From 0f8e2ecce6b8e55942bb56de4f4fdae462f25129 Mon Sep 17 00:00:00 2001
From: Sandro La Bruzzo <sandro.labruzzo@isti.cnr.it>
Date: Fri, 29 Jan 2021 10:45:07 +0100
Subject: [PATCH] Merged Datacite transfrom into this branch

---
 dhp-workflows/dhp-aggregation/pom.xml         | 14 ++++--------
 .../DataciteToOAFTransformation.scala         | 14 +++++++++---
 .../datacite/ImportDatacite.scala             |  1 +
 .../datacite/oozie_app/workflow.xml           | 22 ++++++++++++++++++-
 4 files changed, 37 insertions(+), 14 deletions(-)

diff --git a/dhp-workflows/dhp-aggregation/pom.xml b/dhp-workflows/dhp-aggregation/pom.xml
index 0445e0e1b..b61c3d443 100644
--- a/dhp-workflows/dhp-aggregation/pom.xml
+++ b/dhp-workflows/dhp-aggregation/pom.xml
@@ -37,7 +37,7 @@
         </plugins>
 
     </build>
-  
+
     <dependencies>
 
         <dependency>
@@ -58,15 +58,9 @@
             <groupId>eu.dnetlib.dhp</groupId>
             <artifactId>dhp-common</artifactId>
             <version>${project.version}</version>
-            <exclusions>
-                <exclusion>
-                    <groupId>com.sun.xml.bind</groupId>
-                    <artifactId>jaxb-core</artifactId>
-                </exclusion>
-            </exclusions>
         </dependency>
-        
-         <dependency>
+
+        <dependency>
             <groupId>eu.dnetlib.dhp</groupId>
             <artifactId>dhp-schemas</artifactId>
             <version>${project.version}</version>
@@ -116,4 +110,4 @@
 
     </dependencies>
 
-</project>
+</project>
\ No newline at end of file
diff --git a/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/datacite/DataciteToOAFTransformation.scala b/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/datacite/DataciteToOAFTransformation.scala
index 9418e71da..933f1445f 100644
--- a/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/datacite/DataciteToOAFTransformation.scala
+++ b/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/datacite/DataciteToOAFTransformation.scala
@@ -11,9 +11,10 @@ import org.json4s.JsonAST.{JField, JObject, JString}
 import org.json4s.jackson.JsonMethods.parse
 
 import java.nio.charset.CodingErrorAction
+import java.text.SimpleDateFormat
 import java.time.LocalDate
 import java.time.format.DateTimeFormatter
-import java.util.Locale
+import java.util.{Date, Locale}
 import java.util.regex.Pattern
 import scala.collection.JavaConverters._
 import scala.io.{Codec, Source}
@@ -44,6 +45,8 @@ object DataciteToOAFTransformation {
   codec.onMalformedInput(CodingErrorAction.REPLACE)
   codec.onUnmappableCharacter(CodingErrorAction.REPLACE)
 
+
+
   private val PID_VOCABULARY = "dnet:pid_types"
   val COBJ_VOCABULARY = "dnet:publication_resource"
   val RESULT_VOCABULARY = "dnet:result_typologies"
@@ -298,8 +301,13 @@ object DataciteToOAFTransformation {
     result.setPid(List(pid).asJava)
     result.setId(OafMapperUtils.createOpenaireId(50, s"datacite____::$doi", true))
     result.setOriginalId(List(doi).asJava)
-    result.setDateofcollection(s"${dateOfCollection}")
-    result.setDateoftransformation(s"$ts")
+
+    val d = new Date(dateOfCollection*1000)
+    val ISO8601FORMAT = new SimpleDateFormat("yyyy-MM-dd'T'HH:mm:ssZ", Locale.US)
+
+
+    result.setDateofcollection(ISO8601FORMAT.format(d))
+    result.setDateoftransformation(ISO8601FORMAT.format(ts))
     result.setDataInfo(dataInfo)
 
     val creators = (json \\ "creators").extractOrElse[List[CreatorType]](List())
diff --git a/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/datacite/ImportDatacite.scala b/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/datacite/ImportDatacite.scala
index 06fcbb518..d5edb674a 100644
--- a/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/datacite/ImportDatacite.scala
+++ b/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/datacite/ImportDatacite.scala
@@ -108,6 +108,7 @@ object ImportDatacite {
 
     val cnt = writeSequenceFile(hdfsTargetPath, ts, conf)
 
+
     log.info(s"Imported from Datacite API $cnt documents")
 
     if (cnt  > 0) {
diff --git a/dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/actionmanager/datacite/oozie_app/workflow.xml b/dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/actionmanager/datacite/oozie_app/workflow.xml
index a3caa5e23..047794c9c 100644
--- a/dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/actionmanager/datacite/oozie_app/workflow.xml
+++ b/dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/actionmanager/datacite/oozie_app/workflow.xml
@@ -13,13 +13,25 @@
             <name>nativeInputPath</name>
             <description>the path of the input MDStore</description>
         </property>
+
+
     </parameters>
 
-    <start to="ImportDatacite"/>
+    <start to="resume_from"/>
+
+
     <kill name="Kill">
         <message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
     </kill>
 
+    <decision name="resume_from">
+        <switch>
+            <case to="TransformJob">${wf:conf('resumeFrom') eq 'TransformJob'}</case>
+            <case to="ExportDataset">${wf:conf('resumeFrom') eq 'ExportDataset'}</case>
+            <default to="ImportDatacite"/>
+        </switch>
+    </decision>
+
     <action name="ImportDatacite">
         <spark xmlns="uri:oozie:spark-action:0.2">
             <master>yarn-cluster</master>
@@ -69,6 +81,14 @@
             <arg>-tr</arg><arg>${isLookupUrl}</arg>
             <arg>--master</arg><arg>yarn-cluster</arg>
         </spark>
+        <ok to="DeletePathIfExists"/>
+        <error to="Kill"/>
+    </action>
+
+    <action name="DeletePathIfExists">
+        <fs>
+            <delete path='${mdstoreOutputPath}_raw_AS'/>
+        </fs>
         <ok to="ExportDataset"/>
         <error to="Kill"/>
     </action>