pispis
/
dnet-hadoop
forked from antonis.lempesis/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

cleaned up poms, added descriptions

This commit is contained in:
Claudio Atzori 2020-04-29 18:44:17 +02:00
parent 64d790a266
commit 77ac995770
12 changed files with 145 additions and 129 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
dhp-build/dhp-build-assembly-resources/pom.xml
View File

@ -12,6 +12,8 @@
<artifactId>dhp-build-assembly-resources</artifactId>
<packaging>jar</packaging>
<description>This module contains a set of scripts supporting the build lifecycle for the dnet-hadoop project</description>
<build>
<plugins>
<plugin>

38
dhp-build/dhp-build-properties-maven-plugin/pom.xml
View File

@ -12,6 +12,7 @@
<artifactId>dhp-build-properties-maven-plugin</artifactId>
<packaging>maven-plugin</packaging>
<description>This module is a maven plugin implementing custom properties substitutions in the build lifecycle</description>
<dependencies>
<dependency>
@ -33,7 +34,7 @@
<dependency>
<groupId>org.kuali.maven.plugins</groupId>
<artifactId>properties-maven-plugin</artifactId>
<version>2.0.1</version>
<version>${properties.maven.plugin.version}</version>
</dependency>
<dependency>
<groupId>com.google.code.findbugs</groupId>
@ -83,41 +84,6 @@
</configuration>
</plugin>
</plugins>
<pluginManagement>
<plugins>
<!--This plugin's configuration is used to store Eclipse m2e settings only. It has no influence on the Maven build itself.-->
<plugin>
<groupId>org.eclipse.m2e</groupId>
<artifactId>lifecycle-mapping</artifactId>
<version>1.0.0</version>
<configuration>
<lifecycleMappingMetadata>
<pluginExecutions>
<pluginExecution>
<pluginExecutionFilter>
<groupId>
org.apache.maven.plugins
</groupId>
<artifactId>
maven-plugin-plugin
</artifactId>
<versionRange>
[3.2,)
</versionRange>
<goals>
<goal>descriptor</goal>
</goals>
</pluginExecutionFilter>
<action>
<ignore />
</action>
</pluginExecution>
</pluginExecutions>
</lifecycleMappingMetadata>
</configuration>
</plugin>
</plugins>
</pluginManagement>
</build>
</project>

32
dhp-build/dhp-code-style/pom.xml
View File

@ -11,6 +11,38 @@
<packaging>jar</packaging>
<description>This module contains resources supporting common code style conventions</description>
<distributionManagement>
<snapshotRepository>
<id>dnet45-snapshots</id>
<name>DNet45 Snapshots</name>
<url>http://maven.research-infrastructures.eu/nexus/content/repositories/dnet45-snapshots</url>
<layout>default</layout>
</snapshotRepository>
<repository>
<id>dnet45-releases</id>
<url>http://maven.research-infrastructures.eu/nexus/content/repositories/dnet45-releases</url>
</repository>
</distributionManagement>
<build>
<pluginManagement>
<plugins>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-project-info-reports-plugin</artifactId>
<version>3.0.0</version>
</plugin>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-site-plugin</artifactId>
<version>3.7.1</version>
</plugin>
</plugins>
</pluginManagement>
</build>
<properties>
<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
</properties>

3
dhp-build/pom.xml
View File

@ -8,6 +8,9 @@
</parent>
<artifactId>dhp-build</artifactId>
<packaging>pom</packaging>
<description>This module is a container for the build tools used in dnet-hadoop</description>
<modules>
<module>dhp-code-style</module>
<module>dhp-build-assembly-resources</module>

2
dhp-common/pom.xml
View File

@ -12,6 +12,8 @@
<artifactId>dhp-common</artifactId>
<packaging>jar</packaging>
<description>This module contains common utilities meant to be used across the dnet-hadoop submodules</description>
<dependencies>
<dependency>

2
dhp-schemas/pom.xml
View File

@ -12,7 +12,7 @@
<artifactId>dhp-schemas</artifactId>
<packaging>jar</packaging>
<description>This module contains common schema classes meant to be used across the dnet-hadoop submodules</description>
<dependencies>

1
dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java
View File

@ -1,3 +1,4 @@
package eu.dnetlib.dhp.oa.dedup;
import static java.nio.file.Files.createTempDirectory;

84
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);
@ -61,9 +61,10 @@ 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);
@ -76,7 +77,7 @@ 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);
@ -107,15 +108,15 @@ 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));
@ -134,7 +135,10 @@ 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
@ -175,7 +179,7 @@ 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);
@ -195,7 +199,10 @@ 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")));
@ -212,7 +219,7 @@ 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));
@ -249,7 +256,10 @@ 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")));
@ -264,12 +274,14 @@ 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));
@ -288,7 +300,8 @@ 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");
@ -321,7 +334,8 @@ 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");
@ -351,10 +365,12 @@ 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 {
@ -440,11 +456,15 @@ 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;
}
@ -458,19 +478,23 @@ 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()) {
@ -489,8 +513,10 @@ public class MigrateDbEntitiesApplication extends AbstractMigrationApplication
final String[] arr = sj.split("@@@");
if (arr.length == 3) {
final String issn = StringUtils.isNotBlank(arr[0]) ? arr[0] : null;
final String eissn = StringUtils.isNotBlank(arr[1]) ? arr[1] : null;;
final String lissn = StringUtils.isNotBlank(arr[2]) ? arr[2] : null;;
final String eissn = StringUtils.isNotBlank(arr[1]) ? arr[1] : null;
;
final String lissn = StringUtils.isNotBlank(arr[2]) ? arr[2] : 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/dhp-stats-update/pom.xml
View File

@ -22,6 +22,7 @@
<plugin>
<groupId>pl.project13.maven</groupId>
<artifactId>git-commit-id-plugin</artifactId>
<version>2.1.11</version>
<configuration>
<failOnNoGitDirectory>false</failOnNoGitDirectory>
</configuration>

16
dhp-workflows/dhp-worfklow-profiles/pom.xml
View File

@ -3,13 +3,27 @@
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</artifactId>
<artifactId>dhp-workflows</artifactId>
<groupId>eu.dnetlib.dhp</groupId>
<version>1.1.7-SNAPSHOT</version>
</parent>
<modelVersion>4.0.0</modelVersion>
<artifactId>dhp-worfklow-profiles</artifactId>
<packaging>jar</packaging>
<!--
<build>
<pluginManagement>
<plugins>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-site-plugin</artifactId>
<version>3.7.1</version>
</plugin>
</plugins>
</pluginManagement>
</build>
-->
</project>

60
dhp-workflows/pom.xml
View File

@ -13,6 +13,8 @@
<artifactId>dhp-workflows</artifactId>
<packaging>pom</packaging>
<description>This module is the container for the oozie workflow definitions in dnet-hadoop project</description>
<modules>
<module>dhp-worfklow-profiles</module>
<module>dhp-aggregation</module>
@ -182,7 +184,7 @@
<plugin>
<groupId>org.kuali.maven.plugins</groupId>
<artifactId>properties-maven-plugin</artifactId>
<version>1.3.2</version>
<version>${properties.maven.plugin.version}</version>
<dependencies>
<dependency>
<groupId>eu.dnetlib.dhp</groupId>
@ -562,60 +564,4 @@
</profile>
</profiles>
<build>
<pluginManagement>
<plugins>
<!--This plugin's configuration is used to store Eclipse m2e settings only. It has no influence on the Maven build itself. -->
<plugin>
<groupId>org.eclipse.m2e</groupId>
<artifactId>lifecycle-mapping</artifactId>
<version>1.0.0</version>
<configuration>
<lifecycleMappingMetadata>
<pluginExecutions>
<pluginExecution>
<pluginExecutionFilter>
<groupId>
org.kuali.maven.plugins
</groupId>
<artifactId>
properties-maven-plugin
</artifactId>
<versionRange>
[1.3.2,)
</versionRange>
<goals>
<goal>
read-project-properties
</goal>
<goal>
write-project-properties
</goal>
</goals>
</pluginExecutionFilter>
<action>
<ignore />
</action>
</pluginExecution>
<!-- copy-dependency plugin -->
<pluginExecution>
<pluginExecutionFilter>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-dependency-plugin</artifactId>
<versionRange>[1.0.0,)</versionRange>
<goals>
<goal>copy-dependencies</goal>
</goals>
</pluginExecutionFilter>
<action>
<ignore />
</action>
</pluginExecution>
</pluginExecutions>
</lifecycleMappingMetadata>
</configuration>
</plugin>
</plugins>
</pluginManagement>
</build>
</project>

33
pom.xml
View File

@ -6,14 +6,14 @@
<version>1.1.7-SNAPSHOT</version>
<packaging>pom</packaging>
<url>http://www.d-net.research-infrastructures.eu</url>
<licenses>
<license>
<name>The Apache Software License, Version 2.0</name>
<url>http://www.apache.org/licenses/LICENSE-2.0.txt</url>
<name>GNU Affero General Public License v3.0 or later</name>
<url>https://spdx.org/licenses/AGPL-3.0-or-later.html#licenseText</url>
<distribution>repo</distribution>
<comments>A business-friendly OSS license</comments>
<comments>This program is free software: you can redistribute it and/or modify it under the terms of the
GNU Affero General Public License as published by the Free Software Foundation, either version 3 of the
License, or (at your option) any later version.</comments>
</license>
</licenses>
@ -41,6 +41,8 @@
<tag>HEAD</tag>
</scm>
<description>This module is the root descriptor for the dnet-hadoop project</description>
<pluginRepositories>
</pluginRepositories>
@ -406,6 +408,18 @@
<testOutputDirectory>target/test-classes</testOutputDirectory>
<pluginManagement>
<plugins>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-project-info-reports-plugin</artifactId>
<version>3.0.0</version>
</plugin>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-site-plugin</artifactId>
<version>3.7.1</version>
</plugin>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-compiler-plugin</artifactId>
@ -476,6 +490,14 @@
</plugins>
</pluginManagement>
<plugins>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-site-plugin</artifactId>
</plugin>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-project-info-reports-plugin</artifactId>
</plugin>
<plugin>
<groupId>net.revelc.code.formatter</groupId>
<artifactId>formatter-maven-plugin</artifactId>
@ -583,6 +605,7 @@
<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
<maven.compiler.plugin.version>3.6.0</maven.compiler.plugin.version>
<maven.failsave.plugin.version>2.22.2</maven.failsave.plugin.version>
<properties.maven.plugin.version>2.0.1</properties.maven.plugin.version>
<dhp.cdh.version>cdh5.9.2</dhp.cdh.version>
<dhp.hadoop.version>2.6.0-${dhp.cdh.version}</dhp.hadoop.version>
<dhp.oozie.version>4.1.0-${dhp.cdh.version}</dhp.oozie.version>