D-Net
/
dnet-hadoop
14
5
Fork 11
Code Issues 14 Pull Requests 13 Projects 2 Releases 2 Wiki Activity

[AffRo] used the collectedfrom openaire for all the relations imported as affRo output

This commit is contained in:
Miriam Baglioni 2024-09-25 17:23:49 +02:00
parent d0eba032cd
commit 0765641979
1 changed files with 9 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/bipaffiliations/PrepareAffiliationRelations.java
View File

@ -98,35 +98,26 @@ public class PrepareAffiliationRelations implements Serializable {
private static void createActionSet(SparkSession spark, String crossrefInputPath, String pubmedInputPath,
String openapcInputPath, String dataciteInputPath, String webcrawlInputPath, String publisherlInputPath,
String outputPath) {
List<KeyValue> collectedFromCrossref = OafMapperUtils
.listKeyValues(ModelConstants.CROSSREF_ID, "Crossref");
List<KeyValue> collectedfromOpenAIRE = OafMapperUtils
.listKeyValues(OPENAIRE_DATASOURCE_ID, OPENAIRE_DATASOURCE_NAME);
JavaPairRDD<Text, Text> crossrefRelations = prepareAffiliationRelations(
spark, crossrefInputPath, collectedFromCrossref);
spark, crossrefInputPath, collectedfromOpenAIRE);
List<KeyValue> collectedFromPubmed = OafMapperUtils
.listKeyValues(ModelConstants.PUBMED_CENTRAL_ID, "Pubmed");
JavaPairRDD<Text, Text> pubmedRelations = prepareAffiliationRelations(
spark, pubmedInputPath, collectedFromPubmed);
spark, pubmedInputPath, collectedfromOpenAIRE);
List<KeyValue> collectedFromOpenAPC = OafMapperUtils
.listKeyValues(ModelConstants.OPEN_APC_ID, "OpenAPC");
JavaPairRDD<Text, Text> openAPCRelations = prepareAffiliationRelations(
spark, openapcInputPath, collectedFromOpenAPC);
spark, openapcInputPath, collectedfromOpenAIRE);
List<KeyValue> collectedFromDatacite = OafMapperUtils
.listKeyValues(ModelConstants.DATACITE_ID, "Datacite");
JavaPairRDD<Text, Text> dataciteRelations = prepareAffiliationRelations(
spark, dataciteInputPath, collectedFromDatacite);
spark, dataciteInputPath, collectedfromOpenAIRE);
List<KeyValue> collectedFromWebCrawl = OafMapperUtils
.listKeyValues(OPENAIRE_DATASOURCE_ID, OPENAIRE_DATASOURCE_NAME);
JavaPairRDD<Text, Text> webCrawlRelations = prepareAffiliationRelations(
spark, webcrawlInputPath, collectedFromWebCrawl);
spark, webcrawlInputPath, collectedfromOpenAIRE);
List<KeyValue> collectedfromPublisher = OafMapperUtils
.listKeyValues(OPENAIRE_DATASOURCE_ID, OPENAIRE_DATASOURCE_NAME);
JavaPairRDD<Text, Text> publisherRelations = prepareAffiliationRelationFromPublisher(
spark, publisherlInputPath, collectedfromPublisher);
spark, publisherlInputPath, collectedfromOpenAIRE);
crossrefRelations
.union(pubmedRelations)