You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
221 lines
6.7 KiB
Java
221 lines
6.7 KiB
Java
package eu.dnetlib.ariadneplus.workflows.nodes;
|
|
|
|
import eu.dnetlib.msro.workflows.graph.Arc;
|
|
import eu.dnetlib.msro.workflows.nodes.AsyncJobNode;
|
|
import eu.dnetlib.msro.workflows.nodes.is.ValidateProfilesJobNode;
|
|
import eu.dnetlib.msro.workflows.procs.Env;
|
|
import eu.dnetlib.msro.workflows.procs.Token;
|
|
import eu.dnetlib.msro.workflows.util.ProgressProvider;
|
|
import eu.dnetlib.rmi.manager.MSROException;
|
|
import org.apache.commons.logging.Log;
|
|
import org.apache.commons.logging.LogFactory;
|
|
import org.apache.http.HttpEntity;
|
|
import org.apache.http.HttpResponse;
|
|
import org.apache.http.client.HttpClient;
|
|
import org.apache.http.client.methods.HttpGet;
|
|
import org.apache.http.client.methods.HttpPost;
|
|
import org.apache.http.client.utils.URIBuilder;
|
|
import org.apache.http.impl.client.HttpClients;
|
|
import org.apache.http.util.EntityUtils;
|
|
|
|
import java.io.IOException;
|
|
import java.net.URI;
|
|
import java.net.URISyntaxException;
|
|
import java.util.Arrays;
|
|
import java.util.List;
|
|
|
|
|
|
public class IndexOnESJobNode extends AsyncJobNode {
|
|
|
|
private static final Log log = LogFactory.getLog(IndexOnESJobNode.class);
|
|
|
|
private String publisherEndpoint;
|
|
private String datasourceInterface;
|
|
private String datasource;
|
|
private int currentResourceToIndex = 0;
|
|
private int totalResourceToIndex = 0;
|
|
|
|
// @Override
|
|
// protected String execute(final Env env) throws Exception {
|
|
//
|
|
// int statusCode = -1;
|
|
// String response = "";
|
|
// log.info("IndexOnES endpoint: " + getIndexOnESEndpoint());
|
|
// HttpClient client = null;
|
|
// try {
|
|
// String[] splits = getDatasourceInterface().split("::");
|
|
// String datasource = splits[2];
|
|
// String collectionId = splits[3];
|
|
//
|
|
// URI getURI = new URIBuilder(getIndexOnESEndpoint())
|
|
// .addParameter("datasource", datasource)
|
|
// .addParameter("collectionId", collectionId)
|
|
// .build();
|
|
// client = HttpClients.createDefault();
|
|
// HttpResponse res = client.execute(new HttpGet(getURI));
|
|
// response = EntityUtils.toString(res.getEntity());
|
|
// if (res.getStatusLine()!=null) {
|
|
// statusCode = res.getStatusLine().getStatusCode();
|
|
// }
|
|
//
|
|
// }
|
|
// catch (Throwable t) {
|
|
// log.error(t);
|
|
// throw new MSROException("Indexing on Elastic Search: " + t.getMessage());
|
|
// }
|
|
//
|
|
// finally{
|
|
// }
|
|
//
|
|
// env.setAttribute(WorkflowsConstants.MAIN_LOG_PREFIX + "statusCode", Integer.toString(statusCode));
|
|
// env.setAttribute(WorkflowsConstants.MAIN_LOG_PREFIX + "response", response);
|
|
//
|
|
// if (statusCode!=200) {
|
|
// throw new MSROException("Error from Publisher endpoint [ status code: " + statusCode + " ]");
|
|
// }
|
|
//
|
|
// return Arc.DEFAULT_ARC;
|
|
// }
|
|
|
|
@Override
|
|
protected String execute(final Env env) throws Exception {
|
|
|
|
final String collectionResourceType = "COLLECTION";
|
|
final String recordResourceType = "RECORD";
|
|
try {
|
|
String[] splits = getDatasourceInterface().split("::");
|
|
String datasource = splits[2];
|
|
String collectionId = splits[3];
|
|
|
|
List<String> collectionIdentifiers = selectIdentifiers(datasource, collectionId, collectionResourceType);
|
|
if (!collectionIdentifiers.isEmpty()) {
|
|
collectionIdentifiers.forEach(identifier -> {
|
|
try {
|
|
indexing(datasource, collectionId, collectionResourceType, cleanIdentifier(identifier));
|
|
} catch (Throwable t) {
|
|
log.error(identifier+" "+t);
|
|
}
|
|
});
|
|
}
|
|
|
|
List<String> recordIdentifiers = selectIdentifiers(datasource, collectionId, recordResourceType);
|
|
if (!recordIdentifiers.isEmpty()) {
|
|
recordIdentifiers.forEach(identifier -> {
|
|
try {
|
|
indexing(datasource, collectionId, recordResourceType, cleanIdentifier(identifier));
|
|
} catch (Throwable t) {
|
|
log.error(identifier+" "+t);
|
|
}
|
|
});
|
|
}
|
|
}
|
|
catch (Throwable t) {
|
|
log.error(t);
|
|
throw new MSROException("Indexing on Elastic Search: " + t.getMessage());
|
|
}
|
|
|
|
return Arc.DEFAULT_ARC;
|
|
}
|
|
|
|
public String getPublisherEndpoint() {
|
|
return publisherEndpoint;
|
|
}
|
|
|
|
private String getIndexOnESEndpoint() {
|
|
return publisherEndpoint.concat("/indexOnESByIdentifier");
|
|
}
|
|
|
|
private String getSelectIdentifiersEndpoint() {
|
|
return publisherEndpoint.concat("/selectIdentifiers");
|
|
}
|
|
|
|
public void setPublisherEndpoint(final String publisherEndpoint) {
|
|
this.publisherEndpoint = publisherEndpoint;
|
|
}
|
|
|
|
public String getDatasourceInterface() {
|
|
return datasourceInterface;
|
|
}
|
|
|
|
public void setDatasourceInterface(String datasourceInterface) {
|
|
this.datasourceInterface = datasourceInterface;
|
|
}
|
|
|
|
public String getDatasource() {
|
|
return datasource;
|
|
}
|
|
|
|
public void setDatasource(String datasource) {
|
|
this.datasource = datasource;
|
|
}
|
|
|
|
private String cleanIdentifier(String identifier) {
|
|
String cleaned = identifier;
|
|
try {
|
|
cleaned = identifier
|
|
.replace("[", "")
|
|
.replace("]", "")
|
|
.replace("\"", "");
|
|
}
|
|
catch (Exception e) {
|
|
|
|
}
|
|
return cleaned;
|
|
}
|
|
|
|
private String indexing(String datasource, String collectionId, String resourceType, String identifier) throws IOException, URISyntaxException {
|
|
int statusCode = -1;
|
|
String response = "";
|
|
String result = "";
|
|
String endpoint = getIndexOnESEndpoint();
|
|
HttpClient client = null;
|
|
URI postURI = new URIBuilder(endpoint)
|
|
.addParameter("datasource", datasource)
|
|
.addParameter("collectionId", collectionId)
|
|
.addParameter("resourceType", resourceType)
|
|
.addParameter("identifier", identifier)
|
|
.build();
|
|
client = HttpClients.createDefault();
|
|
HttpResponse res = client.execute(new HttpPost(postURI));
|
|
if (res.getStatusLine()!=null) {
|
|
statusCode = res.getStatusLine().getStatusCode();
|
|
}
|
|
HttpEntity entity = res.getEntity();
|
|
result = EntityUtils.toString(entity);
|
|
return result;
|
|
}
|
|
|
|
private List<String> selectIdentifiers(String datasource, String collectionId, String resourceType) throws Exception {
|
|
int statusCode = -1;
|
|
String response = "";
|
|
List<String> identifiers = null;
|
|
String endpoint = getSelectIdentifiersEndpoint();
|
|
HttpClient client = null;
|
|
URI getURI = new URIBuilder(endpoint)
|
|
.addParameter("datasource", datasource)
|
|
.addParameter("collectionId", collectionId)
|
|
.addParameter("resourceType", resourceType)
|
|
.build();
|
|
client = HttpClients.createDefault();
|
|
HttpResponse res = client.execute(new HttpGet(getURI));
|
|
if (res.getStatusLine()!=null) {
|
|
statusCode = res.getStatusLine().getStatusCode();
|
|
}
|
|
HttpEntity entity = res.getEntity();
|
|
String content = EntityUtils.toString(entity);
|
|
String[] identifiersStr = content.split(",");
|
|
identifiers = Arrays.asList(identifiersStr);
|
|
return identifiers;
|
|
}
|
|
|
|
@Override
|
|
protected void beforeStart(final Token token) {
|
|
token.setProgressProvider(new ProgressProvider() {
|
|
@Override
|
|
public String getProgressDescription() {
|
|
return IndexOnESJobNode.this.currentResourceToIndex + " / " + IndexOnESJobNode.this.totalResourceToIndex;
|
|
}
|
|
});
|
|
}
|
|
}
|