orcid / orcid_pending cleaning backported from master branch

This commit is contained in:
Claudio Atzori 2021-06-14 09:40:50 +02:00
parent dd19c4ac5a
commit 2039bb9f5f
6 changed files with 67 additions and 189 deletions

View File

@ -26,8 +26,9 @@ import eu.dnetlib.dhp.schema.oaf.*;
public class GraphCleaningFunctions extends CleaningFunctions {
public static final String ORCID_CLEANING_REGEX = ".*([0-9]{4}).*[-–—−=].*([0-9]{4}).*[-–—−=].*([0-9]{4}).*[-–—−=].*([0-9x]{4})";
public static final int ORCID_LEN = 19;
public static final String CLEANING_REGEX = "(?:\\n|\\r|\\t)";
public static final String ORCID_PREFIX_REGEX = "^http(s?):\\/\\/orcid\\.org\\/";
public static final String INVALID_AUTHOR_REGEX = ".*deactivated.*";
public static final String TITLE_FILTER_REGEX = "[.*test.*\\W\\d]";
public static final int TITLE_FILTER_RESIDUAL_LENGTH = 10;
@ -281,7 +282,27 @@ public class GraphCleaningFunctions extends CleaningFunctions {
}
}
if (Objects.nonNull(r.getAuthor())) {
final List<Author> authors = Lists.newArrayList();
r
.setAuthor(
r
.getAuthor()
.stream()
.filter(a -> Objects.nonNull(a))
.filter(a -> StringUtils.isNotBlank(a.getFullname()))
.filter(a -> StringUtils.isNotBlank(a.getFullname().replaceAll("[\\W]", "")))
.collect(Collectors.toList()));
boolean nullRank = r
.getAuthor()
.stream()
.anyMatch(a -> Objects.isNull(a.getRank()));
if (nullRank) {
int i = 1;
for (Author author : r.getAuthor()) {
author.setRank(i++);
}
}
for (Author a : r.getAuthor()) {
if (Objects.isNull(a.getPid())) {
a.setPid(Lists.newArrayList());
@ -295,40 +316,52 @@ public class GraphCleaningFunctions extends CleaningFunctions {
.filter(p -> Objects.nonNull(p.getQualifier()))
.filter(p -> StringUtils.isNotBlank(p.getValue()))
.map(p -> {
p.setValue(p.getValue().trim().replaceAll(ORCID_PREFIX_REGEX, ""));
// hack to distinguish orcid from orcid_pending
String pidProvenance = Optional
.ofNullable(p.getDataInfo())
.map(
d -> Optional
.ofNullable(d.getProvenanceaction())
.map(Qualifier::getClassid)
.orElse(""))
.orElse("");
if (p
.getQualifier()
.getClassid()
.toLowerCase()
.contains(ModelConstants.ORCID)) {
if (pidProvenance
.equals(ModelConstants.SYSIMPORT_CROSSWALK_ENTITYREGISTRY)) {
p.getQualifier().setClassid(ModelConstants.ORCID);
} else {
p.getQualifier().setClassid(ModelConstants.ORCID_PENDING);
}
final String orcid = p
.getValue()
.trim()
.toLowerCase()
.replaceAll(ORCID_CLEANING_REGEX, "$1-$2-$3-$4");
if (orcid.length() == ORCID_LEN) {
p.setValue(orcid);
} else {
p.setValue("");
}
}
return p;
})
.filter(p -> StringUtils.isNotBlank(p.getValue()))
.collect(
Collectors
.toMap(
StructuredProperty::getValue, Function.identity(), (p1, p2) -> p1,
p -> p.getQualifier().getClassid() + p.getValue(),
Function.identity(),
(p1, p2) -> p1,
LinkedHashMap::new))
.values()
.stream()
.collect(Collectors.toList()));
}
if (StringUtils.isBlank(a.getFullname())) {
if (StringUtils.isNotBlank(a.getName()) && StringUtils.isNotBlank(a.getSurname())) {
a.setFullname(a.getSurname() + ", " + a.getName());
}
}
if (StringUtils.isNotBlank(a.getFullname()) && isValidAuthorName(a)) {
authors.add(a);
}
}
boolean nullRank = authors
.stream()
.anyMatch(a -> Objects.isNull(a.getRank()));
if (nullRank) {
int i = 1;
for (Author author : authors) {
author.setRank(i++);
}
}
r.setAuthor(authors);
}
if (value instanceof Publication) {

View File

@ -131,6 +131,11 @@ public class OafMapperUtilsTest {
}
@Test
public void testDate() {
System.out.println(GraphCleaningFunctions.cleanDate("23-FEB-1998"));
}
@Test
public void testMergePubs() throws IOException {
Publication p1 = read("publication_1.json", Publication.class);

View File

@ -76,7 +76,9 @@ public abstract class AbstractMdRecordToOafMapper {
protected static final String DATACITE_SCHEMA_KERNEL_3 = "http://datacite.org/schema/kernel-3";
protected static final String DATACITE_SCHEMA_KERNEL_3_SLASH = "http://datacite.org/schema/kernel-3/";
protected static final Qualifier ORCID_PID_TYPE = qualifier(
"ORCID", "Open Researcher and Contributor ID", DNET_PID_TYPES, DNET_PID_TYPES);
ModelConstants.ORCID_PENDING,
ModelConstants.ORCID_CLASSNAME,
DNET_PID_TYPES, DNET_PID_TYPES);
protected static final Qualifier MAG_PID_TYPE = qualifier(
"MAGIdentifier", "Microsoft Academic Graph Identifier", DNET_PID_TYPES, DNET_PID_TYPES);

View File

@ -102,11 +102,6 @@ public class MigrateHdfsMdstoresApplication extends AbstractMigrationApplication
.mapToPair(xml -> new Tuple2<>(new Text(UUID.randomUUID() + ":" + type), new Text(xml)))
// .coalesce(1)
.saveAsHadoopFile(outputPath, Text.class, Text.class, SequenceFileOutputFormat.class, GzipCodec.class);
/*
* .foreach(xml -> { try { writer.append(new Text(UUID.randomUUID() + ":" + type), new Text(xml)); } catch
* (final Exception e) { throw new RuntimeException(e); } });
*/
}
private static String enrichRecord(final Row r) {

View File

@ -1,157 +0,0 @@
<workflow-app name="Test Import of Hdfs Stores" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>graphOutputPath</name>
<description>the target path to store raw graph</description>
</property>
<property>
<name>contentPath</name>
<description>path location to store (or reuse) content from the aggregator</description>
</property>
<property>
<name>mdstoreManagerUrl</name>
<description>the address of the Mdstore Manager</description>
</property>
<property>
<name>isLookupUrl</name>
<description>the address of the lookUp service</description>
</property>
<property>
<name>sparkDriverMemory</name>
<description>memory for driver process</description>
</property>
<property>
<name>sparkExecutorMemory</name>
<description>memory for individual executor</description>
</property>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property>
</parameters>
<global>
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<configuration>
<property>
<name>mapreduce.job.queuename</name>
<value>${queueName}</value>
</property>
<property>
<name>oozie.launcher.mapred.job.queue.name</name>
<value>${oozieLauncherQueueName}</value>
</property>
<property>
<name>oozie.action.sharelib.for.spark</name>
<value>${oozieActionShareLibForSpark2}</value>
</property>
</configuration>
</global>
<start to="ImportODF_hdfs"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="ImportODF_hdfs">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>ImportODF_hdfs</name>
<class>eu.dnetlib.dhp.oa.graph.raw.MigrateHdfsMdstoresApplication</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory ${sparkExecutorMemory}
--executor-cores ${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts>
<arg>--hdfsPath</arg><arg>${contentPath}/odf_records_hdfs</arg>
<arg>--mdstoreManagerUrl</arg><arg>${mdstoreManagerUrl}</arg>
<arg>--mdFormat</arg><arg>ODF</arg>
<arg>--mdLayout</arg><arg>store</arg>
<arg>--mdInterpretation</arg><arg>cleaned</arg>
</spark>
<ok to="GenerateEntities"/>
<error to="Kill"/>
</action>
<action name="GenerateEntities">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>GenerateEntities</name>
<class>eu.dnetlib.dhp.oa.graph.raw.GenerateEntitiesApplication</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory ${sparkExecutorMemory}
--executor-cores ${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts>
<arg>--sourcePaths</arg><arg>${contentPath}/odf_records_hdfs</arg>
<arg>--targetPath</arg><arg>${workingDir}/entities</arg>
<arg>--isLookupUrl</arg><arg>${isLookupUrl}</arg>
<arg>--shouldHashId</arg><arg>${shouldHashId}</arg>
</spark>
<ok to="GenerateGraph"/>
<error to="Kill"/>
</action>
<action name="GenerateGraph">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>GenerateGraph</name>
<class>eu.dnetlib.dhp.oa.graph.raw.DispatchEntitiesApplication</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory ${sparkExecutorMemory}
--executor-cores ${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=7680
</spark-opts>
<arg>--sourcePath</arg><arg>${workingDir}/entities</arg>
<arg>--graphRawPath</arg><arg>${workingDir}/graph_raw</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<end name="End"/>
</workflow-app>

View File

@ -195,8 +195,8 @@ public class MappersTest {
.findFirst()
.get();
assertEquals("0000-0001-6651-1178", pid.getValue());
assertEquals("ORCID", pid.getQualifier().getClassid());
assertEquals("Open Researcher and Contributor ID", pid.getQualifier().getClassname());
assertEquals(ModelConstants.ORCID_PENDING, pid.getQualifier().getClassid());
assertEquals(ModelConstants.ORCID_CLASSNAME, pid.getQualifier().getClassname());
assertEquals(ModelConstants.DNET_PID_TYPES, pid.getQualifier().getSchemeid());
assertEquals(ModelConstants.DNET_PID_TYPES, pid.getQualifier().getSchemename());
assertEquals("Votsi,Nefta", author.get().getFullname());