spyros
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Merge branch 'master' of https://code-repo.d4science.org/D-Net/dnet-hadoop

This commit is contained in:
Claudio Atzori 2020-05-06 11:28:01 +02:00
parent c79e2f5977 8f30a09d84
commit fdfecc9578
4 changed files with 98 additions and 53 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
dhp-workflows/dhp-broker-events/README.md Normal file
View File

@ -0,0 +1,3 @@
# dhp-broker-events
dhp-broker-events is a DNET module responsible
of the production of events for the OpenAIRE Broker Service.

66
dhp-workflows/dhp-broker-events/pom.xml Normal file
View File

@ -0,0 +1,66 @@
<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
<parent>
<artifactId>dhp-workflows</artifactId>
<groupId>eu.dnetlib.dhp</groupId>
<version>1.1.7-SNAPSHOT</version>
</parent>
<modelVersion>4.0.0</modelVersion>
<artifactId>dhp-broker-events</artifactId>
<dependencies>
<dependency>
<groupId>commons-io</groupId>
<artifactId>commons-io</artifactId>
</dependency>
<dependency>
<groupId>org.apache.spark</groupId>
<artifactId>spark-core_2.11</artifactId>
</dependency>
<dependency>
<groupId>org.apache.spark</groupId>
<artifactId>spark-sql_2.11</artifactId>
</dependency>
<dependency>
<groupId>org.apache.spark</groupId>
<artifactId>spark-hive_2.11</artifactId>
<scope>test</scope>
</dependency>
<dependency>
<groupId>eu.dnetlib.dhp</groupId>
<artifactId>dhp-common</artifactId>
<version>${project.version}</version>
</dependency>
<dependency>
<groupId>eu.dnetlib.dhp</groupId>
<artifactId>dhp-schemas</artifactId>
<version>${project.version}</version>
</dependency>
<dependency>
<groupId>com.jayway.jsonpath</groupId>
<artifactId>json-path</artifactId>
</dependency>
<dependency>
<groupId>dom4j</groupId>
<artifactId>dom4j</artifactId>
</dependency>
<dependency>
<groupId>jaxen</groupId>
<artifactId>jaxen</artifactId>
</dependency>
<dependency>
<groupId>eu.dnetlib</groupId>
<artifactId>dnet-openaire-broker-common</artifactId>
<version>[1.0.0,2.0.0)</version>
</dependency>
</dependencies>
</project>

81
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/raw/MigrateDbEntitiesApplication.java
View File

@ -51,7 +51,7 @@ import eu.dnetlib.dhp.schema.oaf.Software;
import eu.dnetlib.dhp.schema.oaf.StructuredProperty;
public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
implements Closeable {
implements Closeable {
private static final Log log = LogFactory.getLog(MigrateDbEntitiesApplication.class);
@ -61,10 +61,9 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
public static void main(final String[] args) throws Exception {
final ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
MigrateDbEntitiesApplication.class
.getResourceAsStream("/eu/dnetlib/dhp/oa/graph/migrate_db_entities_parameters.json")));
IOUtils
.toString(MigrateDbEntitiesApplication.class
.getResourceAsStream("/eu/dnetlib/dhp/oa/graph/migrate_db_entities_parameters.json")));
parser.parseArgument(args);
@ -77,7 +76,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
final boolean processClaims = parser.get("action") != null && parser.get("action").equalsIgnoreCase("claims");
try (final MigrateDbEntitiesApplication smdbe = new MigrateDbEntitiesApplication(hdfsPath, dbUrl, dbUser,
dbPassword)) {
dbPassword)) {
if (processClaims) {
log.info("Processing claims...");
smdbe.execute("queryClaims.sql", smdbe::processClaims);
@ -108,15 +107,15 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
}
public MigrateDbEntitiesApplication(
final String hdfsPath, final String dbUrl, final String dbUser, final String dbPassword)
throws Exception {
final String hdfsPath, final String dbUrl, final String dbUser, final String dbPassword)
throws Exception {
super(hdfsPath);
this.dbClient = new DbClient(dbUrl, dbUser, dbPassword);
this.lastUpdateTimestamp = new Date().getTime();
}
public void execute(final String sqlFile, final Function<ResultSet, List<Oaf>> producer)
throws Exception {
throws Exception {
final String sql = IOUtils.toString(getClass().getResourceAsStream("/eu/dnetlib/dhp/oa/graph/sql/" + sqlFile));
final Consumer<ResultSet> consumer = rs -> producer.apply(rs).forEach(oaf -> emitOaf(oaf));
@ -135,10 +134,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
ds.setId(createOpenaireId(10, rs.getString("datasourceid"), true));
ds.setOriginalId(Arrays.asList(rs.getString("datasourceid")));
ds
.setCollectedfrom(
listKeyValues(
createOpenaireId(10, rs.getString("collectedfromid"), true),
rs.getString("collectedfromname")));
.setCollectedfrom(listKeyValues(createOpenaireId(10, rs.getString("collectedfromid"), true), rs.getString("collectedfromname")));
ds.setPid(new ArrayList<>());
ds.setDateofcollection(asString(rs.getDate("dateofcollection")));
ds.setDateoftransformation(null); // Value not returned by the SQL query
@ -179,7 +175,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
ds.setCertificates(field(rs.getString("certificates"), info));
ds.setPolicies(new ArrayList<>()); // The sql query returns an empty array
ds
.setJournal(prepareJournal(rs.getString("officialname"), rs.getString("journal"), info)); // Journal
.setJournal(prepareJournal(rs.getString("officialname"), rs.getString("journal"), info)); // Journal
ds.setDataInfo(info);
ds.setLastupdatetimestamp(lastUpdateTimestamp);
@ -199,10 +195,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
p.setId(createOpenaireId(40, rs.getString("projectid"), true));
p.setOriginalId(Arrays.asList(rs.getString("projectid")));
p
.setCollectedfrom(
listKeyValues(
createOpenaireId(10, rs.getString("collectedfromid"), true),
rs.getString("collectedfromname")));
.setCollectedfrom(listKeyValues(createOpenaireId(10, rs.getString("collectedfromid"), true), rs.getString("collectedfromname")));
p.setPid(new ArrayList<>());
p.setDateofcollection(asString(rs.getDate("dateofcollection")));
p.setDateoftransformation(asString(rs.getDate("dateoftransformation")));
@ -219,7 +212,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
p.setDuration(field(Integer.toString(rs.getInt("duration")), info));
p.setEcsc39(field(Boolean.toString(rs.getBoolean("ecsc39")), info));
p
.setOamandatepublications(field(Boolean.toString(rs.getBoolean("oamandatepublications")), info));
.setOamandatepublications(field(Boolean.toString(rs.getBoolean("oamandatepublications")), info));
p.setEcarticle29_3(field(Boolean.toString(rs.getBoolean("ecarticle29_3")), info));
p.setSubjects(prepareListOfStructProps(rs.getArray("subjects"), info));
p.setFundingtree(prepareListFields(rs.getArray("fundingtree"), info));
@ -256,10 +249,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
o.setId(createOpenaireId(20, rs.getString("organizationid"), true));
o.setOriginalId(Arrays.asList(rs.getString("organizationid")));
o
.setCollectedfrom(
listKeyValues(
createOpenaireId(10, rs.getString("collectedfromid"), true),
rs.getString("collectedfromname")));
.setCollectedfrom(listKeyValues(createOpenaireId(10, rs.getString("collectedfromid"), true), rs.getString("collectedfromname")));
o.setPid(new ArrayList<>());
o.setDateofcollection(asString(rs.getDate("dateofcollection")));
o.setDateoftransformation(asString(rs.getDate("dateoftransformation")));
@ -274,14 +264,12 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
o.setEclegalperson(field(Boolean.toString(rs.getBoolean("eclegalperson")), info));
o.setEcnonprofit(field(Boolean.toString(rs.getBoolean("ecnonprofit")), info));
o
.setEcresearchorganization(field(Boolean.toString(rs.getBoolean("ecresearchorganization")), info));
.setEcresearchorganization(field(Boolean.toString(rs.getBoolean("ecresearchorganization")), info));
o.setEchighereducation(field(Boolean.toString(rs.getBoolean("echighereducation")), info));
o
.setEcinternationalorganizationeurinterests(
field(Boolean.toString(rs.getBoolean("ecinternationalorganizationeurinterests")), info));
.setEcinternationalorganizationeurinterests(field(Boolean.toString(rs.getBoolean("ecinternationalorganizationeurinterests")), info));
o
.setEcinternationalorganization(
field(Boolean.toString(rs.getBoolean("ecinternationalorganization")), info));
.setEcinternationalorganization(field(Boolean.toString(rs.getBoolean("ecinternationalorganization")), info));
o.setEcenterprise(field(Boolean.toString(rs.getBoolean("ecenterprise")), info));
o.setEcsmevalidated(field(Boolean.toString(rs.getBoolean("ecsmevalidated")), info));
o.setEcnutscode(field(Boolean.toString(rs.getBoolean("ecnutscode")), info));
@ -300,8 +288,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
final DataInfo info = prepareDataInfo(rs);
final String orgId = createOpenaireId(20, rs.getString("organization"), true);
final String dsId = createOpenaireId(10, rs.getString("datasource"), true);
final List<KeyValue> collectedFrom = listKeyValues(
createOpenaireId(10, rs.getString("collectedfromid"), true), rs.getString("collectedfromname"));
final List<KeyValue> collectedFrom = listKeyValues(createOpenaireId(10, rs.getString("collectedfromid"), true), rs.getString("collectedfromname"));
final Relation r1 = new Relation();
r1.setRelType("datasourceOrganization");
@ -334,8 +321,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
final DataInfo info = prepareDataInfo(rs);
final String orgId = createOpenaireId(20, rs.getString("resporganization"), true);
final String projectId = createOpenaireId(40, rs.getString("project"), true);
final List<KeyValue> collectedFrom = listKeyValues(
createOpenaireId(10, rs.getString("collectedfromid"), true), rs.getString("collectedfromname"));
final List<KeyValue> collectedFrom = listKeyValues(createOpenaireId(10, rs.getString("collectedfromid"), true), rs.getString("collectedfromname"));
final Relation r1 = new Relation();
r1.setRelType("projectOrganization");
@ -365,12 +351,10 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
public List<Oaf> processClaims(final ResultSet rs) {
final DataInfo info = dataInfo(
false, null, false, false,
qualifier("user:claim", "user:claim", "dnet:provenanceActions", "dnet:provenanceActions"), "0.9");
final DataInfo info =
dataInfo(false, null, false, false, qualifier("user:claim", "user:claim", "dnet:provenanceActions", "dnet:provenanceActions"), "0.9");
final List<KeyValue> collectedFrom = listKeyValues(
createOpenaireId(10, "infrastruct_::openaire", true), "OpenAIRE");
final List<KeyValue> collectedFrom = listKeyValues(createOpenaireId(10, "infrastruct_::openaire", true), "OpenAIRE");
try {
@ -456,15 +440,11 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
final String inferenceprovenance = rs.getString("inferenceprovenance");
final Boolean inferred = rs.getBoolean("inferred");
final String trust = rs.getString("trust");
return dataInfo(
deletedbyinference, inferenceprovenance, inferred, false,
MigrationConstants.ENTITYREGISTRY_PROVENANCE_ACTION, trust);
return dataInfo(deletedbyinference, inferenceprovenance, inferred, false, MigrationConstants.ENTITYREGISTRY_PROVENANCE_ACTION, trust);
}
private Qualifier prepareQualifierSplitting(final String s) {
if (StringUtils.isBlank(s)) {
return null;
}
if (StringUtils.isBlank(s)) { return null; }
final String[] arr = s.split("@@@");
return arr.length == 4 ? qualifier(arr[0], arr[1], arr[2], arr[3]) : null;
}
@ -478,23 +458,19 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
}
private StructuredProperty prepareStructProp(final String s, final DataInfo dataInfo) {
if (StringUtils.isBlank(s)) {
return null;
}
if (StringUtils.isBlank(s)) { return null; }
final String[] parts = s.split("###");
if (parts.length == 2) {
final String value = parts[0];
final String[] arr = parts[1].split("@@@");
if (arr.length == 4) {
return structuredProperty(value, arr[0], arr[1], arr[2], arr[3], dataInfo);
}
if (arr.length == 4) { return structuredProperty(value, arr[0], arr[1], arr[2], arr[3], dataInfo); }
}
return null;
}
private List<StructuredProperty> prepareListOfStructProps(
final Array array,
final DataInfo dataInfo) throws SQLException {
final Array array,
final DataInfo dataInfo) throws SQLException {
final List<StructuredProperty> res = new ArrayList<>();
if (array != null) {
for (final String s : (String[]) array.getArray()) {
@ -514,9 +490,8 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
if (arr.length == 3) {
final String issn = StringUtils.isNotBlank(arr[0]) ? arr[0].trim() : null;
final String eissn = StringUtils.isNotBlank(arr[1]) ? arr[1].trim() : null;
;
final String lissn = StringUtils.isNotBlank(arr[2]) ? arr[2].trim() : null;
;
if (issn != null || eissn != null || lissn != null) {
return journal(name, issn, eissn, eissn, null, null, null, null, null, null, null, info);
}

1
dhp-workflows/pom.xml
View File

@ -26,6 +26,7 @@
<module>dhp-dedup-scholexplorer</module>
<module>dhp-graph-provision-scholexplorer</module>
<module>dhp-stats-update</module>
<module>dhp-broker-events</module>
</modules>
<pluginRepositories>