alessia.bardi
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Merge remote-tracking branch 'upstream/master'

This commit is contained in:
Miriam Baglioni 2020-02-19 17:03:10 +01:00
parent ab84163bb3 5e5e32cb48
commit e8af7a6b64
12 changed files with 32 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
dhp-common/pom.xml
View File

@ -46,6 +46,18 @@
<groupId>net.sf.saxon</groupId>
<artifactId>Saxon-HE</artifactId>
</dependency>
<dependency>
<groupId>org.slf4j</groupId>
<artifactId>jcl-over-slf4j</artifactId>
</dependency>
<dependency>
<groupId>org.apache.cxf</groupId>
<artifactId>cxf-rt-transports-http</artifactId>
</dependency>
<dependency>
<groupId>eu.dnetlib</groupId>
<artifactId>cnr-rmi-api</artifactId>
</dependency>
</dependencies>
</project>

2
dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/graph/utils/ISLookupClientFactory.java → dhp-common/src/main/java/eu/dnetlib/dhp/utils/ISLookupClientFactory.java
View File

@ -1,4 +1,4 @@
package eu.dnetlib.dhp.graph.utils;
package eu.dnetlib.dhp.utils;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService;
import org.apache.commons.logging.Log;

2
dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/migration/oozie_app/workflow.xml
View File

@ -43,7 +43,7 @@
</property>
</parameters>
<start to="ResetWorkingPath"/>
<start to="ImportODFEntitiesFromMongoDB"/>
<kill name="Kill">

19
dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/migration/sql/queryProjects.sql
View File

@ -28,15 +28,17 @@ SELECT
p.summary AS summary,
p.currency AS currency,
p.totalcost AS totalcost,
p.fundedamount AS fundedamount,
p.fundedamount AS fundedamount,
dc.id AS collectedfromid,
dc.officialname AS collectedfromname,
ctc.code || '@@@' || ctc.name || '@@@' || cts.code || '@@@' || cts.name AS contracttype,
pac.code || '@@@' || pac.name || '@@@' || pas.code || '@@@' || pas.name AS provenanceaction,
array_agg(DISTINCT i.pid || '###' || i.issuertype) AS pid,
array_agg(DISTINCT s.name || '###' || sc.code || '@@@' || sc.name || '@@@' || ss.code || '@@@' || ss.name) AS subjects,
array_agg(DISTINCT fp.path) AS fundingtree
p.contracttype || '@@@' || p.contracttypename || '@@@' || p.contracttypescheme || '@@@' || p.contracttypescheme AS contracttype,
pac.code || '@@@' || pac.name || '@@@' || pas.code || '@@@' || pas.name AS provenanceaction,
array_agg(DISTINCT i.pid || '###' || i.issuertype) AS pid,
array_agg(DISTINCT s.name || '###' || sc.code || '@@@' || sc.name || '@@@' || ss.code || '@@@' || ss.name) AS subjects,
array_agg(DISTINCT fp.path) AS fundingtree
FROM projects p
LEFT OUTER JOIN class pac ON (pac.code = p.provenanceactionclass)
LEFT OUTER JOIN scheme pas ON (pas.code = p.provenanceactionscheme)
@ -54,9 +56,6 @@ SELECT
LEFT OUTER JOIN class sc ON (sc.code = s.semanticclass)
LEFT OUTER JOIN scheme ss ON (ss.code = s.semanticscheme)
LEFT OUTER JOIN class ctc ON (ctc.code = p.contracttypeclass)
LEFT OUTER JOIN scheme cts ON (cts.code = p.contracttypescheme)
GROUP BY
p.id,
p.code,
@ -87,4 +86,4 @@ SELECT
dc.id,
dc.officialname,
pac.code, pac.name, pas.code, pas.name,
ctc.code, ctc.name, cts.code, cts.name;
p.contracttype , p.contracttypename, p.contracttypescheme;

8
dhp-workflows/dhp-graph-provision/pom.xml
View File

@ -66,14 +66,6 @@
<artifactId>zookeeper</artifactId>
</dependency>
<dependency>
<groupId>org.apache.cxf</groupId>
<artifactId>cxf-rt-transports-http</artifactId>
</dependency>
<dependency>
<groupId>eu.dnetlib</groupId>
<artifactId>cnr-rmi-api</artifactId>
</dependency>
<dependency>
<groupId>eu.dnetlib.dhp</groupId>

2
dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/graph/SparkXmlIndexingJob.java
View File

@ -2,8 +2,8 @@ package eu.dnetlib.dhp.graph;
import com.lucidworks.spark.util.SolrSupport;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.graph.utils.ISLookupClientFactory;
import eu.dnetlib.dhp.graph.utils.StreamingInputDocumentFactory;
import eu.dnetlib.dhp.utils.ISLookupClientFactory;
import eu.dnetlib.dhp.utils.saxon.SaxonTransformerFactory;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpDocumentNotFoundException;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpException;

1
dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/graph/utils/ContextMapper.java
View File

@ -1,6 +1,7 @@
package eu.dnetlib.dhp.graph.utils;
import com.google.common.base.Joiner;
import eu.dnetlib.dhp.utils.ISLookupClientFactory;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpException;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService;
import org.dom4j.Document;

1
dhp-workflows/dhp-graph-provision/src/test/resources/eu/dnetlib/dhp/graph/datasource.json
View File

File diff suppressed because one or more lines are too long

5
dhp-workflows/dhp-graph-provision/src/test/resources/eu/dnetlib/dhp/graph/related_entity.json
View File

@ -1,5 +0,0 @@
{
"id": "20|nih_________::6b8108b6d6399f7163a6a7ccdd0efc2d",
"type": "organization",
"legalname": "MCGILL UNIVERSITY"
}

1
dhp-workflows/dhp-graph-provision/src/test/resources/eu/dnetlib/dhp/graph/result.json
View File

File diff suppressed because one or more lines are too long

1
dhp-workflows/dhp-graph-provision/src/test/resources/eu/dnetlib/dhp/graph/software.json
View File

File diff suppressed because one or more lines are too long

7
pom.xml
View File

@ -129,6 +129,13 @@
<scope>provided</scope>
</dependency>
<dependency>
<groupId>org.slf4j</groupId>
<artifactId>jcl-over-slf4j</artifactId>
<version>1.7.25</version>
<scope>provided</scope>
</dependency>
<dependency>
<groupId>org.apache.commons</groupId>
<artifactId>commons-lang3</artifactId>