sabeel
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

minor fix

This commit is contained in:
Sandro La Bruzzo 2021-01-04 17:37:08 +01:00
parent b0dc92786f
commit 8765020b85
4 changed files with 11 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/SparkGenerateDoiBoost.scala
View File

@ -31,7 +31,7 @@ object SparkGenerateDoiBoost {
import spark.implicits._ import spark.implicits._
val hostedByMapPath = parser.get("hostedByMapPath") val hostedByMapPath = parser.get("hostedByMapPath")
val workingDirPath = parser.get("workingDirPath") val workingDirPath = parser.get("workingPath")
implicit val mapEncoderPub: Encoder[Publication] = Encoders.kryo[Publication] implicit val mapEncoderPub: Encoder[Publication] = Encoders.kryo[Publication]

2
dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/mag/SparkPreProcessMAG.scala → dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/mag/SparkProcessMAG.scala
View File

@ -11,7 +11,7 @@ import org.slf4j.{Logger, LoggerFactory}
import scala.collection.JavaConverters._ import scala.collection.JavaConverters._
object SparkPreProcessMAG { object SparkProcessMAG {
def main(args: Array[String]): Unit = { def main(args: Array[String]): Unit = {
val logger: Logger = LoggerFactory.getLogger(getClass) val logger: Logger = LoggerFactory.getLogger(getClass)

2
dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/generate_doiboost_params.json
View File

@ -3,5 +3,5 @@
{"paramName": "hb", "paramLongName":"hostedByMapPath", "paramDescription": "the hosted By Map Path", "paramRequired": true}, {"paramName": "hb", "paramLongName":"hostedByMapPath", "paramDescription": "the hosted By Map Path", "paramRequired": true},
{"paramName": "ap", "paramLongName":"affiliationPath", "paramDescription": "the Affliation Path", "paramRequired": true}, {"paramName": "ap", "paramLongName":"affiliationPath", "paramDescription": "the Affliation Path", "paramRequired": true},
{"paramName": "pa", "paramLongName":"paperAffiliationPath", "paramDescription": "the paperAffiliation Path", "paramRequired": true}, {"paramName": "pa", "paramLongName":"paperAffiliationPath", "paramDescription": "the paperAffiliation Path", "paramRequired": true},
{"paramName": "w", "paramLongName":"workingDirPath", "paramDescription": "the Working Path", "paramRequired": true} {"paramName": "w", "paramLongName":"workingPath", "paramDescription": "the Working Path", "paramRequired": true}
] ]

20
dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/oozie_app/workflow.xml
View File

@ -166,9 +166,9 @@
<!-- MAG SECTION --> <!-- MAG SECTION -->
<action name="ResetMagWorkingPath"> <action name="ResetMagWorkingPath">
<fs> <fs>
<delete path='${inputPathMAG}/dataset'/> <delete path="${inputPathMAG}/dataset"/>
<delete path='${inputPathMAG}/process'/> <delete path="${inputPathMAG}/process"/>
<delete path='${inputPathMAG}/dataset'/> <delete path="${inputPathMAG}/dataset"/>
</fs> </fs>
<ok to="ConvertMagToDataset"/> <ok to="ConvertMagToDataset"/>
<error to="Kill"/> <error to="Kill"/>
@ -191,18 +191,16 @@
<arg>--targetPath</arg><arg>${inputPathMAG}/dataset</arg> <arg>--targetPath</arg><arg>${inputPathMAG}/dataset</arg>
<arg>--master</arg><arg>yarn-cluster</arg> <arg>--master</arg><arg>yarn-cluster</arg>
</spark> </spark>
<ok to="PreprocessMag"/> <ok to="ProcessMAG"/>
<error to="Kill"/> <error to="Kill"/>
</action> </action>
<action name="ProcessMAG">
<action name="PreprocessMag">
<spark xmlns="uri:oozie:spark-action:0.2"> <spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<name>Convert Mag to OAF Dataset</name> <name>Convert Mag to OAF Dataset</name>
<class>eu.dnetlib.doiboost.mag.SparkPreProcessMAG</class> <class>eu.dnetlib.doiboost.mag.SparkProcessMAG</class>
<jar>dhp-doiboost-${projectVersion}.jar</jar> <jar>dhp-doiboost-${projectVersion}.jar</jar>
<spark-opts> <spark-opts>
--executor-memory=${sparkExecutorMemory} --executor-memory=${sparkExecutorMemory}
@ -244,8 +242,6 @@
<error to="Kill"/> <error to="Kill"/>
</action> </action>
<!-- ORCID SECTION --> <!-- ORCID SECTION -->
<action name="PreprocessORCID"> <action name="PreprocessORCID">
<spark xmlns="uri:oozie:spark-action:0.2"> <spark xmlns="uri:oozie:spark-action:0.2">
@ -269,7 +265,7 @@
<error to="Kill"/> <error to="Kill"/>
</action> </action>
<!-- INTERSECTION SECTION--> <!-- INTERSECTION SECTION-->
<action name="CreateDOIBoost"> <action name="CreateDOIBoost">
<spark xmlns="uri:oozie:spark-action:0.2"> <spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn-cluster</master> <master>yarn-cluster</master>
@ -287,7 +283,7 @@
<arg>--hostedByMapPath</arg><arg>${hostedByMapPath}</arg> <arg>--hostedByMapPath</arg><arg>${hostedByMapPath}</arg>
<arg>--affiliationPath</arg><arg>${inputPathMAG}/process/Affiliations</arg> <arg>--affiliationPath</arg><arg>${inputPathMAG}/process/Affiliations</arg>
<arg>--paperAffiliationPath</arg><arg>${inputPathMAG}/process/PaperAuthorAffiliations</arg> <arg>--paperAffiliationPath</arg><arg>${inputPathMAG}/process/PaperAuthorAffiliations</arg>
<arg>--workingDirPath</arg><arg>${workingPath}</arg> <arg>--workingPath</arg><arg>${workingPath}</arg>
<arg>--master</arg><arg>yarn-cluster</arg> <arg>--master</arg><arg>yarn-cluster</arg>
</spark> </spark>
<ok to="GenerateActionSet"/> <ok to="GenerateActionSet"/>