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

avoid to save intermediate dataset before generation of Sequence file

This commit is contained in:
Sandro La Bruzzo 2021-01-04 17:54:57 +01:00
parent e79445a8b4
commit 7834a35768
1 changed files with 15 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/SparkGenerateDOIBoostActionSet.scala
View File

@ -41,30 +41,34 @@ object SparkGenerateDOIBoostActionSet {
val workingDirPath = parser.get("targetPath")
val sequenceFilePath = parser.get("sFilePath")
spark.read.load(dbDatasetPath).as[OafDataset]
val asDataset = spark.read.load(dbDatasetPath).as[OafDataset]
.map(d =>DoiBoostMappingUtil.fixResult(d))
.map(d=>DoiBoostMappingUtil.toActionSet(d))(Encoders.tuple(Encoders.STRING, Encoders.STRING))
.write.mode(SaveMode.Overwrite).save(s"$workingDirPath/actionSet")
// .write.mode(SaveMode.Overwrite).save(s"$workingDirPath/actionSet")
spark.read.load(dbPublicationPath).as[Publication]
val asPublication =spark.read.load(dbPublicationPath).as[Publication]
.map(d=>DoiBoostMappingUtil.toActionSet(d))(Encoders.tuple(Encoders.STRING, Encoders.STRING))
.write.mode(SaveMode.Append).save(s"$workingDirPath/actionSet")
// .write.mode(SaveMode.Append).save(s"$workingDirPath/actionSet")
spark.read.load(dbOrganizationPath).as[Organization]
val asOrganization = spark.read.load(dbOrganizationPath).as[Organization]
.map(d=>DoiBoostMappingUtil.toActionSet(d))(Encoders.tuple(Encoders.STRING, Encoders.STRING))
.write.mode(SaveMode.Append).save(s"$workingDirPath/actionSet")
// .write.mode(SaveMode.Append).save(s"$workingDirPath/actionSet")
spark.read.load(crossRefRelation).as[Relation]
val asCRelation = spark.read.load(crossRefRelation).as[Relation]
.map(d=>DoiBoostMappingUtil.toActionSet(d))(Encoders.tuple(Encoders.STRING, Encoders.STRING))
.write.mode(SaveMode.Append).save(s"$workingDirPath/actionSet")
// .write.mode(SaveMode.Append).save(s"$workingDirPath/actionSet")
spark.read.load(dbaffiliationRelationPath).as[Relation]
val asRelAffiliation = spark.read.load(dbaffiliationRelationPath).as[Relation]
.map(d=>DoiBoostMappingUtil.toActionSet(d))(Encoders.tuple(Encoders.STRING, Encoders.STRING))
.write.mode(SaveMode.Append).save(s"$workingDirPath/actionSet")
// .write.mode(SaveMode.Append).save(s"$workingDirPath/actionSet")
val d: Dataset[(String, String)] =spark.read.load(s"$workingDirPath/actionSet").as[(String,String)]
val d: Dataset[(String, String)] = asDataset.union(asPublication).union(asOrganization).union(asCRelation).union(asRelAffiliation)
// spark.read.load(s"$workingDirPath/actionSet").as[(String,String)]
d.rdd.repartition(6000).map(s => (new Text(s._1), new Text(s._2))).saveAsHadoopFile(s"$sequenceFilePath", classOf[Text], classOf[Text], classOf[SequenceFileOutputFormat[Text,Text]], classOf[GzipCodec])