D-Net
/
dnet-hadoop
14
5
Fork 11
Code Issues 14 Pull Requests 10 Projects 2 Releases 2 Wiki Activity

handle missing instance.pid during bulk cleaning

This commit is contained in:
Claudio Atzori 2021-03-23 09:28:58 +01:00
parent c73072079d
commit 431cbe9955
2 changed files with 17 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
dhp-common/src/main/java/eu/dnetlib/dhp/schema/oaf/CleaningFunctions.java
View File

@ -149,17 +149,21 @@ public class CleaningFunctions {
if (Objects.nonNull(r.getInstance())) {
for (Instance i : r.getInstance()) {
final Set<StructuredProperty> pids = Sets.newHashSet(i.getPid());
i
.setAlternateIdentifier(
Optional
.ofNullable(i.getAlternateIdentifier())
.map(
altId -> altId
.stream()
.filter(p -> !pids.contains(p))
.collect(Collectors.toList()))
.orElse(Lists.newArrayList()));
Optional
.ofNullable(i.getPid())
.ifPresent(pid -> {
final Set<StructuredProperty> pids = Sets.newHashSet(i.getPid());
i
.setAlternateIdentifier(
Optional
.ofNullable(i.getAlternateIdentifier())
.map(
altId -> altId
.stream()
.filter(p -> !pids.contains(p))
.collect(Collectors.toList()))
.orElse(Lists.newArrayList()));
});
if (Objects.isNull(i.getAccessright()) || StringUtils.isBlank(i.getAccessright().getClassid())) {
i

3
dhp-workflows/dhp-graph-mapper/src/test/java/eu/dnetlib/dhp/oa/graph/clean/CleaningFunctionTest.java
View File

@ -19,6 +19,7 @@ import org.mockito.junit.jupiter.MockitoExtension;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.common.vocabulary.VocabularyGroup;
import eu.dnetlib.dhp.schema.common.ModelConstants;
import eu.dnetlib.dhp.schema.oaf.*;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpException;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService;
@ -78,7 +79,7 @@ public class CleaningFunctionTest {
assertEquals("CLOSED", p_out.getInstance().get(0).getAccessright().getClassid());
assertEquals("Closed Access", p_out.getInstance().get(0).getAccessright().getClassname());
Set<String> pidTerms = vocabularies.getTerms("dnet:pid_types");
Set<String> pidTerms = vocabularies.getTerms(ModelConstants.DNET_PID_TYPES);
assertTrue(
p_out
.getPid()