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

extended existing code to import of POCI from open citation #340

Merged
claudio.atzori merged 4 commits from importpoci into beta 2023-10-03 10:52:12 +02:00
Conversation 2 Commits 4 Files Changed 7 +4 -4
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit d7fccdc64b - Show all commits

8
dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/actionmanager/opencitations/oozie_app/workflow.xml
View File

@ -46,7 +46,7 @@
</configuration>
<exec>download.sh</exec>
<argument>${filelist}</argument>
<argument>${workingPath}/Original</argument>
<argument>${workingPath}/${prefix}/Original</argument>
<env-var>HADOOP_USER_NAME=${wf:user()}</env-var>
<file>download.sh</file>
<capture-output/>
@ -59,7 +59,7 @@
<main-class>eu.dnetlib.dhp.actionmanager.opencitations.GetOpenCitationsRefs</main-class>
<arg>--hdfsNameNode</arg><arg>${nameNode}</arg>
<arg>--inputFile</arg><arg>${inputFile}</arg>
<arg>--workingPath</arg><arg>${workingPath}</arg>
<arg>--workingPath</arg><arg>${workingPath}/${prefix}</arg>
<arg>--prefix</arg><arg>${prefix}</arg>
</java>
<ok to="read"/>
@ -83,8 +83,8 @@
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--workingPath</arg><arg>${workingPath}/${prefix}</arg>
<arg>--outputPath</arg><arg>${workingPath}/${prefix}_JSON/</arg>
<arg>--workingPath</arg><arg>${workingPath}/${prefix}/${prefix}</arg>
<arg>--outputPath</arg><arg>${workingPath}/${prefix}/${prefix}_JSON/</arg>
<arg>--delimiter</arg><arg>${delimiter}</arg>
<arg>--inputFile</arg><arg>${inputFileCoci}</arg>
<arg>--format</arg><arg>${prefix}</arg>