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

hostedby patching to work with the updated Crossref contents

This commit is contained in:
Claudio Atzori 2024-05-28 15:28:13 +02:00
parent fb266efbcb
commit db5e18c784
2 changed files with 29 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/hostedbymap/oozie_app/workflow.xml
View File

@ -223,11 +223,13 @@
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
--conf spark.sql.shuffle.partitions=15000
</spark-opts>
<arg>--hostedByMapPath</arg><arg>${hostedByMapPath}</arg>
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
@ -253,11 +255,13 @@
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
--conf spark.sql.shuffle.partitions=15000
</spark-opts>
<arg>--outputPath</arg><arg>${outputPath}/publication</arg>
<arg>--preparedInfoPath</arg><arg>${workingDir}/preparedInfo</arg>
@ -278,6 +282,7 @@
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.executor.memoryOverhead=${sparkExecutorMemory}
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}

19
dhp-workflows/dhp-graph-mapper/src/main/scala/eu/dnetlib/dhp/oa/graph/hostedbymap/SparkApplyHostedByMapToResult.scala
View File

@ -25,6 +25,20 @@ object SparkApplyHostedByMapToResult {
val i = p.getInstance().asScala
if (i.size == 1) {
val inst: Instance = i.head
patchInstance(p, ei, inst)
} else if (i.size == 2) {
if (i.map(ii => ii.getCollectedfrom.getValue).contains("UnpayWall")) {
val inst: Instance = i.filter(ii => "Crossref".equals(ii.getCollectedfrom.getValue)).head
patchInstance(p, ei, inst)
}
}
}
p
})(Encoders.bean(classOf[Publication]))
}
private def patchInstance(p: Publication, ei: EntityInfo, inst: Instance): Unit = {
inst.getHostedby.setKey(ei.getHostedById)
inst.getHostedby.setValue(ei.getName)
if (ei.getOpenAccess) {
@ -39,11 +53,6 @@ object SparkApplyHostedByMapToResult {
inst.getAccessright.setOpenAccessRoute(OpenAccessRoute.gold)
p.setBestaccessright(OafMapperUtils.createBestAccessRights(p.getInstance()));
}
}
}
p
})(Encoders.bean(classOf[Publication]))
}
def main(args: Array[String]): Unit = {