enrico.ottonello
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

[cleaning] cleaning instance.pid and instance.alternateidentifier using the same procedure used to clean result.pid

This commit is contained in:
Claudio Atzori 2021-06-24 12:10:17 +02:00
parent 2e8fd2c531
commit 67afd06cd1
1 changed files with 8 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
dhp-common/src/main/java/eu/dnetlib/dhp/schema/oaf/utils/GraphCleaningFunctions.java
View File

@ -224,24 +224,20 @@ public class GraphCleaningFunctions extends CleaningFunctions {
if (Objects.nonNull(r.getInstance())) {
for (Instance i : r.getInstance()) {
if (Objects.nonNull(i.getPid())) {
i.setPid(processPidCleaning(i.getPid()));
}
if (Objects.nonNull(i.getAlternateIdentifier())) {
i.setAlternateIdentifier(processPidCleaning(i.getAlternateIdentifier()));
}
Optional
.ofNullable(i.getPid())
.ifPresent(pid -> {
final Set<StructuredProperty> pids = pid
.stream()
.filter(Objects::nonNull)
.filter(p -> StringUtils.isNotBlank(p.getValue()))
.collect(Collectors.toCollection(HashSet::new));
final Set<StructuredProperty> pids = Sets.newHashSet(pid);
Optional
.ofNullable(i.getAlternateIdentifier())
.ifPresent(altId -> {
final Set<StructuredProperty> altIds = altId
.stream()
.filter(Objects::nonNull)
.filter(p -> StringUtils.isNotBlank(p.getValue()))
.collect(Collectors.toCollection(HashSet::new));
final Set<StructuredProperty> altIds = Sets.newHashSet(altId);
i.setAlternateIdentifier(Lists.newArrayList(Sets.difference(altIds, pids)));
});
});