D-Net
/
dnet-hadoop
14
5
Fork 11
Code Issues 14 Pull Requests 12 Projects 2 Releases 2 Wiki Activity

hostedbymap #136

Merged
claudio.atzori merged 60 commits from hostedbymap into beta 2021-08-12 17:10:55 +02:00
Conversation 1 Commits 60 Files Changed 46 +1 -5
1 changed files with 1 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit eccf3851b0 - Show all commits

6
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/hostedbymap/SparkProduceHostedByMap.scala
View File

@ -23,10 +23,6 @@ object SparkProduceHostedByMap {
implicit val tupleForJoinEncoder: Encoder[(String, HostedByItemType)] = Encoders.tuple(Encoders.STRING, Encoders.product[HostedByItemType]) implicit val tupleForJoinEncoder: Encoder[(String, HostedByItemType)] = Encoders.tuple(Encoders.STRING, Encoders.product[HostedByItemType])
def toHostedByItemType(input: ((HostedByInfo, HostedByInfo), HostedByInfo)) : HostedByItemType = { def toHostedByItemType(input: ((HostedByInfo, HostedByInfo), HostedByInfo)) : HostedByItemType = {
val openaire: HostedByInfo = input._1._1 val openaire: HostedByInfo = input._1._1
val doaj: HostedByInfo = input._1._2 val doaj: HostedByInfo = input._1._2
@ -217,7 +213,7 @@ object SparkProduceHostedByMap {
.union(doajHostedByDataset(spark, workingDirPath + "/doaj")) .union(doajHostedByDataset(spark, workingDirPath + "/doaj"))
.flatMap(hbi => toList(hbi))).filter(hbi => hbi._2.id.startsWith("10|")) .flatMap(hbi => toList(hbi))).filter(hbi => hbi._2.id.startsWith("10|"))
.map(hbi => toHostedByMap(hbi))(Encoders.STRING) .map(hbi => toHostedByMap(hbi))(Encoders.STRING)
.rdd.saveAsTextFile(outputPath + "/hostedByMap", classOf[GzipCodec]) .rdd.saveAsTextFile(outputPath , classOf[GzipCodec])
} }