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

[CountryPropagation] refactoring

This commit is contained in:
Miriam Baglioni 2022-05-20 09:15:53 +02:00
parent c298c148cb
commit 5e0b8f9b5f
1 changed files with 7 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/countrypropagation/SparkCountryPropagationJob.java
View File

@ -102,7 +102,7 @@ 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()) 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 else
t._1().setCountry(merge(null, t._2().getCountrySet())); t._1().setCountry(merge(null, t._2().getCountrySet()));
@ -113,12 +113,13 @@ public class SparkCountryPropagationJob {
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 = new HashSet<>(); HashSet<String> countries = new HashSet<>();
if(Optional.ofNullable(c1).isPresent()){ if (Optional.ofNullable(c1).isPresent()) {
countries = c1.stream().map(Qualifier::getClassid) countries = c1
.stream()
.map(Qualifier::getClassid)
.collect(Collectors.toCollection(HashSet::new)); .collect(Collectors.toCollection(HashSet::new));
} }
HashSet<String> finalCountries = countries; HashSet<String> finalCountries = countries;
return c2 return c2
.stream() .stream()