dimitris.pierrakos
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

code formatting

This commit is contained in:
Claudio Atzori 2020-10-07 13:14:31 +02:00
parent 1abcabb6e6
commit 8958f20813
7 changed files with 72 additions and 68 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

44
dhp-common/src/main/java/eu/dnetlib/dhp/schema/oaf/utils/IdentifierFactory.java
View File

@ -1,13 +1,14 @@
package eu.dnetlib.dhp.schema.oaf.utils;
import java.io.Serializable;
import java.util.Objects;
import org.apache.commons.lang.StringUtils;
import eu.dnetlib.dhp.schema.oaf.OafEntity;
import eu.dnetlib.dhp.schema.oaf.StructuredProperty;
import eu.dnetlib.dhp.utils.DHPUtils;
import org.apache.commons.lang.StringUtils;
import java.io.Serializable;
import java.util.Objects;
/**
* Factory class for OpenAIRE identifiers in the Graph
@ -16,7 +17,8 @@ public class IdentifierFactory implements Serializable {
public static final String ID_SEPARATOR = "::";
public static final String ID_PREFIX_SEPARATOR = "|";
public final static String ID_REGEX = "^[0-9][0-9]\\"+ID_PREFIX_SEPARATOR+".{12}"+ID_SEPARATOR+"[a-zA-Z0-9]{32}$";
public final static String ID_REGEX = "^[0-9][0-9]\\" + ID_PREFIX_SEPARATOR + ".{12}" + ID_SEPARATOR
+ "[a-zA-Z0-9]{32}$";
public static final int ID_PREFIX_LEN = 12;
public static <T extends OafEntity> String createIdentifier(T entity) {
@ -26,18 +28,18 @@ public class IdentifierFactory implements Serializable {
}
return entity
.getPid()
.stream()
.filter(s -> Objects.nonNull(s.getQualifier()))
.filter(s -> PidType.isValid(s.getQualifier().getClassid()))
.min(new PidComparator<>(entity))
.map(s -> idFromPid(entity, s))
.map(IdentifierFactory::verifyIdSyntax)
.orElseGet(entity::getId);
.getPid()
.stream()
.filter(s -> Objects.nonNull(s.getQualifier()))
.filter(s -> PidType.isValid(s.getQualifier().getClassid()))
.min(new PidComparator<>(entity))
.map(s -> idFromPid(entity, s))
.map(IdentifierFactory::verifyIdSyntax)
.orElseGet(entity::getId);
}
private static String verifyIdSyntax(String s) {
if(StringUtils.isBlank(s) || !s.matches(ID_REGEX)) {
if (StringUtils.isBlank(s) || !s.matches(ID_REGEX)) {
throw new RuntimeException(String.format("malformed id: '%s'", s));
} else {
return s;
@ -46,16 +48,16 @@ public class IdentifierFactory implements Serializable {
private static <T extends OafEntity> String idFromPid(T entity, StructuredProperty s) {
return new StringBuilder()
.append(StringUtils.substringBefore(entity.getId(), ID_PREFIX_SEPARATOR))
.append(ID_PREFIX_SEPARATOR)
.append(createPrefix(s.getQualifier().getClassid()))
.append(ID_SEPARATOR)
.append(DHPUtils.md5(normalizePidValue(s.getValue())))
.toString();
.append(StringUtils.substringBefore(entity.getId(), ID_PREFIX_SEPARATOR))
.append(ID_PREFIX_SEPARATOR)
.append(createPrefix(s.getQualifier().getClassid()))
.append(ID_SEPARATOR)
.append(DHPUtils.md5(normalizePidValue(s.getValue())))
.toString();
}
private static String normalizePidValue(String value) {
//TODO more aggressive cleaning? keep only alphanum and punctuation?
// TODO more aggressive cleaning? keep only alphanum and punctuation?
return value.toLowerCase().replaceAll(" ", "");
}

13
dhp-common/src/main/java/eu/dnetlib/dhp/schema/oaf/utils/PidType.java
View File

@ -1,17 +1,18 @@
package eu.dnetlib.dhp.schema.oaf.utils;
import org.apache.commons.lang3.EnumUtils;
public enum PidType {
// Result
doi, pmid, pmc, handle, arXiv, NCID, GBIF, nct, pdb,
// Result
doi, pmid, pmc, handle, arXiv, NCID, GBIF, nct, pdb,
// Organization
// Organization
GRID, mag_id, urn;
public static boolean isValid(String type) {
return EnumUtils.isValidEnum(PidType.class, type);
}
public static boolean isValid(String type) {
return EnumUtils.isValidEnum(PidType.class, type);
}
}

62
dhp-common/src/test/java/eu/dnetlib/dhp/schema/oaf/utils/IdentifierFactoryTest.java
View File

@ -1,43 +1,47 @@
package eu.dnetlib.dhp.schema.oaf.utils;
import com.fasterxml.jackson.databind.DeserializationFeature;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.schema.oaf.Publication;
import eu.dnetlib.dhp.utils.DHPUtils;
import org.apache.commons.io.IOUtils;
import org.junit.jupiter.api.Test;
import java.io.IOException;
import static org.junit.jupiter.api.Assertions.*;
import java.io.IOException;
import org.apache.commons.io.IOUtils;
import org.junit.jupiter.api.Test;
import com.fasterxml.jackson.databind.DeserializationFeature;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.schema.oaf.Publication;
import eu.dnetlib.dhp.utils.DHPUtils;
public class IdentifierFactoryTest {
private static ObjectMapper OBJECT_MAPPER = new ObjectMapper().configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false);
private static ObjectMapper OBJECT_MAPPER = new ObjectMapper()
.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false);
@Test
public void testCreateIdentifierForPublication() throws IOException {
@Test
public void testCreateIdentifierForPublication() throws IOException {
verifyIdentifier("publication_doi.json", "50|doi_________::" + DHPUtils.md5("10.1016/j.cmet.2011.03.013"));
verifyIdentifier("publication_pmc.json", "50|pmc_________::" + DHPUtils.md5("21459329"));
verifyIdentifier("publication_urn.json", "50|urn_________::" + DHPUtils.md5("urn:nbn:nl:ui:29-f3ed5f9e-edf6-457e-8848-61b58a4075e2"));
verifyIdentifier("publication_doi.json", "50|doi_________::" + DHPUtils.md5("10.1016/j.cmet.2011.03.013"));
verifyIdentifier("publication_pmc.json", "50|pmc_________::" + DHPUtils.md5("21459329"));
verifyIdentifier(
"publication_urn.json",
"50|urn_________::" + DHPUtils.md5("urn:nbn:nl:ui:29-f3ed5f9e-edf6-457e-8848-61b58a4075e2"));
final String defaultID = "50|DansKnawCris::0829b5191605bdbea36d6502b8c1ce1f";
verifyIdentifier("publication_3.json", defaultID);
verifyIdentifier("publication_4.json", defaultID);
verifyIdentifier("publication_5.json", defaultID);
}
final String defaultID = "50|DansKnawCris::0829b5191605bdbea36d6502b8c1ce1f";
verifyIdentifier("publication_3.json", defaultID);
verifyIdentifier("publication_4.json", defaultID);
verifyIdentifier("publication_5.json", defaultID);
}
protected void verifyIdentifier(String filename, String expectedID) throws IOException {
final String json = IOUtils.toString(getClass().getResourceAsStream(filename));
final Publication pub = OBJECT_MAPPER.readValue(json, Publication.class);
protected void verifyIdentifier(String filename, String expectedID) throws IOException {
final String json = IOUtils.toString(getClass().getResourceAsStream(filename));
final Publication pub = OBJECT_MAPPER.readValue(json, Publication.class);
String id = IdentifierFactory.createIdentifier(pub);
assertNotNull(id);
assertEquals(expectedID, id);
}
String id = IdentifierFactory.createIdentifier(pub);
assertNotNull(id);
assertEquals(expectedID, id);
}
}

1
dhp-schemas/src/main/java/eu/dnetlib/dhp/schema/oaf/H2020Programme.java
View File

@ -10,7 +10,6 @@ import java.util.Objects;
* - private String description to store the description of the programme
*/
public class H2020Programme implements Serializable {
private String code;
private String description;

7
dhp-schemas/src/main/java/eu/dnetlib/dhp/schema/oaf/Result.java
View File

@ -1,14 +1,14 @@
package eu.dnetlib.dhp.schema.oaf;
import eu.dnetlib.dhp.schema.common.LicenseComparator;
import java.io.Serializable;
import java.util.ArrayList;
import java.util.Comparator;
import java.util.List;
import java.util.stream.Collectors;
import eu.dnetlib.dhp.schema.common.LicenseComparator;
public class Result extends OafEntity implements Serializable {
private List<Measure> measures;
@ -247,7 +247,8 @@ public class Result extends OafEntity implements Serializable {
instance = mergeLists(instance, r.getInstance());
if (r.getBestaccessright() != null && new LicenseComparator().compare(r.getBestaccessright(), bestaccessright) < 0)
if (r.getBestaccessright() != null
&& new LicenseComparator().compare(r.getBestaccessright(), bestaccessright) < 0)
bestaccessright = r.getBestaccessright();
if (r.getResulttype() != null && compareTrust(this, r) < 0)

10
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/project/PrepareProgramme.java
View File

@ -177,14 +177,12 @@ public class PrepareProgramme {
prepareClassification(h2020Programmes);
h2020Programmes.map(csvProgramme -> OBJECT_MAPPER.writeValueAsString(csvProgramme))
.saveAsTextFile(outputPath);
h2020Programmes
.map(csvProgramme -> OBJECT_MAPPER.writeValueAsString(csvProgramme))
.saveAsTextFile(outputPath);
}
private static void prepareClassification(JavaRDD<CSVProgramme> h2020Programmes) {
Object[] codedescription = h2020Programmes
.map(value -> new Tuple2<>(value.getCode(), value.getTitle()))
@ -255,7 +253,7 @@ public class PrepareProgramme {
}
h2020Programmes.foreach(csvProgramme -> {
if (!csvProgramme.getCode().endsWith(".") && !csvProgramme.getCode().contains("Euratom")
&& !csvProgramme.getCode().equals("H2020-EC"))
&& !csvProgramme.getCode().equals("H2020-EC"))
csvProgramme.setClassification(map.get(csvProgramme.getCode() + "."));
else
csvProgramme.setClassification(map.get(csvProgramme.getCode()));

1
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/project/ProjectSubset.java
View File

@ -10,7 +10,6 @@ public class ProjectSubset implements Serializable {
private String code;
public String getCode() {
return code;
}