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

[CountryPropagation] fix NPE issue

This commit is contained in:
Miriam Baglioni 2022-05-20 09:11:46 +02:00
parent eaf9385ae5
commit c298c148cb
1 changed files with 12 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/countrypropagation/SparkCountryPropagationJob.java
View File

@ -102,21 +102,27 @@ public class SparkCountryPropagationJob {
private static <R extends Result> MapFunction<Tuple2<R, ResultCountrySet>, R> getCountryMergeFn() { private static <R extends Result> MapFunction<Tuple2<R, ResultCountrySet>, R> getCountryMergeFn() {
return t -> { return t -> {
Optional.ofNullable(t._2()).ifPresent(r -> { Optional.ofNullable(t._2()).ifPresent(r -> {
if(Optional.ofNullable(t._1().getCountry()).isPresent())
t._1().getCountry().addAll(merge(t._1().getCountry(), r.getCountrySet())); t._1().getCountry().addAll(merge(t._1().getCountry(), r.getCountrySet()));
else
t._1().setCountry(merge(null, t._2().getCountrySet()));
}); });
return t._1(); return t._1();
}; };
} }
private static List<Country> merge(List<Country> c1, List<CountrySbs> c2) { private static List<Country> merge(List<Country> c1, List<CountrySbs> c2) {
HashSet<String> countries = c1 HashSet<String> countries = new HashSet<>();
.stream() if(Optional.ofNullable(c1).isPresent()){
.map(Qualifier::getClassid) countries = c1.stream().map(Qualifier::getClassid)
.collect(Collectors.toCollection(HashSet::new)); .collect(Collectors.toCollection(HashSet::new));
}
HashSet<String> finalCountries = countries;
return c2 return c2
.stream() .stream()
.filter(c -> !countries.contains(c.getClassid())) .filter(c -> !finalCountries.contains(c.getClassid()))
.map(c -> getCountry(c.getClassid(), c.getClassname())) .map(c -> getCountry(c.getClassid(), c.getClassname()))
.collect(Collectors.toList()); .collect(Collectors.toList());
} }