From 1265dadc90686ad9a687a8bf7e20fd739d663ae9 Mon Sep 17 00:00:00 2001
From: Enrico Ottonello <enrico.ottonello@isti.cnr.it>
Date: Thu, 20 May 2021 19:01:28 +0200
Subject: [PATCH] workflow aligned with stable_ids

---
 .../orcid/SparkDownloadOrcidAuthors.java      |  2 +-
 .../orcidnodoi/oaf/PublicationToOaf.java      |  4 +-
 .../oozie_app/workflow.xml                    | 42 -------------------
 .../orcidnodoi/oozie_app/workflow.xml         | 31 ++++++++++----
 4 files changed, 27 insertions(+), 52 deletions(-)
 delete mode 100644 dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/import_orcid_no_doi/oozie_app/workflow.xml

diff --git a/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/orcid/SparkDownloadOrcidAuthors.java b/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/orcid/SparkDownloadOrcidAuthors.java
index 36b4b073d..8cf070213 100644
--- a/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/orcid/SparkDownloadOrcidAuthors.java
+++ b/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/orcid/SparkDownloadOrcidAuthors.java
@@ -62,7 +62,7 @@ public class SparkDownloadOrcidAuthors {
 			isSparkSessionManaged,
 			spark -> {
 				String lastUpdate = HDFSUtil.readFromTextFile(hdfsServerUri, workingPath, "last_update.txt");
-				logger.info("lastUpdate: ", lastUpdate);
+				logger.info("lastUpdate: {}", lastUpdate);
 				if (StringUtils.isBlank(lastUpdate)) {
 					throw new RuntimeException("last update info not found");
 				}
diff --git a/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/orcidnodoi/oaf/PublicationToOaf.java b/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/orcidnodoi/oaf/PublicationToOaf.java
index 5c3236222..ccae4d976 100644
--- a/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/orcidnodoi/oaf/PublicationToOaf.java
+++ b/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/orcidnodoi/oaf/PublicationToOaf.java
@@ -174,7 +174,9 @@ public class PublicationToOaf implements Serializable {
 					publication
 						.getExternalReference()
 						.add(
-							convertExtRef(extId, classid, classname, "dnet:pid_types", "dnet:pid_types"));
+							convertExtRef(
+								extId, classid, classname, ModelConstants.DNET_PID_TYPES,
+								ModelConstants.DNET_PID_TYPES));
 				}
 			});
 
diff --git a/dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/import_orcid_no_doi/oozie_app/workflow.xml b/dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/import_orcid_no_doi/oozie_app/workflow.xml
deleted file mode 100644
index becdf0974..000000000
--- a/dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/import_orcid_no_doi/oozie_app/workflow.xml
+++ /dev/null
@@ -1,42 +0,0 @@
-<workflow-app name="import_orcid_no_doi" xmlns="uri:oozie:workflow:0.5">
-    <parameters>
-        <property>
-            <name>inputPath</name>
-            <value>/data/orcid_activities_2020/no_doi_dataset</value>
-            <description>path where retrieve the already generated action set</description>
-        </property>
-        <property>
-            <name>outputPath</name>
-            <value>/data/orcid_activities_2020/test_import_orcid_no_doi</value>
-            <description>path where to store the action set</description>
-        </property>
-    </parameters>
-
-    <global>
-        <job-tracker>${jobTracker}</job-tracker>
-        <name-node>${nameNode}</name-node>
-        <configuration>
-            <property>
-                <name>oozie.action.sharelib.for.spark</name>
-                <value>${oozieActionShareLibForSpark2}</value>
-            </property>
-        </configuration>
-    </global>
-
-    <start to="importOrcidNoDoi"/>
-
-    <kill name="Kill">
-        <message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
-    </kill>
-
-    <action name="importOrcidNoDoi">
-        <distcp xmlns="uri:oozie:distcp-action:0.2">
-            <arg>${inputPath}/*</arg>
-            <arg>${outputPath}</arg>
-        </distcp>
-        <ok to="End"/>
-        <error to="Kill"/>
-    </action>
-
-   <end name="End"/>
-</workflow-app>
\ No newline at end of file
diff --git a/dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/orcidnodoi/oozie_app/workflow.xml b/dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/orcidnodoi/oozie_app/workflow.xml
index 6513ff7e1..365c4d5b4 100644
--- a/dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/orcidnodoi/oozie_app/workflow.xml
+++ b/dhp-workflows/dhp-doiboost/src/main/resources/eu/dnetlib/dhp/doiboost/orcidnodoi/oozie_app/workflow.xml
@@ -1,5 +1,15 @@
 <workflow-app name="gen_orcid_no_doi_dataset" xmlns="uri:oozie:workflow:0.5">
     <parameters>
+        <property>
+            <name>workingPath</name>
+            <value>/data/orcid_activities_2020</value>
+            <description>path where the collection workflow stores the ORCID data</description>
+        </property>
+        <property>
+            <name>outputPath</name>
+            <description>path where to store the action set</description>
+        </property>
+
         <property>
             <name>spark2GenNoDoiDatasetMaxExecutors</name>
             <value>40</value>
@@ -35,10 +45,6 @@
             <name>spark2EventLogDir</name>
             <description>spark 2.* event log dir location</description>
         </property>
-        <property>
-            <name>workingPath</name>
-            <description>the working dir base path</description>
-        </property>
     </parameters>
 
     <global>
@@ -83,11 +89,20 @@
                 --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
                 --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
             </spark-opts>
-            <arg>-w</arg><arg>${workingPath}/</arg>
-            <arg>-n</arg><arg>${nameNode}</arg>
-            <arg>-i</arg><arg>last_orcid_dataset</arg>
-            <arg>-oew</arg><arg>no_doi_dataset</arg>
+            <arg>--workingPath</arg><arg>${workingPath}/</arg>
+            <arg>--hdfsServerUri</arg><arg>${nameNode}</arg>
+            <arg>--orcidDataFolder</arg><arg>last_orcid_dataset</arg>
+            <arg>--outputEnrichedWorksPath</arg><arg>no_doi_dataset</arg>
         </spark>
+        <ok to="importOrcidNoDoi"/>
+        <error to="Kill"/>
+    </action>
+
+    <action name="importOrcidNoDoi">
+        <distcp xmlns="uri:oozie:distcp-action:0.2">
+            <arg>${workingPath}/no_doi_dataset/*</arg>
+            <arg>${outputPath}</arg>
+        </distcp>
         <ok to="End"/>
         <error to="Kill"/>
     </action>