article types
translationMap.put("Article", "Article");
translationMap.put("Data Paper", "Article");
translationMap.put("Software Paper", "Article");
translationMap.put("Preprint", "Article");
//thesis types
translationMap.put("Thesis", "Thesis");
translationMap.put("Master thesis", "Thesis");
translationMap.put("Bachelor thesis", "Thesis");
translationMap.put("Doctoral thesis", "Thesis");
}
// thesis types
translationMap.put("Thesis", "Thesis");
translationMap.put("Master thesis", "Thesis");
translationMap.put("Bachelor thesis", "Thesis");
translationMap.put("Doctoral thesis", "Thesis");
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
@Override
public double compare(final Field a, final Field b, final Config conf) {
if (a == null || b == null) {
return -1;
}
if (a == null || b == null) {
return -1;
}
final List<String> sa = ((FieldList) a).stringList();
final List<String> sb = ((FieldList) b).stringList();
final List<String> sa = ((FieldList) a).stringList();
final List<String> sb = ((FieldList) b).stringList();
if (sa.isEmpty() || sb.isEmpty()) {
return -1;
}
if (sa.isEmpty() || sb.isEmpty()) {
return -1;
}
final Set<String> ca = sa.stream().map(this::translate).collect(Collectors.toSet());
final Set<String> cb = sb.stream().map(this::translate).collect(Collectors.toSet());
final Set<String> ca = sa.stream().map(this::translate).collect(Collectors.toSet());
final Set<String> cb = sb.stream().map(this::translate).collect(Collectors.toSet());
// if at least one is a jolly type, it must produce a match
if (ca.contains("*") || cb.contains("*"))
return 1.0;
//if at least one is a jolly type, it must produce a match
if (ca.contains("*") || cb.contains("*"))
return 1.0;
int incommon = Sets.intersection(ca, cb).size();
int incommon = Sets.intersection(ca, cb).size();
// if at least one is in common, it must produce a match
return incommon >= 1 ? 1 : 0;
}
//if at least one is in common, it must produce a match
return incommon >= 1 ? 1 : 0;
}
public String translate(String term) {
return translationMap.getOrDefault(term, term);
}
public String translate(String term){
return translationMap.getOrDefault(term, term);
}
@Override
public double getWeight() {
return super.weight;
}
@Override
public double getWeight() {
return super.weight;
}
@Override
protected double normalize(final double d) {
return d;
}
@Override
protected double normalize(final double d) {
return d;
}
}

52
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/JaroWinkler.java
View File

@ -1,44 +1,46 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
//case class JaroWinkler(w: Double) extends SecondStringDistanceAlgo(w, new com.wcohen.ss.JaroWinkler())
@ComparatorClass("jaroWinkler")
public class JaroWinkler extends AbstractComparator {
public JaroWinkler(Map<String, String> params){
super(params, new com.wcohen.ss.JaroWinkler());
}
public JaroWinkler(Map<String, String> params) {
super(params, new com.wcohen.ss.JaroWinkler());
}
public JaroWinkler(double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}
public JaroWinkler(double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}
protected JaroWinkler(double weight, AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
protected JaroWinkler(double weight, AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
@Override
public double distance(String a, String b, final Config conf) {
String ca = cleanup(a);
String cb = cleanup(b);
@Override
public double distance(String a, String b, final Config conf) {
String ca = cleanup(a);
String cb = cleanup(b);
return normalize(ssalgo.score(ca, cb));
}
return normalize(ssalgo.score(ca, cb));
}
@Override
public double getWeight() {
return super.weight;
}
@Override
public double getWeight() {
return super.weight;
}
@Override
protected double normalize(double d) {
return d;
}
@Override
protected double normalize(double d) {
return d;
}
}

100
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/JaroWinklerNormalizedName.java
View File

@ -1,72 +1,74 @@
package eu.dnetlib.pace.tree;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import eu.dnetlib.pace.config.Config;
import java.util.Map;
import java.util.Set;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("jaroWinklerNormalizedName")
public class JaroWinklerNormalizedName extends AbstractComparator {
private Map<String, String> params;
private Map<String, String> params;
public JaroWinklerNormalizedName(Map<String, String> params){
super(params, new com.wcohen.ss.JaroWinkler());
this.params = params;
}
public JaroWinklerNormalizedName(Map<String, String> params) {
super(params, new com.wcohen.ss.JaroWinkler());
this.params = params;
}
public JaroWinklerNormalizedName(double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}
public JaroWinklerNormalizedName(double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}
protected JaroWinklerNormalizedName(double weight, AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
protected JaroWinklerNormalizedName(double weight, AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
@Override
public double distance(String a, String b, final Config conf) {
String ca = cleanup(a);
String cb = cleanup(b);
@Override
public double distance(String a, String b, final Config conf) {
String ca = cleanup(a);
String cb = cleanup(b);
ca = normalize(ca);
cb = normalize(cb);
ca = normalize(ca);
cb = normalize(cb);
ca = filterAllStopWords(ca);
cb = filterAllStopWords(cb);
ca = filterAllStopWords(ca);
cb = filterAllStopWords(cb);
Set<String> keywords1 = getKeywords(ca, conf.translationMap(), Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> keywords2 = getKeywords(cb, conf.translationMap(), Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> keywords1 = getKeywords(
ca, conf.translationMap(), Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> keywords2 = getKeywords(
cb, conf.translationMap(), Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> cities1 = getCities(ca, Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> cities2 = getCities(cb, Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> cities1 = getCities(ca, Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> cities2 = getCities(cb, Integer.parseInt(params.getOrDefault("windowSize", "4")));
ca = removeKeywords(ca, keywords1);
ca = removeKeywords(ca, cities1);
cb = removeKeywords(cb, keywords2);
cb = removeKeywords(cb, cities2);
ca = removeKeywords(ca, keywords1);
ca = removeKeywords(ca, cities1);
cb = removeKeywords(cb, keywords2);
cb = removeKeywords(cb, cities2);
ca = ca.replaceAll("[ ]{2,}", " ");
cb = cb.replaceAll("[ ]{2,}", " ");
ca = ca.replaceAll("[ ]{2,}", " ");
cb = cb.replaceAll("[ ]{2,}", " ");
if (ca.isEmpty() && cb.isEmpty())
return 1.0;
else
return normalize(ssalgo.score(ca,cb));
}
if (ca.isEmpty() && cb.isEmpty())
return 1.0;
else
return normalize(ssalgo.score(ca, cb));
}
@Override
public double getWeight() {
return super.weight;
}
@Override
public double getWeight() {
return super.weight;
}
@Override
protected double normalize(double d) {
return d;
}
@Override
protected double normalize(double d) {
return d;
}
}

11
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/JaroWinklerTitle.java
View File

@ -1,18 +1,19 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import eu.dnetlib.pace.config.Config;
import java.util.Map;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
//case class JaroWinkler(w: Double) extends SecondStringDistanceAlgo(w, new com.wcohen.ss.JaroWinkler())
@ComparatorClass("jaroWinklerTitle")
public class JaroWinklerTitle extends AbstractComparator {
public JaroWinklerTitle(Map<String, String> params){
public JaroWinklerTitle(Map<String, String> params) {
super(params, new com.wcohen.ss.JaroWinkler());
}

106
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/JsonListMatch.java
View File

@ -1,77 +1,81 @@
package eu.dnetlib.pace.tree;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import eu.dnetlib.pace.util.MapDocumentUtil;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
package eu.dnetlib.pace.tree;
import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.stream.Collectors;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import eu.dnetlib.pace.util.MapDocumentUtil;
@ComparatorClass("jsonListMatch")
public class JsonListMatch extends AbstractComparator {
private static final Log log = LogFactory.getLog(JsonListMatch.class);
private Map<String, String> params;
private static final Log log = LogFactory.getLog(JsonListMatch.class);
private Map<String, String> params;
private String MODE; //"percentage" or "count"
private String MODE; // "percentage" or "count"
public JsonListMatch(final Map<String, String> params) {
super(params);
this.params = params;
public JsonListMatch(final Map<String, String> params) {
super(params);
this.params = params;
MODE = params.getOrDefault("mode", "percentage");
}
MODE = params.getOrDefault("mode", "percentage");
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
@Override
public double compare(final Field a, final Field b, final Config conf) {
final List<String> sa = ((FieldList) a).stringList();
final List<String> sb = ((FieldList) b).stringList();
final List<String> sa = ((FieldList) a).stringList();
final List<String> sb = ((FieldList) b).stringList();
if (sa.isEmpty() || sb.isEmpty()) {
return -1;
}
if (sa.isEmpty() || sb.isEmpty()) {
return -1;
}
final Set<String> ca = sa.stream().map(this::toComparableString).collect(Collectors.toSet());
final Set<String> cb = sb.stream().map(this::toComparableString).collect(Collectors.toSet());
final Set<String> ca = sa.stream().map(this::toComparableString).collect(Collectors.toSet());
final Set<String> cb = sb.stream().map(this::toComparableString).collect(Collectors.toSet());
int incommon = Sets.intersection(ca, cb).size();
int simDiff = Sets.symmetricDifference(ca, cb).size();
int incommon = Sets.intersection(ca, cb).size();
int simDiff = Sets.symmetricDifference(ca, cb).size();
if (incommon + simDiff == 0) {
return 0.0;
}
if (incommon + simDiff == 0) {
return 0.0;
}
if (MODE.equals("percentage"))
return (double)incommon / (incommon + simDiff);
else
return incommon;
if (MODE.equals("percentage"))
return (double) incommon / (incommon + simDiff);
else
return incommon;
}
}
//converts every json into a comparable string basing on parameters
private String toComparableString(String json){
// converts every json into a comparable string basing on parameters
private String toComparableString(String json) {
StringBuilder st = new StringBuilder(); //to build the string used for comparisons basing on the jpath into parameters
StringBuilder st = new StringBuilder(); // to build the string used for comparisons basing on the jpath into
// parameters
//for each path in the param list
for (String key: params.keySet().stream().filter(k -> k.contains("jpath")).collect(Collectors.toList())) {
String path = params.get(key);
String value = MapDocumentUtil.getJPathString(path, json);
if (value == null || value.isEmpty())
value = "";
st.append( value + "::");
}
// for each path in the param list
for (String key : params.keySet().stream().filter(k -> k.contains("jpath")).collect(Collectors.toList())) {
String path = params.get(key);
String value = MapDocumentUtil.getJPathString(path, json);
if (value == null || value.isEmpty())
value = "";
st.append(value + "::");
}
st.setLength(st.length()-2);
return st.toString();
}
st.setLength(st.length() - 2);
return st.toString();
}
}

61
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/KeywordMatch.java
View File

@ -1,47 +1,50 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import java.util.Set;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
import java.util.Set;
@ComparatorClass("keywordMatch")
public class KeywordMatch extends AbstractComparator {
Map<String, String> params;
Map<String, String> params;
public KeywordMatch(Map<String, String> params) {
super(params);
this.params = params;
}
public KeywordMatch(Map<String, String> params) {
super(params);
this.params = params;
}
@Override
public double distance(final String a, final String b, final Config conf) {
@Override
public double distance(final String a, final String b, final Config conf) {
String ca = cleanup(a);
String cb = cleanup(b);
String ca = cleanup(a);
String cb = cleanup(b);
ca = normalize(ca);
cb = normalize(cb);
ca = normalize(ca);
cb = normalize(cb);
ca = filterAllStopWords(ca);
cb = filterAllStopWords(cb);
ca = filterAllStopWords(ca);
cb = filterAllStopWords(cb);
Set<String> keywords1 = getKeywords(ca, conf.translationMap(), Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> keywords2 = getKeywords(cb, conf.translationMap(), Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> keywords1 = getKeywords(
ca, conf.translationMap(), Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> keywords2 = getKeywords(
cb, conf.translationMap(), Integer.parseInt(params.getOrDefault("windowSize", "4")));
Set<String> codes1 = toCodes(keywords1, conf.translationMap());
Set<String> codes2 = toCodes(keywords2, conf.translationMap());
Set<String> codes1 = toCodes(keywords1, conf.translationMap());
Set<String> codes2 = toCodes(keywords2, conf.translationMap());
//if no cities are detected, the comparator gives 1.0
if (codes1.isEmpty() && codes2.isEmpty())
return 1.0;
else {
if (codes1.isEmpty() ^ codes2.isEmpty())
return -1.0; //undefined if one of the two has no keywords
return commonElementsPercentage(codes1, codes2);
}
}
// if no cities are detected, the comparator gives 1.0
if (codes1.isEmpty() && codes2.isEmpty())
return 1.0;
else {
if (codes1.isEmpty() ^ codes2.isEmpty())
return -1.0; // undefined if one of the two has no keywords
return commonElementsPercentage(codes1, codes2);
}
}
}

8
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/Level2JaroWinkler.java
View File

@ -1,15 +1,17 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
@ComparatorClass("level2JaroWinkler")
public class Level2JaroWinkler extends AbstractComparator {
public Level2JaroWinkler(Map<String, String> params){
public Level2JaroWinkler(Map<String, String> params) {
super(params, new com.wcohen.ss.Level2JaroWinkler());
}

13
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/Level2JaroWinklerTitle.java
View File

@ -1,16 +1,18 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import eu.dnetlib.pace.config.Config;
import java.util.Map;
@ComparatorClass("level2JaroWinklerTitle")
public class Level2JaroWinklerTitle extends AbstractComparator {
public Level2JaroWinklerTitle(Map<String,String> params){
public Level2JaroWinklerTitle(Map<String, String> params) {
super(params, new com.wcohen.ss.Level2JaroWinkler());
}
@ -29,7 +31,8 @@ public class Level2JaroWinklerTitle extends AbstractComparator {
final boolean check = checkNumbers(ca, cb);
if (check) return 0.5;
if (check)
return 0.5;
return ssalgo.score(ca, cb);
}

8
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/Level2Levenstein.java
View File

@ -1,15 +1,17 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
@ComparatorClass("level2Levenstein")
public class Level2Levenstein extends AbstractComparator {
public Level2Levenstein(Map<String,String> params){
public Level2Levenstein(Map<String, String> params) {
super(params, new com.wcohen.ss.Level2Levenstein());
}

8
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/Levenstein.java
View File

@ -1,15 +1,17 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
@ComparatorClass("levenstein")
public class Levenstein extends AbstractComparator {
public Levenstein(Map<String,String> params){
public Levenstein(Map<String, String> params) {
super(params, new com.wcohen.ss.Levenstein());
}

17
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/LevensteinTitle.java
View File

@ -1,21 +1,23 @@
package eu.dnetlib.pace.tree;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import eu.dnetlib.pace.config.Config;
import java.util.Map;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("levensteinTitle")
public class LevensteinTitle extends AbstractComparator {
private static final Log log = LogFactory.getLog(LevensteinTitle.class);
public LevensteinTitle(Map<String,String> params){
public LevensteinTitle(Map<String, String> params) {
super(params, new com.wcohen.ss.Levenstein());
}
@ -34,7 +36,8 @@ public class LevensteinTitle extends AbstractComparator {
final boolean check = checkNumbers(ca, cb);
if (check) return 0.5;
if (check)
return 0.5;
return normalize(ssalgo.score(ca, cb), ca.length(), cb.length());
}

11
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/LevensteinTitleIgnoreVersion.java
View File

@ -1,12 +1,13 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import eu.dnetlib.pace.config.Config;
import java.util.Map;
/**
* Compared compare between two titles, ignoring version numbers. Suitable for Software entities.
@ -14,7 +15,7 @@ import java.util.Map;
@ComparatorClass("levensteinTitleIgnoreVersion")
public class LevensteinTitleIgnoreVersion extends AbstractComparator {
public LevensteinTitleIgnoreVersion(Map<String,String> params){
public LevensteinTitleIgnoreVersion(Map<String, String> params) {
super(params, new com.wcohen.ss.Levenstein());
}

101
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/ListContainsMatch.java
View File

@ -1,17 +1,19 @@
package eu.dnetlib.pace.tree;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
package eu.dnetlib.pace.tree;
import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.stream.Collectors;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
/**
* The Class Contains match
*
@ -20,55 +22,54 @@ import java.util.stream.Collectors;
@ComparatorClass("listContainsMatch")
public class ListContainsMatch extends AbstractComparator {
private Map<String, String> params;
private boolean CASE_SENSITIVE;
private String STRING;
private String AGGREGATOR;
private Map<String, String> params;
private boolean CASE_SENSITIVE;
private String STRING;
private String AGGREGATOR;
public ListContainsMatch(Map<String, String> params) {
super(params);
this.params = params;
public ListContainsMatch(Map<String, String> params) {
super(params);
this.params = params;
//read parameters
CASE_SENSITIVE = Boolean.parseBoolean(params.getOrDefault("caseSensitive", "false"));
STRING = params.get("string");
AGGREGATOR = params.get("bool");
}
// read parameters
CASE_SENSITIVE = Boolean.parseBoolean(params.getOrDefault("caseSensitive", "false"));
STRING = params.get("string");
AGGREGATOR = params.get("bool");
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
@Override
public double compare(final Field a, final Field b, final Config conf) {
List<String> sa = ((FieldList) a).stringList();
List<String> sb = ((FieldList) b).stringList();
List<String> sa = ((FieldList) a).stringList();
List<String> sb = ((FieldList) b).stringList();
if (sa.isEmpty() || sb.isEmpty()) {
return -1;
}
if (sa.isEmpty() || sb.isEmpty()) {
return -1;
}
if (!CASE_SENSITIVE) {
sa = sa.stream().map(String::toLowerCase).collect(Collectors.toList());
sb = sb.stream().map(String::toLowerCase).collect(Collectors.toList());
STRING = STRING.toLowerCase();
}
if (!CASE_SENSITIVE) {
sa = sa.stream().map(String::toLowerCase).collect(Collectors.toList());
sb = sb.stream().map(String::toLowerCase).collect(Collectors.toList());
STRING = STRING.toLowerCase();
}
switch(AGGREGATOR) {
case "AND":
if(sa.contains(STRING) && sb.contains(STRING))
return 1.0;
break;
case "OR":
if(sa.contains(STRING) || sb.contains(STRING))
return 1.0;
break;
case "XOR":
if(sa.contains(STRING) ^ sb.contains(STRING))
return 1.0;
break;
default:
return 0.0;
}
return 0.0;
switch (AGGREGATOR) {
case "AND":
if (sa.contains(STRING) && sb.contains(STRING))
return 1.0;
break;
case "OR":
if (sa.contains(STRING) || sb.contains(STRING))
return 1.0;
break;
case "XOR":
if (sa.contains(STRING) ^ sb.contains(STRING))
return 1.0;
break;
default:
return 0.0;
}
return 0.0;
}
}
}

10
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/MustBeDifferent.java
View File

@ -1,16 +1,18 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import eu.dnetlib.pace.config.Config;
import java.util.Map;
@ComparatorClass("mustBeDifferent")
public class MustBeDifferent extends AbstractComparator {
public MustBeDifferent(Map<String,String> params){
public MustBeDifferent(Map<String, String> params) {
super(params, new com.wcohen.ss.Levenstein());
}

7
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/NullDistanceAlgo.java
View File

@ -1,12 +1,13 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.tree.support.Comparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
/**
* Not all fields of a document need to partecipate in the compare measure. We model those fields as having a
* NullDistanceAlgo.
@ -14,7 +15,7 @@ import java.util.Map;
@ComparatorClass("null")
public class NullDistanceAlgo implements Comparator {
public NullDistanceAlgo(Map<String, String> params){
public NullDistanceAlgo(Map<String, String> params) {
}
@Override

37
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/NumbersComparator.java
View File

@ -1,34 +1,35 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
@ComparatorClass("numbersComparator")
public class NumbersComparator extends AbstractComparator {
Map<String, String> params;
Map<String, String> params;
public NumbersComparator(Map<String, String> params) {
super(params);
this.params = params;
}
public NumbersComparator(Map<String, String> params) {
super(params);
this.params = params;
}
@Override
public double distance(String a, String b, Config conf) {
@Override
public double distance(String a, String b, Config conf) {
//extracts numbers from the field
String numbers1 = getNumbers(nfd(a));
String numbers2 = getNumbers(nfd(b));
// extracts numbers from the field
String numbers1 = getNumbers(nfd(a));
String numbers2 = getNumbers(nfd(b));
if (numbers1.isEmpty() || numbers2.isEmpty())
return -1.0;
if (numbers1.isEmpty() || numbers2.isEmpty())
return -1.0;
int n1 = Integer.parseInt(numbers1);
int n2 = Integer.parseInt(numbers2);
int n1 = Integer.parseInt(numbers1);
int n2 = Integer.parseInt(numbers2);
return Math.abs(n1 - n2);
}
return Math.abs(n1 - n2);
}
}

38
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/NumbersMatch.java
View File

@ -1,35 +1,35 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
@ComparatorClass("numbersMatch")
public class NumbersMatch extends AbstractComparator {
public NumbersMatch(Map<String, String> params) {
super(params);
}
public NumbersMatch(Map<String, String> params) {
super(params);
}
@Override
public double distance(String a, String b, Config conf) {
@Override
public double distance(String a, String b, Config conf) {
// extracts numbers from the field
String numbers1 = getNumbers(nfd(a));
String numbers2 = getNumbers(nfd(b));
//extracts numbers from the field
String numbers1 = getNumbers(nfd(a));
String numbers2 = getNumbers(nfd(b));
if (numbers1.isEmpty() && numbers2.isEmpty())
return 1.0;
if (numbers1.isEmpty() && numbers2.isEmpty())
return 1.0;
if (numbers1.isEmpty() || numbers2.isEmpty())
return -1.0;
if (numbers1.isEmpty() || numbers2.isEmpty())
return -1.0;
if (numbers1.equals(numbers2))
return 1.0;
if (numbers1.equals(numbers2))
return 1.0;
return 0.0;
}
return 0.0;
}
}

38
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/RomansMatch.java
View File

@ -1,35 +1,35 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
@ComparatorClass("romansMatch")
public class RomansMatch extends AbstractComparator {
public RomansMatch(Map<String, String> params) {
super(params);
}
public RomansMatch(Map<String, String> params) {
super(params);
}
@Override
public double distance(String a, String b, Config conf) {
@Override
public double distance(String a, String b, Config conf) {
// extracts romans from the field
String romans1 = getRomans(nfd(a));
String romans2 = getRomans(nfd(b));
//extracts romans from the field
String romans1 = getRomans(nfd(a));
String romans2 = getRomans(nfd(b));
if (romans1.isEmpty() && romans2.isEmpty())
return 1.0;
if (romans1.isEmpty() && romans2.isEmpty())
return 1.0;
if (romans1.isEmpty() || romans2.isEmpty())
return -1.0;
if (romans1.isEmpty() || romans2.isEmpty())
return -1.0;
if (romans1.equals(romans2))
return 1.0;
if (romans1.equals(romans2))
return 1.0;
return 0.0;
}
return 0.0;
}
}

51
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/SizeMatch.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.tree;
import java.util.List;
@ -18,34 +19,34 @@ import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("sizeMatch")
public class SizeMatch extends AbstractComparator {
/**
* Instantiates a new size match.
*
* @param params
* the parameters
*/
public SizeMatch(final Map<String, String> params) {
super(params);
}
/**
* Instantiates a new size match.
*
* @param params
* the parameters
*/
public SizeMatch(final Map<String, String> params) {
super(params);
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
@Override
public double compare(final Field a, final Field b, final Config conf) {
if (a.isEmpty() || b.isEmpty())
return -1;
if (a.isEmpty() || b.isEmpty())
return -1;
return Iterables.size(a) == Iterables.size(b) ? 1 : 0;
}
return Iterables.size(a) == Iterables.size(b) ? 1 : 0;
}
/**
* Checks if is empty.
*
* @param a
* the a
* @return true, if is empty
*/
protected boolean isEmpty(final Iterable<?> a) {
return (a == null) || Iterables.isEmpty(a);
}
/**
* Checks if is empty.
*
* @param a
* the a
* @return true, if is empty
*/
protected boolean isEmpty(final Iterable<?> a) {
return (a == null) || Iterables.isEmpty(a);
}
}

10
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/SortedJaroWinkler.java
View File

@ -1,18 +1,20 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.tree.support.AbstractSortedComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
/**
* The Class SortedJaroWinkler.
*/
@ComparatorClass("sortedJaroWinkler")
public class SortedJaroWinkler extends AbstractSortedComparator {
public SortedJaroWinkler(Map<String,String> params){
public SortedJaroWinkler(Map<String, String> params) {
super(params, new com.wcohen.ss.Levenstein());
}
@ -40,7 +42,6 @@ public class SortedJaroWinkler extends AbstractSortedComparator {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.compare.DistanceAlgo#getWeight()
*/
@Override
@ -50,7 +51,6 @@ public class SortedJaroWinkler extends AbstractSortedComparator {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.compare.SecondStringDistanceAlgo#normalize(double)
*/
@Override

10
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/SortedLevel2JaroWinkler.java
View File

@ -1,11 +1,13 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.tree.support.AbstractSortedComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
/**
* The Class SortedJaroWinkler.
*/
@ -22,7 +24,7 @@ public class SortedLevel2JaroWinkler extends AbstractSortedComparator {
super(weight, new com.wcohen.ss.Level2JaroWinkler());
}
public SortedLevel2JaroWinkler(final Map<String, String> params){
public SortedLevel2JaroWinkler(final Map<String, String> params) {
super(params, new com.wcohen.ss.Level2JaroWinkler());
}
@ -40,7 +42,6 @@ public class SortedLevel2JaroWinkler extends AbstractSortedComparator {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.compare.DistanceAlgo#getWeight()
*/
@Override
@ -50,7 +51,6 @@ public class SortedLevel2JaroWinkler extends AbstractSortedComparator {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.compare.SecondStringDistanceAlgo#normalize(double)
*/
@Override

87
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/StringContainsMatch.java
View File

@ -1,11 +1,12 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
/**
* The Class Contains match
*
@ -14,53 +15,53 @@ import java.util.Map;
@ComparatorClass("stringContainsMatch")
public class StringContainsMatch extends AbstractComparator {
private Map<String, String> params;
private Map<String, String> params;
private boolean CASE_SENSITIVE;
private String STRING;
private String AGGREGATOR;
private boolean CASE_SENSITIVE;
private String STRING;
private String AGGREGATOR;
public StringContainsMatch(Map<String, String> params) {
super(params);
this.params = params;
public StringContainsMatch(Map<String, String> params) {
super(params);
this.params = params;
//read parameters
CASE_SENSITIVE = Boolean.parseBoolean(params.getOrDefault("caseSensitive", "false"));
STRING = params.get("string");
AGGREGATOR = params.get("aggregator");
// read parameters
CASE_SENSITIVE = Boolean.parseBoolean(params.getOrDefault("caseSensitive", "false"));
STRING = params.get("string");
AGGREGATOR = params.get("aggregator");
}
}
@Override
public double distance(final String a, final String b, final Config conf) {
@Override
public double distance(final String a, final String b, final Config conf) {
String ca = a;
String cb = b;
if (!CASE_SENSITIVE) {
ca = a.toLowerCase();
cb = b.toLowerCase();
STRING = STRING.toLowerCase();
}
String ca = a;
String cb = b;
if (!CASE_SENSITIVE) {
ca = a.toLowerCase();
cb = b.toLowerCase();
STRING = STRING.toLowerCase();
}
if (AGGREGATOR != null) {
switch (AGGREGATOR) {
case "AND":
if (ca.contains(STRING) && cb.contains(STRING))
return 1.0;
break;
case "OR":
if (ca.contains(STRING) || cb.contains(STRING))
return 1.0;
break;
case "XOR":
if (ca.contains(STRING) ^ cb.contains(STRING))
return 1.0;
break;
default:
return 0.0;
}
}
if (AGGREGATOR != null) {
switch (AGGREGATOR) {
case "AND":
if (ca.contains(STRING) && cb.contains(STRING))
return 1.0;
break;
case "OR":
if (ca.contains(STRING) || cb.contains(STRING))
return 1.0;
break;
case "XOR":
if (ca.contains(STRING) ^ cb.contains(STRING))
return 1.0;
break;
default:
return 0.0;
}
}
return 0.0;
}
return 0.0;
}
}

71
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/StringListMatch.java
View File

@ -1,54 +1,57 @@
package eu.dnetlib.pace.tree;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
package eu.dnetlib.pace.tree;
import java.util.HashSet;
import java.util.Map;
import java.util.Set;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("stringListMatch")
public class StringListMatch extends AbstractComparator {
private static final Log log = LogFactory.getLog(StringListMatch.class);
private Map<String, String> params;
private static final Log log = LogFactory.getLog(StringListMatch.class);
private Map<String, String> params;
final private String TYPE; //percentage or count
final private String TYPE; // percentage or count
public StringListMatch(final Map<String, String> params) {
super(params);
this.params = params;
public StringListMatch(final Map<String, String> params) {
super(params);
this.params = params;
TYPE = params.getOrDefault("type", "percentage");
}
TYPE = params.getOrDefault("type", "percentage");
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
@Override
public double compare(final Field a, final Field b, final Config conf) {
final Set<String> pa = new HashSet<>(((FieldList) a).stringList());
final Set<String> pb = new HashSet<>(((FieldList) b).stringList());
final Set<String> pa = new HashSet<>(((FieldList) a).stringList());
final Set<String> pb = new HashSet<>(((FieldList) b).stringList());
if (pa.isEmpty() || pb.isEmpty()) {
return -1; //return undefined if one of the two lists is empty
}
if (pa.isEmpty() || pb.isEmpty()) {
return -1; // return undefined if one of the two lists is empty
}
int incommon = Sets.intersection(pa, pb).size();
int simDiff = Sets.symmetricDifference(pa, pb).size();
int incommon = Sets.intersection(pa, pb).size();
int simDiff = Sets.symmetricDifference(pa, pb).size();
if (incommon + simDiff == 0) {
return 0.0;
}
if (incommon + simDiff == 0) {
return 0.0;
}
if(TYPE.equals("percentage"))
return (double)incommon / (incommon + simDiff);
else
return incommon;
if (TYPE.equals("percentage"))
return (double) incommon / (incommon + simDiff);
else
return incommon;
}
}
}

17
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/SubStringLevenstein.java
View File

@ -1,17 +1,18 @@
package eu.dnetlib.pace.tree;
import eu.dnetlib.pace.config.Config;
import java.util.Map;
import org.apache.commons.lang3.StringUtils;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.config.Type;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import java.util.Map;
/**
* The Class SubStringLevenstein.
*/
@ -31,7 +32,7 @@ public class SubStringLevenstein extends AbstractComparator {
super(w, new com.wcohen.ss.Levenstein());
}
public SubStringLevenstein(Map<String, String> params){
public SubStringLevenstein(Map<String, String> params) {
super(params, new com.wcohen.ss.Levenstein());
this.limit = Integer.parseInt(params.getOrDefault("limit", "1"));
}
@ -66,8 +67,8 @@ public class SubStringLevenstein extends AbstractComparator {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.compare.SecondStringDistanceAlgo#compare(eu.dnetlib.pace.model.Field, eu.dnetlib.pace.model.Field)
* @see eu.dnetlib.pace.compare.SecondStringDistanceAlgo#compare(eu.dnetlib.pace.model.Field,
* eu.dnetlib.pace.model.Field)
*/
@Override
public double distance(final Field a, final Field b, final Config conf) {
@ -79,7 +80,6 @@ public class SubStringLevenstein extends AbstractComparator {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.compare.DistanceAlgo#getWeight()
*/
@Override
@ -89,7 +89,6 @@ public class SubStringLevenstein extends AbstractComparator {
/*
* (non-Javadoc)
*
* @see eu.dnetlib.pace.compare.SecondStringDistanceAlgo#normalize(double)
*/
@Override

31
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/TitleVersionMatch.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.tree;
import java.util.List;
@ -17,24 +18,24 @@ import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("titleVersionMatch")
public class TitleVersionMatch extends AbstractComparator {
public TitleVersionMatch(final Map<String, String> params) {
super(params);
}
public TitleVersionMatch(final Map<String, String> params) {
super(params);
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
final String valueA = getFirstValue(a);
final String valueB = getFirstValue(b);
@Override
public double compare(final Field a, final Field b, final Config conf) {
final String valueA = getFirstValue(a);
final String valueB = getFirstValue(b);
if (valueA.isEmpty() || valueB.isEmpty())
return -1;
if (valueA.isEmpty() || valueB.isEmpty())
return -1;
return notNull(valueA) && notNull(valueB) && !checkNumbers(valueA, valueB) ? 1 : 0;
}
return notNull(valueA) && notNull(valueB) && !checkNumbers(valueA, valueB) ? 1 : 0;
}
@Override
public String toString() {
return getClass().getSimpleName() + ":" + super.toString();
}
@Override
public String toString() {
return getClass().getSimpleName() + ":" + super.toString();
}
}

80
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/UrlMatcher.java
View File

@ -1,59 +1,61 @@
package eu.dnetlib.pace.tree;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import org.apache.commons.lang3.StringUtils;
package eu.dnetlib.pace.tree;
import java.net.MalformedURLException;
import java.net.URL;
import java.util.Map;
import org.apache.commons.lang3.StringUtils;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.tree.support.ComparatorClass;
@ComparatorClass("urlMatcher")
public class UrlMatcher extends Levenstein {
private Map<String, String> params;
private Map<String, String> params;
public UrlMatcher(Map<String, String> params){
super(params);
this.params = params;
}
public UrlMatcher(Map<String, String> params) {
super(params);
this.params = params;
}
public UrlMatcher(double weight, Map<String, String> params) {
super(weight);
this.params = params;
}
public UrlMatcher(double weight, Map<String, String> params) {
super(weight);
this.params = params;
}
public void setParams(Map<String, String> params) {
this.params = params;
}
public void setParams(Map<String, String> params) {
this.params = params;
}
@Override
public double distance(Field a, Field b, final Config conf) {
final URL urlA = asUrl(getFirstValue(a));
final URL urlB = asUrl(getFirstValue(b));
@Override
public double distance(Field a, Field b, final Config conf) {
final URL urlA = asUrl(getFirstValue(a));
final URL urlB = asUrl(getFirstValue(b));
if (!urlA.getHost().equalsIgnoreCase(urlB.getHost())) {
return 0.0;
}
if (!urlA.getHost().equalsIgnoreCase(urlB.getHost())) {
return 0.0;
}
Double hostW = Double.parseDouble(params.getOrDefault("host", "0.5"));
Double pathW = Double.parseDouble(params.getOrDefault("path", "0.5"));
Double hostW = Double.parseDouble(params.getOrDefault("host", "0.5"));
Double pathW = Double.parseDouble(params.getOrDefault("path", "0.5"));
if (StringUtils.isBlank(urlA.getPath()) || StringUtils.isBlank(urlB.getPath())) {
return hostW * 0.5;
}
if (StringUtils.isBlank(urlA.getPath()) || StringUtils.isBlank(urlB.getPath())) {
return hostW * 0.5;
}
return hostW + pathW * super.distance(urlA.getPath(), urlB.getPath(), conf);
}
return hostW + pathW * super.distance(urlA.getPath(), urlB.getPath(), conf);
}
private URL asUrl(final String value) {
try {
return new URL(value);
} catch (MalformedURLException e) {
// should not happen as checked by pace typing
throw new IllegalStateException("invalid URL: " + value);
}
}
private URL asUrl(final String value) {
try {
return new URL(value);
} catch (MalformedURLException e) {
// should not happen as checked by pace typing
throw new IllegalStateException("invalid URL: " + value);
}
}
}

56
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/YearMatch.java
View File

@ -1,12 +1,14 @@
package eu.dnetlib.pace.tree;
import java.util.Map;
import org.apache.commons.lang3.StringUtils;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.tree.support.AbstractComparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import org.apache.commons.lang3.StringUtils;
import java.util.Map;
/**
* Returns true if the year of the date field in the given documents are the same, false when any of the two is invalid or it's missing.
@ -16,36 +18,36 @@ import java.util.Map;
@ComparatorClass("yearMatch")
public class YearMatch extends AbstractComparator {
private int limit = 4;
private int limit = 4;
public YearMatch(final Map<String, String> params) {
super(params);
}
public YearMatch(final Map<String, String> params) {
super(params);
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
final String valueA = getNumbers(getFirstValue(a));
final String valueB = getNumbers(getFirstValue(b));
@Override
public double compare(final Field a, final Field b, final Config conf) {
final String valueA = getNumbers(getFirstValue(a));
final String valueB = getNumbers(getFirstValue(b));
if (valueA.isEmpty() || valueB.isEmpty())
return -1;
if (valueA.isEmpty() || valueB.isEmpty())
return -1;
final boolean lengthMatch = checkLength(valueA) && checkLength(valueB);
final boolean onemissing = valueA.isEmpty() || valueB.isEmpty();
final boolean lengthMatch = checkLength(valueA) && checkLength(valueB);
final boolean onemissing = valueA.isEmpty() || valueB.isEmpty();
return lengthMatch && valueA.equals(valueB) || onemissing ? 1 : 0;
}
return lengthMatch && valueA.equals(valueB) || onemissing ? 1 : 0;
}
protected boolean checkLength(final String s) {
return s.length() == limit;
}
protected boolean checkLength(final String s) {
return s.length() == limit;
}
protected String getFirstValue(final Field value) {
return (value != null) && !value.isEmpty() ? StringUtils.left(value.stringValue(), limit) : "";
}
protected String getFirstValue(final Field value) {
return (value != null) && !value.isEmpty() ? StringUtils.left(value.stringValue(), limit) : "";
}
@Override
public String toString() {
return getClass().getSimpleName() + ":" + super.toString();
}
@Override
public String toString() {
return getClass().getSimpleName() + ":" + super.toString();
}
}

196
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/AbstractComparator.java
View File

@ -1,124 +1,130 @@
package eu.dnetlib.pace.tree.support;
import java.util.List;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.config.Type;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
import java.util.List;
import java.util.Map;
public abstract class AbstractComparator extends AbstractPaceFunctions implements Comparator {
/** The ssalgo. */
protected AbstractStringDistance ssalgo;
/** The ssalgo. */
protected AbstractStringDistance ssalgo;
/** The weight. */
protected double weight = 0.0;
/** The weight. */
protected double weight = 0.0;
private Map<String, String> params;
private Map<String, String> params;
protected AbstractComparator(Map<String, String> params) {
this.params = params;
}
protected AbstractComparator(Map<String, String> params) {
this.params = params;
}
protected AbstractComparator(Map<String, String> params, final AbstractStringDistance ssalgo){
this.params = params;
this.weight = 1.0;
this.ssalgo = ssalgo;
}
protected AbstractComparator(Map<String, String> params, final AbstractStringDistance ssalgo) {
this.params = params;
this.weight = 1.0;
this.ssalgo = ssalgo;
}
/**
* Instantiates a new second string compare algo.
*
* @param weight
* the weight
* @param ssalgo
* the ssalgo
*/
protected AbstractComparator(final double weight, final AbstractStringDistance ssalgo) {
this.ssalgo = ssalgo;
this.weight = weight;
}
/**
* Instantiates a new second string compare algo.
*
* @param weight
* the weight
* @param ssalgo
* the ssalgo
*/
protected AbstractComparator(final double weight, final AbstractStringDistance ssalgo) {
this.ssalgo = ssalgo;
this.weight = weight;
}
protected AbstractComparator(final AbstractStringDistance ssalgo){
this.ssalgo = ssalgo;
}
protected AbstractComparator(final AbstractStringDistance ssalgo) {
this.ssalgo = ssalgo;
}
/**
* Normalize.
*
* @param d
* the d
* @return the double
*/
protected double normalize(double d) {
return d;
}
/**
* Normalize.
*
* @param d
* the d
* @return the double
*/
protected double normalize(double d) {
return d;
}
/**
* Distance.
*
* @param a
* the a
* @param b
* the b
* @return the double
*/
public double distance(final String a, final String b, final Config conf) {
/**
* Distance.
*
* @param a
* the a
* @param b
* the b
* @return the double
*/
public double distance(final String a, final String b, final Config conf) {
if (a.isEmpty() || b.isEmpty()) {
return -1; //return -1 if a field is missing
}
double score = ssalgo.score(a, b);
return normalize(score);
}
if (a.isEmpty() || b.isEmpty()) {
return -1; // return -1 if a field is missing
}
double score = ssalgo.score(a, b);
return normalize(score);
}
/**
* Distance.
*
* @param a
* the a
* @param b
* the b
* @return the double
*/
protected double distance(final List<String> a, final List<String> b, final Config conf) {
return distance(concat(a), concat(b), conf);
}
/**
* Distance.
*
* @param a
* the a
* @param b
* the b
* @return the double
*/
protected double distance(final List<String> a, final List<String> b, final Config conf) {
return distance(concat(a), concat(b), conf);
}
public double distance(final Field a, final Field b, final Config conf) {
if (a.getType().equals(Type.String) && b.getType().equals(Type.String)) return distance(a.stringValue(), b.stringValue(), conf);
if (a.getType().equals(Type.List) && b.getType().equals(Type.List)) return distance(toList(a), toList(b), conf);
public double distance(final Field a, final Field b, final Config conf) {
if (a.getType().equals(Type.String) && b.getType().equals(Type.String))
return distance(a.stringValue(), b.stringValue(), conf);
if (a.getType().equals(Type.List) && b.getType().equals(Type.List))
return distance(toList(a), toList(b), conf);
throw new IllegalArgumentException("invalid types\n- A: " + a.toString() + "\n- B: " + b.toString());
}
throw new IllegalArgumentException("invalid types\n- A: " + a.toString() + "\n- B: " + b.toString());
}
@Override
public double compare(final Field a, final Field b, final Config conf) {
if (a.isEmpty() || b.isEmpty())
return -1;
if (a.getType().equals(Type.String) && b.getType().equals(Type.String)) return distance(a.stringValue(), b.stringValue(), conf);
if (a.getType().equals(Type.List) && b.getType().equals(Type.List)) return distance(toList(a), toList(b), conf);
@Override
public double compare(final Field a, final Field b, final Config conf) {
if (a.isEmpty() || b.isEmpty())
return -1;
if (a.getType().equals(Type.String) && b.getType().equals(Type.String))
return distance(a.stringValue(), b.stringValue(), conf);
if (a.getType().equals(Type.List) && b.getType().equals(Type.List))
return distance(toList(a), toList(b), conf);
throw new IllegalArgumentException("invalid types\n- A: " + a.toString() + "\n- B: " + b.toString());
}
throw new IllegalArgumentException("invalid types\n- A: " + a.toString() + "\n- B: " + b.toString());
}
/**
* To list.
*
* @param list
* the list
* @return the list
*/
protected List<String> toList(final Field list) {
return ((FieldList) list).stringList();
}
/**
* To list.
*
* @param list
* the list
* @return the list
*/
protected List<String> toList(final Field list) {
return ((FieldList) list).stringList();
}
public double getWeight(){
return this.weight;
}
public double getWeight() {
return this.weight;
}
}

54
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/AbstractSortedComparator.java
View File

@ -1,38 +1,40 @@
package eu.dnetlib.pace.tree.support;
import com.google.common.collect.Lists;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
package eu.dnetlib.pace.tree.support;
import java.util.Collections;
import java.util.List;
import java.util.Map;
import com.google.common.collect.Lists;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldList;
public abstract class AbstractSortedComparator extends AbstractComparator {
/**
* Instantiates a new sorted second string compare algo.
*
* @param weight
* the weight
* @param ssalgo
* the ssalgo
*/
protected AbstractSortedComparator(final double weight, final AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
/**
* Instantiates a new sorted second string compare algo.
*
* @param weight
* the weight
* @param ssalgo
* the ssalgo
*/
protected AbstractSortedComparator(final double weight, final AbstractStringDistance ssalgo) {
super(weight, ssalgo);
}
protected AbstractSortedComparator(final Map<String, String> params, final AbstractStringDistance ssalgo){
super(Double.parseDouble(params.get("weight")), ssalgo);
}
protected AbstractSortedComparator(final Map<String, String> params, final AbstractStringDistance ssalgo) {
super(Double.parseDouble(params.get("weight")), ssalgo);
}
@Override
protected List<String> toList(final Field list) {
FieldList fl = (FieldList) list;
List<String> values = Lists.newArrayList(fl.stringList());
Collections.sort(values);
return values;
}
@Override
protected List<String> toList(final Field list) {
FieldList fl = (FieldList) list;
List<String> values = Lists.newArrayList(fl.stringList());
Collections.sort(values);
return values;
}
}

27
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/AggType.java
View File

@ -1,24 +1,21 @@
package eu.dnetlib.pace.tree.support;
import eu.dnetlib.pace.util.PaceException;
public enum AggType {
W_MEAN, //weighted mean
AVG, //average
SUM,
MAX,
MIN,
AND, //used for necessary conditions
OR; //used for sufficient conditions
W_MEAN, // weighted mean
AVG, // average
SUM, MAX, MIN, AND, // used for necessary conditions
OR; // used for sufficient conditions
public static AggType getEnum(String value) {
public static AggType getEnum(String value) {
try {
return AggType.valueOf(value);
}
catch (IllegalArgumentException e) {
throw new PaceException("Undefined aggregation type", e);
}
}
try {
return AggType.valueOf(value);
} catch (IllegalArgumentException e) {
throw new PaceException("Undefined aggregation type", e);
}
}
}

10
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/Comparator.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.tree.support;
import eu.dnetlib.pace.config.Config;
@ -5,10 +6,9 @@ import eu.dnetlib.pace.model.Field;
public interface Comparator {
/*
* return : -1 -> can't decide (i.e. missing field)
* >0 -> similarity degree (depends on the algorithm)
* */
public double compare(Field a, Field b, Config conf);
/*
* return : -1 -> can't decide (i.e. missing field) >0 -> similarity degree (depends on the algorithm)
*/
public double compare(Field a, Field b, Config conf);
}

3
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/ComparatorClass.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.tree.support;
import java.lang.annotation.ElementType;
@ -9,5 +10,5 @@ import java.lang.annotation.Target;
@Target(ElementType.TYPE)
public @interface ComparatorClass {
public String value();
public String value();
}

114
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/FieldConf.java
View File

@ -1,82 +1,84 @@
package eu.dnetlib.pace.tree.support;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.pace.util.PaceException;
import java.io.IOException;
import java.io.Serializable;
import java.util.Map;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.pace.util.PaceException;
/**
* The class that defines the configuration of each field in the decision tree.
* */
public class FieldConf implements Serializable {
private String field; //name of the field on which apply the comparator
private String comparator; //comparator name
private double weight = 1.0; //weight for the field (to be used in the aggregation)
private Map<String,String> params; //parameters
private String field; // name of the field on which apply the comparator
private String comparator; // comparator name
private double weight = 1.0; // weight for the field (to be used in the aggregation)
private Map<String, String> params; // parameters
private boolean countIfUndefined;
private boolean countIfUndefined;
public boolean isCountIfUndefined() {
return countIfUndefined;
}
public boolean isCountIfUndefined() {
return countIfUndefined;
}
public void setCountIfUndefined(boolean countIfUndefined) {
this.countIfUndefined = countIfUndefined;
}
public void setCountIfUndefined(boolean countIfUndefined) {
this.countIfUndefined = countIfUndefined;
}
public FieldConf() {
}
public FieldConf() {
}
public FieldConf(String field, String comparator, double weight, Map<String, String> params, boolean countIfUndefined) {
this.field = field;
this.comparator = comparator;
this.weight = weight;
this.params = params;
this.countIfUndefined = countIfUndefined;
}
public FieldConf(String field, String comparator, double weight, Map<String, String> params,
boolean countIfUndefined) {
this.field = field;
this.comparator = comparator;
this.weight = weight;
this.params = params;
this.countIfUndefined = countIfUndefined;
}
public String getField() {
return field;
}
public String getField() {
return field;
}
public void setField(String field) {
this.field = field;
}
public void setField(String field) {
this.field = field;
}
public String getComparator() {
return comparator;
}
public String getComparator() {
return comparator;
}
public void setComparator(String comparator) {
this.comparator = comparator;
}
public void setComparator(String comparator) {
this.comparator = comparator;
}
public double getWeight() {
return weight;
}
public double getWeight() {
return weight;
}
public void setWeight(double weight) {
this.weight = weight;
}
public void setWeight(double weight) {
this.weight = weight;
}
public Map<String, String> getParams() {
return params;
}
public Map<String, String> getParams() {
return params;
}
public void setParams(Map<String, String> params) {
this.params = params;
}
public void setParams(Map<String, String> params) {
this.params = params;
}
@Override
public String toString() {
try {
return new ObjectMapper().writeValueAsString(this);
} catch (IOException e) {
throw new PaceException("Impossible to convert to JSON: ", e);
}
}
@Override
public String toString() {
try {
return new ObjectMapper().writeValueAsString(this);
} catch (IOException e) {
throw new PaceException("Impossible to convert to JSON: ", e);
}
}
}

127
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/FieldStats.java
View File

@ -1,89 +1,90 @@
package eu.dnetlib.pace.tree.support;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.util.PaceException;
import java.io.IOException;
import java.io.Serializable;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.util.PaceException;
/**
* The class that contains the result of each comparison in the decision tree
* */
public class FieldStats implements Serializable {
private double weight; //weight for the field (to be used in the aggregation)
private double threshold; //threshold for the field (to be used in some kind of aggregations)
private double result; //the result of the comparison
private Field a;
private Field b;
private double weight; // weight for the field (to be used in the aggregation)
private double threshold; // threshold for the field (to be used in some kind of aggregations)
private double result; // the result of the comparison
private Field a;
private Field b;
private boolean countIfUndefined;
private boolean countIfUndefined;
public FieldStats(double weight, double threshold, double result, boolean countIfUndefined, Field a, Field b) {
this.weight = weight;
this.threshold = threshold;
this.result = result;
this.countIfUndefined = countIfUndefined;
this.a = a;
this.b = b;
}
public FieldStats(double weight, double threshold, double result, boolean countIfUndefined, Field a, Field b) {
this.weight = weight;
this.threshold = threshold;
this.result = result;
this.countIfUndefined = countIfUndefined;
this.a = a;
this.b = b;
}
public double getThreshold() {
return threshold;
}
public double getThreshold() {
return threshold;
}
public void setThreshold(double threshold) {
this.threshold = threshold;
}
public void setThreshold(double threshold) {
this.threshold = threshold;
}
public double getWeight() {
return weight;
}
public double getWeight() {
return weight;
}
public void setWeight(double weight) {
this.weight = weight;
}
public void setWeight(double weight) {
this.weight = weight;
}
public double getResult() {
return result;
}
public double getResult() {
return result;
}
public void setResult(double result) {
this.result = result;
}
public void setResult(double result) {
this.result = result;
}
public boolean isCountIfUndefined() {
return countIfUndefined;
}
public boolean isCountIfUndefined() {
return countIfUndefined;
}
public void setCountIfUndefined(boolean countIfUndefined) {
this.countIfUndefined = countIfUndefined;
}
public void setCountIfUndefined(boolean countIfUndefined) {
this.countIfUndefined = countIfUndefined;
}
public Field getA() {
return a;
}
public Field getA() {
return a;
}
public void setA(Field a) {
this.a = a;
}
public void setA(Field a) {
this.a = a;
}
public Field getB() {
return b;
}
public Field getB() {
return b;
}
public void setB(Field b) {
this.b = b;
}
public void setB(Field b) {
this.b = b;
}
@Override
public String toString(){
try {
return new ObjectMapper().writeValueAsString(this);
} catch (IOException e) {
throw new PaceException("Impossible to convert to JSON: ", e);
}
}
@Override
public String toString() {
try {
return new ObjectMapper().writeValueAsString(this);
} catch (IOException e) {
throw new PaceException("Impossible to convert to JSON: ", e);
}
}
}

20
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/MatchType.java
View File

@ -1,18 +1,16 @@
package eu.dnetlib.pace.tree.support;
public enum MatchType {
MATCH,
NO_MATCH,
UNDEFINED;
MATCH, NO_MATCH, UNDEFINED;
public static MatchType parse(String value) {
public static MatchType parse(String value) {
try {
return MatchType.valueOf(value);
}
catch (IllegalArgumentException e) {
return MatchType.UNDEFINED; //return UNDEFINED if the enum is not parsable
}
}
try {
return MatchType.valueOf(value);
} catch (IllegalArgumentException e) {
return MatchType.UNDEFINED; // return UNDEFINED if the enum is not parsable
}
}
}

228
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/TreeNodeDef.java
View File

@ -1,150 +1,160 @@
package eu.dnetlib.pace.tree.support;
import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.config.PaceConfig;
import eu.dnetlib.pace.model.MapDocument;
import eu.dnetlib.pace.util.PaceException;
import org.apache.commons.lang3.StringUtils;
package eu.dnetlib.pace.tree.support;
import java.io.IOException;
import java.io.Serializable;
import java.io.StringWriter;
import java.util.List;
import org.apache.commons.lang3.StringUtils;
import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.config.PaceConfig;
import eu.dnetlib.pace.model.MapDocument;
import eu.dnetlib.pace.util.PaceException;
public class TreeNodeDef implements Serializable {
final static String CROSS_COMPARE = "crossCompare";
final static String CROSS_COMPARE = "crossCompare";
private List<FieldConf> fields;
private AggType aggregation;
private List<FieldConf> fields;
private AggType aggregation;
private double threshold;
private double threshold;
private String positive;
private String negative;
private String undefined;
private String positive;
private String negative;
private String undefined;
boolean ignoreUndefined;
boolean ignoreUndefined;
public TreeNodeDef(List<FieldConf> fields, AggType aggregation, double threshold, String positive, String negative, String undefined, boolean ignoreUndefined) {
this.fields = fields;
this.aggregation = aggregation;
this.threshold = threshold;
this.positive = positive;
this.negative = negative;
this.undefined = undefined;
this.ignoreUndefined = ignoreUndefined;
}
public TreeNodeDef(List<FieldConf> fields, AggType aggregation, double threshold, String positive, String negative,
String undefined, boolean ignoreUndefined) {
this.fields = fields;
this.aggregation = aggregation;
this.threshold = threshold;
this.positive = positive;
this.negative = negative;
this.undefined = undefined;
this.ignoreUndefined = ignoreUndefined;
}
public TreeNodeDef() {}
public TreeNodeDef() {
}
//function for the evaluation of the node
public TreeNodeStats evaluate(MapDocument doc1, MapDocument doc2, Config conf) {
// function for the evaluation of the node
public TreeNodeStats evaluate(MapDocument doc1, MapDocument doc2, Config conf) {
TreeNodeStats stats = new TreeNodeStats();
TreeNodeStats stats = new TreeNodeStats();
//for each field in the node, it computes the
for (FieldConf fieldConf : fields) {
// for each field in the node, it computes the
for (FieldConf fieldConf : fields) {
double weight = fieldConf.getWeight();
double weight = fieldConf.getWeight();
double result;
double result;
//if the param specifies a cross comparison (i.e. compare elements from different fields), compute the result for both sides and return the maximum
if(fieldConf.getParams().keySet().stream().anyMatch(k -> k.contains(CROSS_COMPARE))) {
String crossField = fieldConf.getParams().get(CROSS_COMPARE);
double result1 = comparator(fieldConf).compare(doc1.getFieldMap().get(fieldConf.getField()), doc2.getFieldMap().get(crossField), conf);
double result2 = comparator(fieldConf).compare(doc1.getFieldMap().get(crossField), doc2.getFieldMap().get(fieldConf.getField()), conf);
result = Math.max(result1,result2);
}
else {
result = comparator(fieldConf).compare(doc1.getFieldMap().get(fieldConf.getField()), doc2.getFieldMap().get(fieldConf.getField()), conf);
}
// if the param specifies a cross comparison (i.e. compare elements from different fields), compute the
// result for both sides and return the maximum
if (fieldConf.getParams().keySet().stream().anyMatch(k -> k.contains(CROSS_COMPARE))) {
String crossField = fieldConf.getParams().get(CROSS_COMPARE);
double result1 = comparator(fieldConf)
.compare(doc1.getFieldMap().get(fieldConf.getField()), doc2.getFieldMap().get(crossField), conf);
double result2 = comparator(fieldConf)
.compare(doc1.getFieldMap().get(crossField), doc2.getFieldMap().get(fieldConf.getField()), conf);
result = Math.max(result1, result2);
} else {
result = comparator(fieldConf)
.compare(
doc1.getFieldMap().get(fieldConf.getField()), doc2.getFieldMap().get(fieldConf.getField()),
conf);
}
stats.addFieldStats(
fieldConf.getComparator() + " on " + fieldConf.getField() + " " + fields.indexOf(fieldConf),
new FieldStats(
weight,
Double.parseDouble(fieldConf.getParams().getOrDefault("threshold", "1.0")),
result,
fieldConf.isCountIfUndefined(),
doc1.getFieldMap().get(fieldConf.getField()),
doc2.getFieldMap().get(fieldConf.getField())
));
}
stats
.addFieldStats(
fieldConf.getComparator() + " on " + fieldConf.getField() + " " + fields.indexOf(fieldConf),
new FieldStats(
weight,
Double.parseDouble(fieldConf.getParams().getOrDefault("threshold", "1.0")),
result,
fieldConf.isCountIfUndefined(),
doc1.getFieldMap().get(fieldConf.getField()),
doc2.getFieldMap().get(fieldConf.getField())));
}
return stats;
}
return stats;
}
private Comparator comparator(final FieldConf field){
private Comparator comparator(final FieldConf field) {
return PaceConfig.resolver.getComparator(field.getComparator(), field.getParams());
}
return PaceConfig.resolver.getComparator(field.getComparator(), field.getParams());
}
public List<FieldConf> getFields() {
return fields;
}
public List<FieldConf> getFields() {
return fields;
}
public void setFields(List<FieldConf> fields) {
this.fields = fields;
}
public void setFields(List<FieldConf> fields) {
this.fields = fields;
}
public AggType getAggregation() {
return aggregation;
}
public AggType getAggregation() {
return aggregation;
}
public void setAggregation(AggType aggregation) {
this.aggregation = aggregation;
}
public void setAggregation(AggType aggregation) {
this.aggregation = aggregation;
}
public double getThreshold() {
return threshold;
}
public double getThreshold() {
return threshold;
}
public void setThreshold(double threshold) {
this.threshold = threshold;
}
public void setThreshold(double threshold) {
this.threshold = threshold;
}
public String getPositive() {
return positive;
}
public String getPositive() {
return positive;
}
public void setPositive(String positive) {
this.positive = positive;
}
public void setPositive(String positive) {
this.positive = positive;
}
public String getNegative() {
return negative;
}
public String getNegative() {
return negative;
}
public void setNegative(String negative) {
this.negative = negative;
}
public void setNegative(String negative) {
this.negative = negative;
}
public String getUndefined() {
return undefined;
}
public String getUndefined() {
return undefined;
}
public void setUndefined(String undefined) {
this.undefined = undefined;
}
public void setUndefined(String undefined) {
this.undefined = undefined;
}
public boolean isIgnoreUndefined() {
return ignoreUndefined;
}
public boolean isIgnoreUndefined() {
return ignoreUndefined;
}
public void setIgnoreUndefined(boolean ignoreUndefined) {
this.ignoreUndefined = ignoreUndefined;
}
public void setIgnoreUndefined(boolean ignoreUndefined) {
this.ignoreUndefined = ignoreUndefined;
}
@Override
public String toString() {
try {
return new ObjectMapper().writeValueAsString(this);
} catch (IOException e) {
throw new PaceException("Impossible to convert to JSON: ", e);
}
}
@Override
public String toString() {
try {
return new ObjectMapper().writeValueAsString(this);
} catch (IOException e) {
throw new PaceException("Impossible to convert to JSON: ", e);
}
}
}

218
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/TreeNodeStats.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.tree.support;
import java.io.Serializable;
@ -6,129 +7,128 @@ import java.util.Map;
public class TreeNodeStats implements Serializable {
private Map<String, FieldStats> results; //this is an accumulator for the results of the node
private Map<String, FieldStats> results; // this is an accumulator for the results of the node
public TreeNodeStats(){
this.results = new HashMap<>();
}
public TreeNodeStats() {
this.results = new HashMap<>();
}
public Map<String, FieldStats> getResults() {
return results;
}
public Map<String, FieldStats> getResults() {
return results;
}
public void addFieldStats(String id, FieldStats fieldStats){
this.results.put(id, fieldStats);
}
public void addFieldStats(String id, FieldStats fieldStats) {
this.results.put(id, fieldStats);
}
public int fieldsCount(){
return this.results.size();
}
public int fieldsCount() {
return this.results.size();
}
public int undefinedCount(){
int undefinedCount = 0;
for(FieldStats fs: this.results.values()){
if(fs.getResult() == -1)
undefinedCount ++;
}
return undefinedCount;
}
public int undefinedCount() {
int undefinedCount = 0;
for (FieldStats fs : this.results.values()) {
if (fs.getResult() == -1)
undefinedCount++;
}
return undefinedCount;
}
public double scoreSum(){
double scoreSum = 0.0;
for(FieldStats fs: this.results.values()){
if(fs.getResult()>=0.0) {
scoreSum += fs.getResult();
}
}
return scoreSum;
}
public double scoreSum() {
double scoreSum = 0.0;
for (FieldStats fs : this.results.values()) {
if (fs.getResult() >= 0.0) {
scoreSum += fs.getResult();
}
}
return scoreSum;
}
//return the sum of the weights without considering the fields with countIfMissing=false && result=-1
public double weightSum(){
double weightSum = 0.0;
for(FieldStats fs: this.results.values()){
if(fs.getResult()>=0.0 || (fs.getResult()<0.0 && fs.isCountIfUndefined())) {
weightSum += fs.getWeight();
}
}
return weightSum;
}
// return the sum of the weights without considering the fields with countIfMissing=false && result=-1
public double weightSum() {
double weightSum = 0.0;
for (FieldStats fs : this.results.values()) {
if (fs.getResult() >= 0.0 || (fs.getResult() < 0.0 && fs.isCountIfUndefined())) {
weightSum += fs.getWeight();
}
}
return weightSum;
}
public double weightedScoreSum(){
double weightedScoreSum = 0.0;
for(FieldStats fs: this.results.values()){
if(fs.getResult()>=0.0) {
weightedScoreSum += fs.getResult()*fs.getWeight();
}
}
return weightedScoreSum;
}
public double weightedScoreSum() {
double weightedScoreSum = 0.0;
for (FieldStats fs : this.results.values()) {
if (fs.getResult() >= 0.0) {
weightedScoreSum += fs.getResult() * fs.getWeight();
}
}
return weightedScoreSum;
}
public double max(){
double max = -1.0;
for(FieldStats fs: this.results.values()){
if(fs.getResult()>max)
max = fs.getResult();
}
return max;
}
public double max() {
double max = -1.0;
for (FieldStats fs : this.results.values()) {
if (fs.getResult() > max)
max = fs.getResult();
}
return max;
}
public double min(){
double min = 100.0; //random high value
for(FieldStats fs: this.results.values()){
if(fs.getResult()<min) {
if (fs.getResult()>=0.0 || (fs.getResult() == -1 && fs.isCountIfUndefined()))
min = fs.getResult();
}
}
return min;
}
public double min() {
double min = 100.0; // random high value
for (FieldStats fs : this.results.values()) {
if (fs.getResult() < min) {
if (fs.getResult() >= 0.0 || (fs.getResult() == -1 && fs.isCountIfUndefined()))
min = fs.getResult();
}
}
return min;
}
//if at least one is true, return 1.0
public double or(){
for (FieldStats fieldStats : this.results.values()) {
if (fieldStats.getResult() >= fieldStats.getThreshold())
return 1.0;
}
return 0.0;
}
// if at least one is true, return 1.0
public double or() {
for (FieldStats fieldStats : this.results.values()) {
if (fieldStats.getResult() >= fieldStats.getThreshold())
return 1.0;
}
return 0.0;
}
//if at least one is false, return 0.0
public double and() {
for (FieldStats fieldStats : this.results.values()) {
// if at least one is false, return 0.0
public double and() {
for (FieldStats fieldStats : this.results.values()) {
if (fieldStats.getResult() == -1) {
if (fieldStats.isCountIfUndefined())
return 0.0;
}
else {
if (fieldStats.getResult() < fieldStats.getThreshold())
return 0.0;
}
if (fieldStats.getResult() == -1) {
if (fieldStats.isCountIfUndefined())
return 0.0;
} else {
if (fieldStats.getResult() < fieldStats.getThreshold())
return 0.0;
}
}
return 1.0;
}
}
return 1.0;
}
public double getFinalScore(AggType aggregation){
public double getFinalScore(AggType aggregation) {
switch (aggregation){
case AVG:
return scoreSum()/fieldsCount();
case SUM:
return scoreSum();
case MAX:
return max();
case MIN:
return min();
case W_MEAN:
return weightedScoreSum()/weightSum();
case OR:
return or();
case AND:
return and();
default:
return 0.0;
}
}
switch (aggregation) {
case AVG:
return scoreSum() / fieldsCount();
case SUM:
return scoreSum();
case MAX:
return max();
case MIN:
return min();
case W_MEAN:
return weightedScoreSum() / weightSum();
case OR:
return or();
case AND:
return and();
default:
return 0.0;
}
}
}

39
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/TreeProcessor.java
View File

@ -1,16 +1,17 @@
package eu.dnetlib.pace.tree.support;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import eu.dnetlib.pace.config.Config;
import eu.dnetlib.pace.model.MapDocument;
import eu.dnetlib.pace.util.PaceException;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
/**
* The compare between two documents is given by the weighted mean of the field distances
*/
public class TreeProcessor{
public class TreeProcessor {
private static final Log log = LogFactory.getLog(TreeProcessor.class);
@ -21,35 +22,34 @@ public class TreeProcessor{
}
public boolean compare(final MapDocument a, final MapDocument b) {
//evaluate the decision tree
// evaluate the decision tree
return evaluateTree(a, b).getResult() == MatchType.MATCH;
}
public TreeStats evaluateTree(final MapDocument doc1, final MapDocument doc2){
public TreeStats evaluateTree(final MapDocument doc1, final MapDocument doc2) {
TreeStats treeStats = new TreeStats();
String current = "start";
while (MatchType.parse(current)==MatchType.UNDEFINED) {
while (MatchType.parse(current) == MatchType.UNDEFINED) {
TreeNodeDef currentNode = config.decisionTree().get(current);
//throw an exception if the node doesn't exist
// throw an exception if the node doesn't exist
if (currentNode == null)
throw new PaceException("Missing tree node: " + current);
TreeNodeStats stats = currentNode.evaluate(doc1, doc2, config);
treeStats.addNodeStats(current, stats);
//if ignoreUndefined=false the miss is considered as undefined
if (!currentNode.isIgnoreUndefined() && stats.undefinedCount()>0) {
// if ignoreUndefined=false the miss is considered as undefined
if (!currentNode.isIgnoreUndefined() && stats.undefinedCount() > 0) {
current = currentNode.getUndefined();
}
//if ignoreUndefined=true the miss is ignored and the score computed anyway
// if ignoreUndefined=true the miss is ignored and the score computed anyway
else if (stats.getFinalScore(currentNode.getAggregation()) >= currentNode.getThreshold()) {
current = currentNode.getPositive();
}
else {
} else {
current = currentNode.getNegative();
}
@ -63,25 +63,24 @@ public class TreeProcessor{
String current = "start";
double score = 0.0;
while (MatchType.parse(current)==MatchType.UNDEFINED) {
while (MatchType.parse(current) == MatchType.UNDEFINED) {
TreeNodeDef currentNode = config.decisionTree().get(current);
//throw an exception if the node doesn't exist
// throw an exception if the node doesn't exist
if (currentNode == null)
throw new PaceException("The Tree Node doesn't exist: " + current);
TreeNodeStats stats = currentNode.evaluate(doc1, doc2, config);
score = stats.getFinalScore(currentNode.getAggregation());
//if ignoreUndefined=false the miss is considered as undefined
if (!currentNode.isIgnoreUndefined() && stats.undefinedCount()>0) {
// if ignoreUndefined=false the miss is considered as undefined
if (!currentNode.isIgnoreUndefined() && stats.undefinedCount() > 0) {
current = currentNode.getUndefined();
}
//if ignoreUndefined=true the miss is ignored and the score computed anyway
// if ignoreUndefined=true the miss is ignored and the score computed anyway
else if (stats.getFinalScore(currentNode.getAggregation()) >= currentNode.getThreshold()) {
current = currentNode.getPositive();
}
else {
} else {
current = currentNode.getNegative();
}

69
dhp-pace-core/src/main/java/eu/dnetlib/pace/tree/support/TreeStats.java
View File

@ -1,51 +1,52 @@
package eu.dnetlib.pace.tree.support;
import eu.dnetlib.pace.util.PaceException;
import com.fasterxml.jackson.databind.ObjectMapper;
package eu.dnetlib.pace.tree.support;
import java.io.IOException;
import java.util.HashMap;
import java.util.Map;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.pace.util.PaceException;
public class TreeStats {
//<layer_id, <field:comparator, result>>
Map<String, TreeNodeStats> stats;
MatchType result;
// <layer_id, <field:comparator, result>>
Map<String, TreeNodeStats> stats;
MatchType result;
public TreeStats(){
this.stats = new HashMap<>();
this.result = MatchType.NO_MATCH;
}
public TreeStats() {
this.stats = new HashMap<>();
this.result = MatchType.NO_MATCH;
}
public MatchType getResult(){
return this.result;
}
public MatchType getResult() {
return this.result;
}
public void setResult(MatchType result){
this.result = result;
}
public void setResult(MatchType result) {
this.result = result;
}
public Map<String, TreeNodeStats> getStats() {
return stats;
}
public Map<String, TreeNodeStats> getStats() {
return stats;
}
public void setStats(Map<String, TreeNodeStats> stats) {
this.stats = stats;
}
public void setStats(Map<String, TreeNodeStats> stats) {
this.stats = stats;
}
public void addNodeStats(String layerID, TreeNodeStats treeNodeStats){
this.stats.put(layerID, treeNodeStats);
}
@Override
public String toString(){
try {
return new ObjectMapper().writerWithDefaultPrettyPrinter().writeValueAsString(this);
} catch (IOException e) {
throw new PaceException("Impossible to convert to JSON: ", e);
}
}
public void addNodeStats(String layerID, TreeNodeStats treeNodeStats) {
this.stats.put(layerID, treeNodeStats);
}
@Override
public String toString() {
try {
return new ObjectMapper().writerWithDefaultPrettyPrinter().writeValueAsString(this);
} catch (IOException e) {
throw new PaceException("Impossible to convert to JSON: ", e);
}
}
}

299
dhp-pace-core/src/main/java/eu/dnetlib/pace/util/BlockProcessor.java
View File

@ -1,192 +1,217 @@
package eu.dnetlib.pace.util;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.clustering.NGramUtils;
import eu.dnetlib.pace.config.DedupConfig;
import eu.dnetlib.pace.config.WfConfig;
import eu.dnetlib.pace.tree.support.TreeProcessor;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.MapDocument;
import eu.dnetlib.pace.model.MapDocumentComparator;
import java.util.*;
import org.apache.commons.lang3.StringUtils;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import java.util.*;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.clustering.NGramUtils;
import eu.dnetlib.pace.config.DedupConfig;
import eu.dnetlib.pace.config.WfConfig;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.MapDocument;
import eu.dnetlib.pace.model.MapDocumentComparator;
import eu.dnetlib.pace.tree.support.TreeProcessor;
public class BlockProcessor {
public static final List<String> accumulators= new ArrayList<>();
public static final List<String> accumulators = new ArrayList<>();
private static final Log log = LogFactory.getLog(BlockProcessor.class);
private static final Log log = LogFactory.getLog(BlockProcessor.class);
private DedupConfig dedupConf;
private DedupConfig dedupConf;
public static void constructAccumulator( final DedupConfig dedupConf) {
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "records per hash key = 1"));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "missing " + dedupConf.getWf().getOrderField()));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), String.format("Skipped records for count(%s) >= %s", dedupConf.getWf().getOrderField(), dedupConf.getWf().getGroupMaxSize())));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "skip list"));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "dedupSimilarity (x2)"));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "d < " + dedupConf.getWf().getThreshold()));
}
public static void constructAccumulator(final DedupConfig dedupConf) {
accumulators.add(String.format("%s::%s", dedupConf.getWf().getEntityType(), "records per hash key = 1"));
accumulators
.add(
String
.format(
"%s::%s", dedupConf.getWf().getEntityType(), "missing " + dedupConf.getWf().getOrderField()));
accumulators
.add(
String
.format(
"%s::%s", dedupConf.getWf().getEntityType(),
String
.format(
"Skipped records for count(%s) >= %s", dedupConf.getWf().getOrderField(),
dedupConf.getWf().getGroupMaxSize())));
accumulators.add(String.format("%s::%s", dedupConf.getWf().getEntityType(), "skip list"));
accumulators.add(String.format("%s::%s", dedupConf.getWf().getEntityType(), "dedupSimilarity (x2)"));
accumulators
.add(String.format("%s::%s", dedupConf.getWf().getEntityType(), "d < " + dedupConf.getWf().getThreshold()));
}
public BlockProcessor(DedupConfig dedupConf) {
this.dedupConf = dedupConf;
}
public BlockProcessor(DedupConfig dedupConf) {
this.dedupConf = dedupConf;
}
public void processSortedBlock(final String key, final List<MapDocument> documents, final Reporter context) {
if (documents.size() > 1) {
public void processSortedBlock(final String key, final List<MapDocument> documents, final Reporter context) {
if (documents.size() > 1) {
// log.info("reducing key: '" + key + "' records: " + q.size());
process(prepare(documents), context);
process(prepare(documents), context);
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "records per hash key = 1", 1);
}
}
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "records per hash key = 1", 1);
}
}
public void process(final String key, final Iterable<MapDocument> documents, final Reporter context) {
public void process(final String key, final Iterable<MapDocument> documents, final Reporter context) {
final Queue<MapDocument> q = prepare(documents);
final Queue<MapDocument> q = prepare(documents);
if (q.size() > 1) {
if (q.size() > 1) {
// log.info("reducing key: '" + key + "' records: " + q.size());
process(simplifyQueue(q, key, context), context);
process(simplifyQueue(q, key, context), context);
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "records per hash key = 1", 1);
}
}
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "records per hash key = 1", 1);
}
}
private Queue<MapDocument> prepare(final Iterable<MapDocument> documents) {
final Queue<MapDocument> queue = new PriorityQueue<>(100, new MapDocumentComparator(dedupConf.getWf().getOrderField()));
private Queue<MapDocument> prepare(final Iterable<MapDocument> documents) {
final Queue<MapDocument> queue = new PriorityQueue<>(100,
new MapDocumentComparator(dedupConf.getWf().getOrderField()));
final Set<String> seen = new HashSet<String>();
final int queueMaxSize = dedupConf.getWf().getQueueMaxSize();
final Set<String> seen = new HashSet<String>();
final int queueMaxSize = dedupConf.getWf().getQueueMaxSize();
documents.forEach(doc -> {
if (queue.size() <= queueMaxSize) {
final String id = doc.getIdentifier();
documents.forEach(doc -> {
if (queue.size() <= queueMaxSize) {
final String id = doc.getIdentifier();
if (!seen.contains(id)) {
seen.add(id);
queue.add(doc);
}
}
});
if (!seen.contains(id)) {
seen.add(id);
queue.add(doc);
}
}
});
return queue;
}
return queue;
}
private Queue<MapDocument> simplifyQueue(final Queue<MapDocument> queue, final String ngram, final Reporter context) {
final Queue<MapDocument> q = new LinkedList<>();
private Queue<MapDocument> simplifyQueue(final Queue<MapDocument> queue, final String ngram,
final Reporter context) {
final Queue<MapDocument> q = new LinkedList<>();
String fieldRef = "";
final List<MapDocument> tempResults = Lists.newArrayList();
String fieldRef = "";
final List<MapDocument> tempResults = Lists.newArrayList();
while (!queue.isEmpty()) {
final MapDocument result = queue.remove();
while (!queue.isEmpty()) {
final MapDocument result = queue.remove();
final String orderFieldName = dedupConf.getWf().getOrderField();
final Field orderFieldValue = result.values(orderFieldName);
if (!orderFieldValue.isEmpty()) {
final String field = NGramUtils.cleanupForOrdering(orderFieldValue.stringValue());
if (field.equals(fieldRef)) {
tempResults.add(result);
} else {
populateSimplifiedQueue(q, tempResults, context, fieldRef, ngram);
tempResults.clear();
tempResults.add(result);
fieldRef = field;
}
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "missing " + dedupConf.getWf().getOrderField(), 1);
}
}
populateSimplifiedQueue(q, tempResults, context, fieldRef, ngram);
final String orderFieldName = dedupConf.getWf().getOrderField();
final Field orderFieldValue = result.values(orderFieldName);
if (!orderFieldValue.isEmpty()) {
final String field = NGramUtils.cleanupForOrdering(orderFieldValue.stringValue());
if (field.equals(fieldRef)) {
tempResults.add(result);
} else {
populateSimplifiedQueue(q, tempResults, context, fieldRef, ngram);
tempResults.clear();
tempResults.add(result);
fieldRef = field;
}
} else {
context
.incrementCounter(
dedupConf.getWf().getEntityType(), "missing " + dedupConf.getWf().getOrderField(), 1);
}
}
populateSimplifiedQueue(q, tempResults, context, fieldRef, ngram);
return q;
}
return q;
}
private void populateSimplifiedQueue(final Queue<MapDocument> q,
final List<MapDocument> tempResults,
final Reporter context,
final String fieldRef,
final String ngram) {
WfConfig wf = dedupConf.getWf();
if (tempResults.size() < wf.getGroupMaxSize()) {
q.addAll(tempResults);
} else {
context.incrementCounter(wf.getEntityType(), String.format("Skipped records for count(%s) >= %s", wf.getOrderField(), wf.getGroupMaxSize()), tempResults.size());
private void populateSimplifiedQueue(final Queue<MapDocument> q,
final List<MapDocument> tempResults,
final Reporter context,
final String fieldRef,
final String ngram) {
WfConfig wf = dedupConf.getWf();
if (tempResults.size() < wf.getGroupMaxSize()) {
q.addAll(tempResults);
} else {
context
.incrementCounter(
wf.getEntityType(),
String.format("Skipped records for count(%s) >= %s", wf.getOrderField(), wf.getGroupMaxSize()),
tempResults.size());
// log.info("Skipped field: " + fieldRef + " - size: " + tempResults.size() + " - ngram: " + ngram);
}
}
}
}
private void process(final Queue<MapDocument> queue, final Reporter context) {
private void process(final Queue<MapDocument> queue, final Reporter context) {
while (!queue.isEmpty()) {
while (!queue.isEmpty()) {
final MapDocument pivot = queue.remove();
final String idPivot = pivot.getIdentifier();
final MapDocument pivot = queue.remove();
final String idPivot = pivot.getIdentifier();
WfConfig wf = dedupConf.getWf();
final Field fieldsPivot = pivot.values(wf.getOrderField());
final String fieldPivot = (fieldsPivot == null) || fieldsPivot.isEmpty() ? "" : fieldsPivot.stringValue();
WfConfig wf = dedupConf.getWf();
final Field fieldsPivot = pivot.values(wf.getOrderField());
final String fieldPivot = (fieldsPivot == null) || fieldsPivot.isEmpty() ? "" : fieldsPivot.stringValue();
if (fieldPivot != null) {
int i = 0;
for (final MapDocument curr : queue) {
final String idCurr = curr.getIdentifier();
if (fieldPivot != null) {
int i = 0;
for (final MapDocument curr : queue) {
final String idCurr = curr.getIdentifier();
if (mustSkip(idCurr)) {
if (mustSkip(idCurr)) {
context.incrementCounter(wf.getEntityType(), "skip list", 1);
context.incrementCounter(wf.getEntityType(), "skip list", 1);
break;
}
break;
}
if (i > wf.getSlidingWindowSize()) {
break;
}
if (i > wf.getSlidingWindowSize()) {
break;
}
final Field fieldsCurr = curr.values(wf.getOrderField());
final String fieldCurr = (fieldsCurr == null) || fieldsCurr.isEmpty() ? null : fieldsCurr.stringValue();
final Field fieldsCurr = curr.values(wf.getOrderField());
final String fieldCurr = (fieldsCurr == null) || fieldsCurr.isEmpty() ? null
: fieldsCurr.stringValue();
if (!idCurr.equals(idPivot) && (fieldCurr != null)) {
if (!idCurr.equals(idPivot) && (fieldCurr != null)) {
final TreeProcessor treeProcessor = new TreeProcessor(dedupConf);
final TreeProcessor treeProcessor = new TreeProcessor(dedupConf);
emitOutput(treeProcessor.compare(pivot, curr), idPivot, idCurr, context);
emitOutput(treeProcessor.compare(pivot, curr), idPivot, idCurr, context);
}
}
}
}
}
}
}
}
}
}
private void emitOutput(final boolean result, final String idPivot, final String idCurr, final Reporter context) {
private void emitOutput(final boolean result, final String idPivot, final String idCurr, final Reporter context) {
if (result) {
writeSimilarity(context, idPivot, idCurr);
context.incrementCounter(dedupConf.getWf().getEntityType(), "dedupSimilarity (x2)", 1);
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "d < " + dedupConf.getWf().getThreshold(), 1);
}
}
if (result) {
writeSimilarity(context, idPivot, idCurr);
context.incrementCounter(dedupConf.getWf().getEntityType(), "dedupSimilarity (x2)", 1);
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "d < " + dedupConf.getWf().getThreshold(), 1);
}
}
private boolean mustSkip(final String idPivot) {
return dedupConf.getWf().getSkipList().contains(getNsPrefix(idPivot));
}
private boolean mustSkip(final String idPivot) {
return dedupConf.getWf().getSkipList().contains(getNsPrefix(idPivot));
}
private String getNsPrefix(final String id) {
return StringUtils.substringBetween(id, "|", "::");
}
private String getNsPrefix(final String id) {
return StringUtils.substringBetween(id, "|", "::");
}
private void writeSimilarity(final Reporter context, final String from, final String to) {
final String type = dedupConf.getWf().getEntityType();
private void writeSimilarity(final Reporter context, final String from, final String to) {
final String type = dedupConf.getWf().getEntityType();
context.emit(type, from, to);
context.emit(type, to, from);
}
context.emit(type, from, to);
context.emit(type, to, from);
}
}

386
dhp-pace-core/src/main/java/eu/dnetlib/pace/util/BlockProcessorForTesting.java
View File

@ -1,6 +1,14 @@
package eu.dnetlib.pace.util;
import java.util.*;
import org.apache.commons.lang3.StringUtils;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import com.google.common.collect.Lists;
import eu.dnetlib.pace.clustering.NGramUtils;
import eu.dnetlib.pace.config.DedupConfig;
import eu.dnetlib.pace.config.WfConfig;
@ -9,240 +17,260 @@ import eu.dnetlib.pace.model.MapDocument;
import eu.dnetlib.pace.model.MapDocumentComparator;
import eu.dnetlib.pace.tree.*;
import eu.dnetlib.pace.tree.support.TreeProcessor;
import org.apache.commons.lang3.StringUtils;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import java.util.*;
public class BlockProcessorForTesting {
public static final List<String> accumulators= new ArrayList<>();
public static final List<String> accumulators = new ArrayList<>();
private static final Log log = LogFactory.getLog(eu.dnetlib.pace.util.BlockProcessorForTesting.class);
private static final Log log = LogFactory.getLog(eu.dnetlib.pace.util.BlockProcessorForTesting.class);
private DedupConfig dedupConf;
private DedupConfig dedupConf;
public static void constructAccumulator( final DedupConfig dedupConf) {
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "records per hash key = 1"));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "missing " + dedupConf.getWf().getOrderField()));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), String.format("Skipped records for count(%s) >= %s", dedupConf.getWf().getOrderField(), dedupConf.getWf().getGroupMaxSize())));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "skip list"));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "dedupSimilarity (x2)"));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "d < " + dedupConf.getWf().getThreshold()));
}
public static void constructAccumulator(final DedupConfig dedupConf) {
accumulators.add(String.format("%s::%s", dedupConf.getWf().getEntityType(), "records per hash key = 1"));
accumulators
.add(
String
.format(
"%s::%s", dedupConf.getWf().getEntityType(), "missing " + dedupConf.getWf().getOrderField()));
accumulators
.add(
String
.format(
"%s::%s", dedupConf.getWf().getEntityType(),
String
.format(
"Skipped records for count(%s) >= %s", dedupConf.getWf().getOrderField(),
dedupConf.getWf().getGroupMaxSize())));
accumulators.add(String.format("%s::%s", dedupConf.getWf().getEntityType(), "skip list"));
accumulators.add(String.format("%s::%s", dedupConf.getWf().getEntityType(), "dedupSimilarity (x2)"));
accumulators
.add(String.format("%s::%s", dedupConf.getWf().getEntityType(), "d < " + dedupConf.getWf().getThreshold()));
}
public BlockProcessorForTesting(DedupConfig dedupConf) {
this.dedupConf = dedupConf;
}
public BlockProcessorForTesting(DedupConfig dedupConf) {
this.dedupConf = dedupConf;
}
public void processSortedBlock(final String key, final List<MapDocument> documents, final Reporter context, boolean useTree, boolean noMatch) {
if (documents.size() > 1) {
public void processSortedBlock(final String key, final List<MapDocument> documents, final Reporter context,
boolean useTree, boolean noMatch) {
if (documents.size() > 1) {
// log.info("reducing key: '" + key + "' records: " + q.size());
process(prepare(documents), context, useTree, noMatch);
process(prepare(documents), context, useTree, noMatch);
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "records per hash key = 1", 1);
}
}
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "records per hash key = 1", 1);
}
}
public void process(final String key, final Iterable<MapDocument> documents, final Reporter context, boolean useTree, boolean noMatch) {
public void process(final String key, final Iterable<MapDocument> documents, final Reporter context,
boolean useTree, boolean noMatch) {
final Queue<MapDocument> q = prepare(documents);
final Queue<MapDocument> q = prepare(documents);
if (q.size() > 1) {
if (q.size() > 1) {
// log.info("reducing key: '" + key + "' records: " + q.size());
process(simplifyQueue(q, key, context), context, useTree, noMatch);
process(simplifyQueue(q, key, context), context, useTree, noMatch);
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "records per hash key = 1", 1);
}
}
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "records per hash key = 1", 1);
}
}
private Queue<MapDocument> prepare(final Iterable<MapDocument> documents) {
final Queue<MapDocument> queue = new PriorityQueue<>(100, new MapDocumentComparator(dedupConf.getWf().getOrderField()));
private Queue<MapDocument> prepare(final Iterable<MapDocument> documents) {
final Queue<MapDocument> queue = new PriorityQueue<>(100,
new MapDocumentComparator(dedupConf.getWf().getOrderField()));
final Set<String> seen = new HashSet<String>();
final int queueMaxSize = dedupConf.getWf().getQueueMaxSize();
final Set<String> seen = new HashSet<String>();
final int queueMaxSize = dedupConf.getWf().getQueueMaxSize();
documents.forEach(doc -> {
if (queue.size() <= queueMaxSize) {
final String id = doc.getIdentifier();
documents.forEach(doc -> {
if (queue.size() <= queueMaxSize) {
final String id = doc.getIdentifier();
if (!seen.contains(id)) {
seen.add(id);
queue.add(doc);
}
}
});
if (!seen.contains(id)) {
seen.add(id);
queue.add(doc);
}
}
});
return queue;
}
return queue;
}
private Queue<MapDocument> simplifyQueue(final Queue<MapDocument> queue, final String ngram, final Reporter context) {
final Queue<MapDocument> q = new LinkedList<>();
private Queue<MapDocument> simplifyQueue(final Queue<MapDocument> queue, final String ngram,
final Reporter context) {
final Queue<MapDocument> q = new LinkedList<>();
String fieldRef = "";
final List<MapDocument> tempResults = Lists.newArrayList();
String fieldRef = "";
final List<MapDocument> tempResults = Lists.newArrayList();
while (!queue.isEmpty()) {
final MapDocument result = queue.remove();
while (!queue.isEmpty()) {
final MapDocument result = queue.remove();
final String orderFieldName = dedupConf.getWf().getOrderField();
final Field orderFieldValue = result.values(orderFieldName);
if (!orderFieldValue.isEmpty()) {
final String field = NGramUtils.cleanupForOrdering(orderFieldValue.stringValue());
if (field.equals(fieldRef)) {
tempResults.add(result);
} else {
populateSimplifiedQueue(q, tempResults, context, fieldRef, ngram);
tempResults.clear();
tempResults.add(result);
fieldRef = field;
}
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "missing " + dedupConf.getWf().getOrderField(), 1);
}
}
populateSimplifiedQueue(q, tempResults, context, fieldRef, ngram);
final String orderFieldName = dedupConf.getWf().getOrderField();
final Field orderFieldValue = result.values(orderFieldName);
if (!orderFieldValue.isEmpty()) {
final String field = NGramUtils.cleanupForOrdering(orderFieldValue.stringValue());
if (field.equals(fieldRef)) {
tempResults.add(result);
} else {
populateSimplifiedQueue(q, tempResults, context, fieldRef, ngram);
tempResults.clear();
tempResults.add(result);
fieldRef = field;
}
} else {
context
.incrementCounter(
dedupConf.getWf().getEntityType(), "missing " + dedupConf.getWf().getOrderField(), 1);
}
}
populateSimplifiedQueue(q, tempResults, context, fieldRef, ngram);
return q;
}
return q;
}
private void populateSimplifiedQueue(final Queue<MapDocument> q,
final List<MapDocument> tempResults,
final Reporter context,
final String fieldRef,
final String ngram) {
WfConfig wf = dedupConf.getWf();
if (tempResults.size() < wf.getGroupMaxSize()) {
q.addAll(tempResults);
} else {
context.incrementCounter(wf.getEntityType(), String.format("Skipped records for count(%s) >= %s", wf.getOrderField(), wf.getGroupMaxSize()), tempResults.size());
private void populateSimplifiedQueue(final Queue<MapDocument> q,
final List<MapDocument> tempResults,
final Reporter context,
final String fieldRef,
final String ngram) {
WfConfig wf = dedupConf.getWf();
if (tempResults.size() < wf.getGroupMaxSize()) {
q.addAll(tempResults);
} else {
context
.incrementCounter(
wf.getEntityType(),
String.format("Skipped records for count(%s) >= %s", wf.getOrderField(), wf.getGroupMaxSize()),
tempResults.size());
// log.info("Skipped field: " + fieldRef + " - size: " + tempResults.size() + " - ngram: " + ngram);
}
}
}
}
private void process(final Queue<MapDocument> queue, final Reporter context, boolean useTree, boolean noMatch) {
private void process(final Queue<MapDocument> queue, final Reporter context, boolean useTree, boolean noMatch) {
while (!queue.isEmpty()) {
while (!queue.isEmpty()) {
final MapDocument pivot = queue.remove();
final String idPivot = pivot.getIdentifier();
final MapDocument pivot = queue.remove();
final String idPivot = pivot.getIdentifier();
WfConfig wf = dedupConf.getWf();
final Field fieldsPivot = pivot.values(wf.getOrderField());
final String fieldPivot = (fieldsPivot == null) || fieldsPivot.isEmpty() ? "" : fieldsPivot.stringValue();
WfConfig wf = dedupConf.getWf();
final Field fieldsPivot = pivot.values(wf.getOrderField());
final String fieldPivot = (fieldsPivot == null) || fieldsPivot.isEmpty() ? "" : fieldsPivot.stringValue();
if (fieldPivot != null) {
int i = 0;
for (final MapDocument curr : queue) {
final String idCurr = curr.getIdentifier();
if (fieldPivot != null) {
int i = 0;
for (final MapDocument curr : queue) {
final String idCurr = curr.getIdentifier();
if (mustSkip(idCurr)) {
if (mustSkip(idCurr)) {
context.incrementCounter(wf.getEntityType(), "skip list", 1);
context.incrementCounter(wf.getEntityType(), "skip list", 1);
break;
}
break;
}
if (i > wf.getSlidingWindowSize()) {
break;
}
if (i > wf.getSlidingWindowSize()) {
break;
}
final Field fieldsCurr = curr.values(wf.getOrderField());
final String fieldCurr = (fieldsCurr == null) || fieldsCurr.isEmpty() ? null : fieldsCurr.stringValue();
final Field fieldsCurr = curr.values(wf.getOrderField());
final String fieldCurr = (fieldsCurr == null) || fieldsCurr.isEmpty() ? null
: fieldsCurr.stringValue();
if (!idCurr.equals(idPivot) && (fieldCurr != null)) {
if (!idCurr.equals(idPivot) && (fieldCurr != null)) {
//draws no match relations (test purpose)
if (noMatch) {
emitOutput(!new TreeProcessor(dedupConf).compare(pivot, curr), idPivot, idCurr, context);
}
else {
//use the decision tree implementation or the "normal" implementation of the similarity score (valid only for publications)
if (useTree)
emitOutput(new TreeProcessor(dedupConf).compare(pivot, curr), idPivot, idCurr, context);
else
emitOutput(publicationCompare(pivot, curr, dedupConf), idPivot, idCurr, context);
}
// draws no match relations (test purpose)
if (noMatch) {
emitOutput(!new TreeProcessor(dedupConf).compare(pivot, curr), idPivot, idCurr, context);
} else {
// use the decision tree implementation or the "normal" implementation of the similarity
// score (valid only for publications)
if (useTree)
emitOutput(new TreeProcessor(dedupConf).compare(pivot, curr), idPivot, idCurr, context);
else
emitOutput(publicationCompare(pivot, curr, dedupConf), idPivot, idCurr, context);
}
// if(new TreeProcessor(dedupConf).compare(pivot, curr) != publicationCompare(pivot, curr, dedupConf)) {
// emitOutput(true, idPivot, idCurr, context);
// }
}
}
}
}
}
}
}
}
}
}
protected static boolean compareInstanceType(MapDocument a, MapDocument b, DedupConfig conf) {
Map<String, String> params = new HashMap<>();
InstanceTypeMatch instanceTypeMatch = new InstanceTypeMatch(params);
double compare = instanceTypeMatch.compare(a.getFieldMap().get("instance"), b.getFieldMap().get("instance"), conf);
return compare>=1.0;
}
protected static boolean compareInstanceType(MapDocument a, MapDocument b, DedupConfig conf) {
Map<String, String> params = new HashMap<>();
InstanceTypeMatch instanceTypeMatch = new InstanceTypeMatch(params);
double compare = instanceTypeMatch
.compare(a.getFieldMap().get("instance"), b.getFieldMap().get("instance"), conf);
return compare >= 1.0;
}
private boolean publicationCompare(MapDocument a, MapDocument b, DedupConfig config) {
//if the score gives 1, the publications are equivalent
Map<String, String> params = new HashMap<>();
params.put("jpath_value", "$.value");
params.put("jpath_classid", "$.qualifier.classid");
params.put("mode", "count");
private boolean publicationCompare(MapDocument a, MapDocument b, DedupConfig config) {
// if the score gives 1, the publications are equivalent
Map<String, String> params = new HashMap<>();
params.put("jpath_value", "$.value");
params.put("jpath_classid", "$.qualifier.classid");
params.put("mode", "count");
double score = 0.0;
double score = 0.0;
//levenstein title
LevensteinTitle levensteinTitle = new LevensteinTitle(params);
if(levensteinTitle.compare(a.getFieldMap().get("title"), b.getFieldMap().get("title"), config) >= 0.9) {
score += 0.2;
}
// levenstein title
LevensteinTitle levensteinTitle = new LevensteinTitle(params);
if (levensteinTitle.compare(a.getFieldMap().get("title"), b.getFieldMap().get("title"), config) >= 0.9) {
score += 0.2;
}
//pid
JsonListMatch jsonListMatch = new JsonListMatch(params);
if (jsonListMatch.compare(a.getFieldMap().get("pid"), b.getFieldMap().get("pid"), config) >= 1.0) {
score += 0.5;
}
// pid
JsonListMatch jsonListMatch = new JsonListMatch(params);
if (jsonListMatch.compare(a.getFieldMap().get("pid"), b.getFieldMap().get("pid"), config) >= 1.0) {
score += 0.5;
}
//title version
TitleVersionMatch titleVersionMatch = new TitleVersionMatch(params);
double result1 = titleVersionMatch.compare(a.getFieldMap().get("title"), b.getFieldMap().get("title"), config);
if(result1<0 || result1>=1.0) {
score += 0.1;
}
// title version
TitleVersionMatch titleVersionMatch = new TitleVersionMatch(params);
double result1 = titleVersionMatch.compare(a.getFieldMap().get("title"), b.getFieldMap().get("title"), config);
if (result1 < 0 || result1 >= 1.0) {
score += 0.1;
}
//authors match
params.remove("mode");
AuthorsMatch authorsMatch = new AuthorsMatch(params);
double result2 = authorsMatch.compare(a.getFieldMap().get("authors"), b.getFieldMap().get("authors"), config);
if(result2 <0|| result2>=0.6) {
score += 0.2;
}
// authors match
params.remove("mode");
AuthorsMatch authorsMatch = new AuthorsMatch(params);
double result2 = authorsMatch.compare(a.getFieldMap().get("authors"), b.getFieldMap().get("authors"), config);
if (result2 < 0 || result2 >= 0.6) {
score += 0.2;
}
return score>=0.5;
}
return score >= 0.5;
}
private void emitOutput(final boolean result, final String idPivot, final String idCurr, final Reporter context) {
private void emitOutput(final boolean result, final String idPivot, final String idCurr, final Reporter context) {
if (result) {
writeSimilarity(context, idPivot, idCurr);
context.incrementCounter(dedupConf.getWf().getEntityType(), "dedupSimilarity (x2)", 1);
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "d < " + dedupConf.getWf().getThreshold(), 1);
}
}
if (result) {
writeSimilarity(context, idPivot, idCurr);
context.incrementCounter(dedupConf.getWf().getEntityType(), "dedupSimilarity (x2)", 1);
} else {
context.incrementCounter(dedupConf.getWf().getEntityType(), "d < " + dedupConf.getWf().getThreshold(), 1);
}
}
private boolean mustSkip(final String idPivot) {
return dedupConf.getWf().getSkipList().contains(getNsPrefix(idPivot));
}
private boolean mustSkip(final String idPivot) {
return dedupConf.getWf().getSkipList().contains(getNsPrefix(idPivot));
}
private String getNsPrefix(final String id) {
return StringUtils.substringBetween(id, "|", "::");
}
private String getNsPrefix(final String id) {
return StringUtils.substringBetween(id, "|", "::");
}
private void writeSimilarity(final Reporter context, final String from, final String to) {
final String type = dedupConf.getWf().getEntityType();
private void writeSimilarity(final Reporter context, final String from, final String to) {
final String type = dedupConf.getWf().getEntityType();
context.emit(type, from, to);
}
context.emit(type, from, to);
}
}

15
dhp-pace-core/src/main/java/eu/dnetlib/pace/util/Capitalise.java
View File

@ -1,15 +1,18 @@
package eu.dnetlib.pace.util;
import org.apache.commons.lang3.text.WordUtils;
import com.google.common.base.Function;
import org.apache.commons.lang3.text.WordUtils;
public class Capitalise implements Function<String, String> {
private final char[] DELIM = {' ', '-'};
private final char[] DELIM = {
' ', '-'
};
@Override
public String apply(final String s) {
return WordUtils.capitalize(s.toLowerCase(), DELIM);
}
@Override
public String apply(final String s) {
return WordUtils.capitalize(s.toLowerCase(), DELIM);
}
};

4925
dhp-pace-core/src/main/java/eu/dnetlib/pace/util/DiffPatchMatch.java
View File

File diff suppressed because it is too large Load Diff

1
dhp-pace-core/src/main/java/eu/dnetlib/pace/util/DotAbbreviations.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.pace.util;
import com.google.common.base.Function;

293
dhp-pace-core/src/main/java/eu/dnetlib/pace/util/MapDocumentUtil.java
View File

@ -1,166 +1,177 @@
package eu.dnetlib.pace.util;
import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.jayway.jsonpath.Configuration;
import com.jayway.jsonpath.JsonPath;
import com.jayway.jsonpath.Option;
import eu.dnetlib.pace.config.DedupConfig;
import eu.dnetlib.pace.config.Type;
import eu.dnetlib.pace.model.*;
import net.minidev.json.JSONArray;
package eu.dnetlib.pace.util;
import java.math.BigDecimal;
import java.util.*;
import java.util.function.Predicate;
import java.util.stream.Collectors;
import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.jayway.jsonpath.Configuration;
import com.jayway.jsonpath.JsonPath;
import com.jayway.jsonpath.Option;
import eu.dnetlib.pace.config.DedupConfig;
import eu.dnetlib.pace.config.Type;
import eu.dnetlib.pace.model.*;
import net.minidev.json.JSONArray;
public class MapDocumentUtil {
public static final String URL_REGEX = "^(http|https|ftp)\\://.*";
public static Predicate<String> urlFilter = s -> s.trim().matches(URL_REGEX);
public static final String URL_REGEX = "^(http|https|ftp)\\://.*";
public static Predicate<String> urlFilter = s -> s.trim().matches(URL_REGEX);
public static MapDocument asMapDocumentWithJPath(DedupConfig conf, final String json) {
MapDocument m = new MapDocument();
m.setIdentifier(getJPathString(conf.getWf().getIdPath(), json));
Map<String, Field> stringField = new HashMap<>();
conf.getPace().getModel().forEach(fdef -> {
switch (fdef.getType()) {
case String:
case Int:
stringField.put(fdef.getName(), new FieldValueImpl(fdef.getType(), fdef.getName(), truncateValue(getJPathString(fdef.getPath(), json), fdef.getLength())));
break;
case URL:
String uv = getJPathString(fdef.getPath(), json);
if (!urlFilter.test(uv)) uv = "";
stringField.put(fdef.getName(), new FieldValueImpl(fdef.getType(), fdef.getName(), uv));
break;
case List:
case JSON:
FieldListImpl fi = new FieldListImpl(fdef.getName(), fdef.getType());
truncateList(getJPathList(fdef.getPath(), json, fdef.getType()), fdef.getSize())
.stream()
.map(item -> new FieldValueImpl(Type.String, fdef.getName(), item))
.forEach(fi::add);
stringField.put(fdef.getName(), fi);
break;
case DoubleArray:
stringField.put(
fdef.getName(),
new FieldValueImpl(Type.DoubleArray,
fdef.getName(),
getJPathArray(fdef.getPath(), json))
);
break;
case StringConcat:
String[] jpaths = fdef.getPath().split("\\|\\|\\|");
stringField.put(
fdef.getName(),
new FieldValueImpl(Type.String,
fdef.getName(),
truncateValue(Arrays.stream(jpaths).map(jpath -> getJPathString(jpath, json)).collect(Collectors.joining(" ")),
fdef.getLength())
)
);
break;
}
});
m.setFieldMap(stringField);
return m;
}
public static MapDocument asMapDocumentWithJPath(DedupConfig conf, final String json) {
MapDocument m = new MapDocument();
m.setIdentifier(getJPathString(conf.getWf().getIdPath(), json));
Map<String, Field> stringField = new HashMap<>();
conf.getPace().getModel().forEach(fdef -> {
switch (fdef.getType()) {
case String:
case Int:
stringField
.put(
fdef.getName(), new FieldValueImpl(fdef.getType(), fdef.getName(),
truncateValue(getJPathString(fdef.getPath(), json), fdef.getLength())));
break;
case URL:
String uv = getJPathString(fdef.getPath(), json);
if (!urlFilter.test(uv))
uv = "";
stringField.put(fdef.getName(), new FieldValueImpl(fdef.getType(), fdef.getName(), uv));
break;
case List:
case JSON:
FieldListImpl fi = new FieldListImpl(fdef.getName(), fdef.getType());
truncateList(getJPathList(fdef.getPath(), json, fdef.getType()), fdef.getSize())
.stream()
.map(item -> new FieldValueImpl(Type.String, fdef.getName(), item))
.forEach(fi::add);
stringField.put(fdef.getName(), fi);
break;
case DoubleArray:
stringField
.put(
fdef.getName(),
new FieldValueImpl(Type.DoubleArray,
fdef.getName(),
getJPathArray(fdef.getPath(), json)));
break;
case StringConcat:
String[] jpaths = fdef.getPath().split("\\|\\|\\|");
stringField
.put(
fdef.getName(),
new FieldValueImpl(Type.String,
fdef.getName(),
truncateValue(
Arrays
.stream(jpaths)
.map(jpath -> getJPathString(jpath, json))
.collect(Collectors.joining(" ")),
fdef.getLength())));
break;
}
});
m.setFieldMap(stringField);
return m;
}
public static List<String> getJPathList(String path, String json, Type type) {
if (type == Type.List)
return JsonPath.using(Configuration.defaultConfiguration().addOptions(Option.ALWAYS_RETURN_LIST, Option.SUPPRESS_EXCEPTIONS)).parse(json).read(path);
Object jresult;
List<String> result = new ArrayList<>();
try {
jresult = JsonPath.read(json, path);
} catch (Throwable e) {
return result;
}
if (jresult instanceof JSONArray) {
public static List<String> getJPathList(String path, String json, Type type) {
if (type == Type.List)
return JsonPath
.using(
Configuration
.defaultConfiguration()
.addOptions(Option.ALWAYS_RETURN_LIST, Option.SUPPRESS_EXCEPTIONS))
.parse(json)
.read(path);
Object jresult;
List<String> result = new ArrayList<>();
try {
jresult = JsonPath.read(json, path);
} catch (Throwable e) {
return result;
}
if (jresult instanceof JSONArray) {
((JSONArray) jresult).forEach(it -> {
((JSONArray) jresult).forEach(it -> {
try {
result.add(new ObjectMapper().writeValueAsString(it));
} catch (JsonProcessingException e) {
try {
result.add(new ObjectMapper().writeValueAsString(it));
} catch (JsonProcessingException e) {
}
}
);
return result;
}
}
});
return result;
}
if (jresult instanceof LinkedHashMap) {
try {
result.add(new ObjectMapper().writeValueAsString(jresult));
} catch (JsonProcessingException e) {
if (jresult instanceof LinkedHashMap) {
try {
result.add(new ObjectMapper().writeValueAsString(jresult));
} catch (JsonProcessingException e) {
}
return result;
}
if (jresult instanceof String) {
result.add((String) jresult);
}
return result;
}
}
return result;
}
if (jresult instanceof String) {
result.add((String) jresult);
}
return result;
}
public static String getJPathString(final String jsonPath, final String json) {
try {
Object o = JsonPath.read(json, jsonPath);
if (o instanceof String)
return (String) o;
if (o instanceof JSONArray && ((JSONArray) o).size() > 0)
return (String) ((JSONArray) o).get(0);
return "";
} catch (Exception e) {
return "";
}
}
public static String getJPathString(final String jsonPath, final String json) {
try {
Object o = JsonPath.read(json, jsonPath);
if (o instanceof String)
return (String)o;
if (o instanceof JSONArray && ((JSONArray)o).size()>0)
return (String)((JSONArray)o).get(0);
return "";
} catch (Exception e) {
return "";
}
}
public static double[] getJPathArray(final String jsonPath, final String json) {
try {
Object o = JsonPath.read(json, jsonPath);
if (o instanceof double[])
return (double[]) o;
if (o instanceof JSONArray) {
Object[] objects = ((JSONArray) o).toArray();
double[] array = new double[objects.length];
for (int i = 0; i < objects.length; i++) {
if (objects[i] instanceof BigDecimal)
array[i] = ((BigDecimal) objects[i]).doubleValue();
else
array[i] = (double) objects[i];
}
return array;
}
return new double[0];
} catch (Exception e) {
e.printStackTrace();
return new double[0];
}
}
public static double[] getJPathArray(final String jsonPath, final String json) {
try {
Object o = JsonPath.read(json, jsonPath);
if (o instanceof double[])
return (double[]) o;
if (o instanceof JSONArray) {
Object[] objects = ((JSONArray) o).toArray();
double[] array = new double[objects.length];
for (int i = 0; i < objects.length; i++) {
if (objects[i] instanceof BigDecimal)
array[i] = ((BigDecimal)objects[i]).doubleValue();
else
array[i] = (double) objects[i];
}
return array;
}
return new double[0];
}
catch (Exception e) {
e.printStackTrace();
return new double[0];
}
}
public static String truncateValue(String value, int length) {
if (value == null)
return "";
if (length == -1 || length > value.length())
return value;
public static String truncateValue(String value, int length) {
if (value == null)
return "";
return value.substring(0, length);
}
if (length == -1 || length > value.length())
return value;
public static List<String> truncateList(List<String> list, int size) {
if (size == -1 || size > list.size())
return list;
return value.substring(0, length);
}
public static List<String> truncateList(List<String> list, int size) {
if (size == -1 || size > list.size())
return list;
return list.subList(0, size);
}
return list.subList(0, size);
}
}

13
dhp-pace-core/src/main/java/eu/dnetlib/pace/util/PaceException.java
View File

@ -1,13 +1,14 @@
package eu.dnetlib.pace.util;
public class PaceException extends RuntimeException {
public PaceException(String s, Throwable e){
super(s, e);
}
public PaceException(String s, Throwable e) {
super(s, e);
}
public PaceException(String s){
super(s);
}
public PaceException(String s) {
super(s);
}
}

76
dhp-pace-core/src/main/java/eu/dnetlib/pace/util/PaceResolver.java
View File

@ -1,49 +1,61 @@
package eu.dnetlib.pace.util;
import eu.dnetlib.pace.clustering.ClusteringClass;
import eu.dnetlib.pace.clustering.ClusteringFunction;
import eu.dnetlib.pace.tree.support.Comparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
import org.reflections.Reflections;
package eu.dnetlib.pace.util;
import java.io.Serializable;
import java.lang.reflect.InvocationTargetException;
import java.util.Map;
import java.util.stream.Collectors;
import org.reflections.Reflections;
import eu.dnetlib.pace.clustering.ClusteringClass;
import eu.dnetlib.pace.clustering.ClusteringFunction;
import eu.dnetlib.pace.tree.support.Comparator;
import eu.dnetlib.pace.tree.support.ComparatorClass;
public class PaceResolver implements Serializable {
public static final Reflections CLUSTERING_RESOLVER = new Reflections("eu.dnetlib.pace.clustering");
public static final Reflections COMPARATOR_RESOLVER = new Reflections("eu.dnetlib.pace.tree");
public static final Reflections CLUSTERING_RESOLVER = new Reflections("eu.dnetlib.pace.clustering");
public static final Reflections COMPARATOR_RESOLVER = new Reflections("eu.dnetlib.pace.tree");
private final Map<String, Class<ClusteringFunction>> clusteringFunctions;
private final Map<String, Class<Comparator>> comparators;
private final Map<String, Class<ClusteringFunction>> clusteringFunctions;
private final Map<String, Class<Comparator>> comparators;
public PaceResolver() {
public PaceResolver() {
this.clusteringFunctions = CLUSTERING_RESOLVER.getTypesAnnotatedWith(ClusteringClass.class).stream()
.filter(ClusteringFunction.class::isAssignableFrom)
.collect(Collectors.toMap(cl -> cl.getAnnotation(ClusteringClass.class).value(), cl -> (Class<ClusteringFunction>)cl));
this.clusteringFunctions = CLUSTERING_RESOLVER
.getTypesAnnotatedWith(ClusteringClass.class)
.stream()
.filter(ClusteringFunction.class::isAssignableFrom)
.collect(
Collectors
.toMap(
cl -> cl.getAnnotation(ClusteringClass.class).value(), cl -> (Class<ClusteringFunction>) cl));
this.comparators = COMPARATOR_RESOLVER.getTypesAnnotatedWith(ComparatorClass.class).stream()
.filter(Comparator.class::isAssignableFrom)
.collect(Collectors.toMap(cl -> cl.getAnnotation(ComparatorClass.class).value(), cl -> (Class<Comparator>)cl));
}
this.comparators = COMPARATOR_RESOLVER
.getTypesAnnotatedWith(ComparatorClass.class)
.stream()
.filter(Comparator.class::isAssignableFrom)
.collect(
Collectors.toMap(cl -> cl.getAnnotation(ComparatorClass.class).value(), cl -> (Class<Comparator>) cl));
}
public ClusteringFunction getClusteringFunction(String name, Map<String, Integer> params) throws PaceException {
try {
return clusteringFunctions.get(name).getDeclaredConstructor(Map.class).newInstance(params);
} catch (InstantiationException | IllegalAccessException | InvocationTargetException | NoSuchMethodException e) {
throw new PaceException(name + " not found ", e);
}
}
public ClusteringFunction getClusteringFunction(String name, Map<String, Integer> params) throws PaceException {
try {
return clusteringFunctions.get(name).getDeclaredConstructor(Map.class).newInstance(params);
} catch (InstantiationException | IllegalAccessException | InvocationTargetException
| NoSuchMethodException e) {
throw new PaceException(name + " not found ", e);
}
}
public Comparator getComparator(String name, Map<String, String> params) throws PaceException {
try {
return comparators.get(name).getDeclaredConstructor(Map.class).newInstance(params);
} catch (InstantiationException | IllegalAccessException | InvocationTargetException | NoSuchMethodException | NullPointerException e) {
throw new PaceException(name + " not found ", e);
}
}
public Comparator getComparator(String name, Map<String, String> params) throws PaceException {
try {
return comparators.get(name).getDeclaredConstructor(Map.class).newInstance(params);
} catch (InstantiationException | IllegalAccessException | InvocationTargetException | NoSuchMethodException
| NullPointerException e) {
throw new PaceException(name + " not found ", e);
}
}
}

6
dhp-pace-core/src/main/java/eu/dnetlib/pace/util/Reporter.java
View File

@ -1,11 +1,11 @@
package eu.dnetlib.pace.util;
package eu.dnetlib.pace.util;
import java.io.Serializable;
public interface Reporter extends Serializable {
void incrementCounter(String counterGroup, String counterName, long delta);
void incrementCounter(String counterGroup, String counterName, long delta);
void emit(String type, String from, String to);
void emit(String type, String from, String to);
}

23
dhp-pace-core/src/test/java/eu/dnetlib/pace/AbstractPaceTest.java
View File

@ -1,11 +1,5 @@
package eu.dnetlib.pace;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Type;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldListImpl;
import eu.dnetlib.pace.model.FieldValueImpl;
import org.apache.commons.io.IOUtils;
package eu.dnetlib.pace;
import java.io.IOException;
import java.io.StringWriter;
@ -13,6 +7,14 @@ import java.nio.charset.StandardCharsets;
import java.util.List;
import java.util.stream.Collectors;
import org.apache.commons.io.IOUtils;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Type;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldListImpl;
import eu.dnetlib.pace.model.FieldValueImpl;
public abstract class AbstractPaceTest extends AbstractPaceFunctions {
protected String readFromClasspath(final String filename) {
@ -41,9 +43,12 @@ public abstract class AbstractPaceTest extends AbstractPaceFunctions {
return new FieldValueImpl(Type.DoubleArray, "array", a);
}
protected Field createFieldList(List<String> strings, String fieldName){
protected Field createFieldList(List<String> strings, String fieldName) {
List<FieldValueImpl> fieldValueStream = strings.stream().map(s -> new FieldValueImpl(Type.String, fieldName, s)).collect(Collectors.toList());
List<FieldValueImpl> fieldValueStream = strings
.stream()
.map(s -> new FieldValueImpl(Type.String, fieldName, s))
.collect(Collectors.toList());
FieldListImpl a = new FieldListImpl();
a.addAll(fieldValueStream);

31
dhp-pace-core/src/test/java/eu/dnetlib/pace/clustering/ClusteringFunctionTest.java
View File

@ -1,17 +1,20 @@
package eu.dnetlib.pace.clustering;
import com.google.common.collect.Lists;
import com.google.common.collect.Maps;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.AbstractPaceTest;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.DedupConfig;
import org.junit.jupiter.api.*;
package eu.dnetlib.pace.clustering;
import java.util.Map;
import java.util.Set;
import java.util.stream.Collectors;
import org.junit.jupiter.api.*;
import com.google.common.collect.Lists;
import com.google.common.collect.Maps;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.AbstractPaceTest;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.DedupConfig;
public class ClusteringFunctionTest extends AbstractPaceTest {
private static Map<String, Integer> params;
@ -20,7 +23,11 @@ public class ClusteringFunctionTest extends AbstractPaceTest {
@BeforeAll
public static void setUp() throws Exception {
params = Maps.newHashMap();
conf = DedupConfig.load(AbstractPaceFunctions.readFromClasspath("/eu/dnetlib/pace/config/organization.current.conf.json", ClusteringFunctionTest.class));
conf = DedupConfig
.load(
AbstractPaceFunctions
.readFromClasspath(
"/eu/dnetlib/pace/config/organization.current.conf.json", ClusteringFunctionTest.class));
}
@Test
@ -210,7 +217,7 @@ public class ClusteringFunctionTest extends AbstractPaceTest {
}
@Test
public void testPersonClustering(){
public void testPersonClustering() {
final ClusteringFunction cf = new PersonClustering(params);
final String s = "Abd-Alla, Abo-el-nour N.";
@ -224,7 +231,7 @@ public class ClusteringFunctionTest extends AbstractPaceTest {
}
@Test
public void testPersonHash(){
public void testPersonHash() {
final ClusteringFunction cf = new PersonHash(params);
final String s = "Manghi, Paolo";
@ -238,7 +245,7 @@ public class ClusteringFunctionTest extends AbstractPaceTest {
}
@Test
public void testLastNameFirstInitial(){
public void testLastNameFirstInitial() {
final ClusteringFunction cf = new LastNameFirstInitial(params);
final String s = "LI Yonghong";

Some files were not shown because too many files have changed in this diff Show More