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

changed message in log

This commit is contained in:
Miriam Baglioni 2020-05-29 10:41:39 +02:00
parent 8b6e886fb6
commit 6f1eea28b6
1 changed files with 81 additions and 79 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

160
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/project/ReadProjectsFromDB.java
View File

@ -1,15 +1,5 @@
package eu.dnetlib.dhp.actionmanager.project;
import com.fasterxml.jackson.databind.ObjectMapper; package eu.dnetlib.dhp.actionmanager.project;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.common.DbClient;
import org.apache.commons.io.IOUtils;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FSDataOutputStream;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import java.io.BufferedWriter; import java.io.BufferedWriter;
import java.io.Closeable; import java.io.Closeable;
@ -22,92 +12,104 @@ import java.util.List;
import java.util.function.Consumer; import java.util.function.Consumer;
import java.util.function.Function; import java.util.function.Function;
import org.apache.commons.io.IOUtils;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FSDataOutputStream;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.common.DbClient;
public class ReadProjectsFromDB implements Closeable { public class ReadProjectsFromDB implements Closeable {
private final DbClient dbClient; private final DbClient dbClient;
private static final Log log = LogFactory.getLog(ReadProjectsFromDB.class); private static final Log log = LogFactory.getLog(ReadProjectsFromDB.class);
private final Configuration conf; private final Configuration conf;
private final BufferedWriter writer; private final BufferedWriter writer;
private final ObjectMapper OBJECT_MAPPER = new ObjectMapper(); private final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
private final static String query = "SELECT code " + private final static String query = "SELECT code " +
"from projects where id like 'corda__h2020%' " ; "from projects where id like 'corda__h2020%' ";
public static void main(final String[] args) throws Exception { public static void main(final String[] args) throws Exception {
final ArgumentApplicationParser parser = new ArgumentApplicationParser( final ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils IOUtils
.toString( .toString(
ReadProjectsFromDB.class ReadProjectsFromDB.class
.getResourceAsStream( .getResourceAsStream(
"/eu/dnetlib/dhp/actionmanager/project/read_projects_db.json"))); "/eu/dnetlib/dhp/actionmanager/project/read_projects_db.json")));
parser.parseArgument(args); parser.parseArgument(args);
final String dbUrl = parser.get("postgresUrl"); final String dbUrl = parser.get("postgresUrl");
final String dbUser = parser.get("postgresUser"); final String dbUser = parser.get("postgresUser");
final String dbPassword = parser.get("postgresPassword"); final String dbPassword = parser.get("postgresPassword");
final String hdfsPath = parser.get("hdfsPath") ; final String hdfsPath = parser.get("hdfsPath");
final String hdfsNameNode = parser.get("hdfsNameNode"); final String hdfsNameNode = parser.get("hdfsNameNode");
try (final ReadProjectsFromDB rbl = new ReadProjectsFromDB(hdfsPath, hdfsNameNode, dbUrl, dbUser, try (final ReadProjectsFromDB rbl = new ReadProjectsFromDB(hdfsPath, hdfsNameNode, dbUrl, dbUser,
dbPassword)) { dbPassword)) {
log.info("Processing blacklist..."); log.info("Processing projects...");
rbl.execute(query, rbl::processProjectsEntry); rbl.execute(query, rbl::processProjectsEntry);
} }
} }
public void execute(final String sql, final Function<ResultSet, List<ProjectSubset>> producer) throws Exception {
final Consumer<ResultSet> consumer = rs -> producer.apply(rs).forEach(r -> writeProject(r)); public void execute(final String sql, final Function<ResultSet, List<ProjectSubset>> producer) throws Exception {
dbClient.processResults(sql, consumer); final Consumer<ResultSet> consumer = rs -> producer.apply(rs).forEach(r -> writeProject(r));
}
public List<ProjectSubset> processProjectsEntry(ResultSet rs) { dbClient.processResults(sql, consumer);
try { }
ProjectSubset p = new ProjectSubset();
p.setCode(rs.getString("code"));
return Arrays.asList(p); public List<ProjectSubset> processProjectsEntry(ResultSet rs) {
try {
ProjectSubset p = new ProjectSubset();
p.setCode(rs.getString("code"));
} catch (final Exception e) { return Arrays.asList(p);
throw new RuntimeException(e);
}
}
protected void writeProject(final ProjectSubset r) { } catch (final Exception e) {
try { throw new RuntimeException(e);
writer.write(OBJECT_MAPPER.writeValueAsString(r)); }
writer.newLine(); }
} catch (final Exception e) {
throw new RuntimeException(e);
}
}
public ReadProjectsFromDB( protected void writeProject(final ProjectSubset r) {
final String hdfsPath, String hdfsNameNode, final String dbUrl, final String dbUser, final String dbPassword) try {
throws Exception { writer.write(OBJECT_MAPPER.writeValueAsString(r));
writer.newLine();
} catch (final Exception e) {
throw new RuntimeException(e);
}
}
this.dbClient = new DbClient(dbUrl, dbUser, dbPassword); public ReadProjectsFromDB(
this.conf = new Configuration(); final String hdfsPath, String hdfsNameNode, final String dbUrl, final String dbUser, final String dbPassword)
this.conf.set("fs.defaultFS", hdfsNameNode); throws Exception {
FileSystem fileSystem = FileSystem.get(this.conf);
Path hdfsWritePath = new Path(hdfsPath);
FSDataOutputStream fsDataOutputStream = null;
if (fileSystem.exists(hdfsWritePath)) {
fileSystem.delete(hdfsWritePath, false);
}
fsDataOutputStream = fileSystem.create(hdfsWritePath);
this.dbClient = new DbClient(dbUrl, dbUser, dbPassword);
this.conf = new Configuration();
this.conf.set("fs.defaultFS", hdfsNameNode);
FileSystem fileSystem = FileSystem.get(this.conf);
Path hdfsWritePath = new Path(hdfsPath);
FSDataOutputStream fsDataOutputStream = null;
if (fileSystem.exists(hdfsWritePath)) {
fileSystem.delete(hdfsWritePath, false);
}
fsDataOutputStream = fileSystem.create(hdfsWritePath);
this.writer = new BufferedWriter(new OutputStreamWriter(fsDataOutputStream, StandardCharsets.UTF_8)); this.writer = new BufferedWriter(new OutputStreamWriter(fsDataOutputStream, StandardCharsets.UTF_8));
} }
@Override @Override
public void close() throws IOException { public void close() throws IOException {
dbClient.close(); dbClient.close();
writer.close(); writer.close();
} }
} }