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

ResultTypeComparator gives priority to Records collectedfrom Crossref

This commit is contained in:
Claudio Atzori 2020-11-26 13:09:19 +01:00
parent 5151850a19
commit fa66e5b6b8
1 changed files with 24 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
dhp-common/src/main/java/eu/dnetlib/dhp/schema/oaf/ResultTypeComparator.java
View File

@ -2,8 +2,14 @@
package eu.dnetlib.dhp.schema.oaf;
import java.util.Comparator;
import java.util.HashSet;
import java.util.Optional;
import java.util.stream.Collectors;
import java.util.stream.Stream;
import com.google.common.collect.Sets;
import eu.dnetlib.dhp.schema.common.ModelConstants;
import static eu.dnetlib.dhp.schema.common.ModelConstants.CROSSREF_ID;
public class ResultTypeComparator implements Comparator<Result> {
@ -17,6 +23,16 @@ public class ResultTypeComparator implements Comparator<Result> {
if (right == null)
return -1;
HashSet<String> lCf = getCollectedFromIds(left);
HashSet<String> rCf = getCollectedFromIds(right);
if (lCf.contains(CROSSREF_ID) && !rCf.contains(CROSSREF_ID)) {
return -1;
}
if (!lCf.contains(CROSSREF_ID) && rCf.contains(CROSSREF_ID)) {
return 1;
}
String lClass = left.getResulttype().getClassid();
String rClass = right.getResulttype().getClassid();
@ -46,4 +62,12 @@ public class ResultTypeComparator implements Comparator<Result> {
// Else (but unlikely), lexicographical ordering will do.
return lClass.compareTo(rClass);
}
protected HashSet<String> getCollectedFromIds(Result left) {
return Optional.ofNullable(left.getCollectedfrom())
.map(cf -> cf.stream()
.map(c -> c.getKey())
.collect(Collectors.toCollection(HashSet::new)))
.orElse(new HashSet<>());
}
}