#22002

Committed 01 Apr 2024 07:56PM CUT coverage: 20.716% (+0.5%) from 20.173%

Build # #22002

Build Type

push

github

Committed by

web-flow

Commit Message

Merge pull request #10453 from IQSS/develop

Merge 6.2 into master

Run Details

704 of 2679 new or added lines in 152 files covered. (26.28%)

81 existing lines in 49 files now uncovered.

17160 of 82836 relevant lines covered (20.72%)

0.21 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

0.0

/src/main/java/edu/harvard/iq/dataverse/engine/command/impl/ImportDatasetCommand.java

package edu.harvard.iq.dataverse.engine.command.impl;

import edu.harvard.iq.dataverse.Dataset;
import edu.harvard.iq.dataverse.engine.command.CommandContext;
import edu.harvard.iq.dataverse.engine.command.DataverseRequest;
import edu.harvard.iq.dataverse.engine.command.exception.CommandException;
import edu.harvard.iq.dataverse.engine.command.exception.CommandExecutionException;
import edu.harvard.iq.dataverse.engine.command.exception.IllegalCommandException;
import edu.harvard.iq.dataverse.engine.command.exception.PermissionException;
import edu.harvard.iq.dataverse.pidproviders.PidProvider;
import edu.harvard.iq.dataverse.pidproviders.PidUtil;

import java.io.IOException;
import java.util.Collections;
import java.util.logging.Level;
import java.util.logging.Logger;
import org.apache.commons.httpclient.HttpClient;
import org.apache.commons.httpclient.methods.GetMethod;
import static org.apache.commons.lang3.StringUtils.isEmpty;

/**
 * Imports a dataset from a different system. This command validates that the PID
 * of the new dataset exists, and then inserts the new dataset into the database.
 * 
 * <b>NOTE:</b> At the moment, this command only supports a single version in the 
 * dataset, and was tested with package files only.
 * 
 * @author michael
 */
public class ImportDatasetCommand extends AbstractCreateDatasetCommand {
    
    private static final Logger logger = Logger.getLogger(ImportDatasetCommand.class.getName());
    
    /**
     * Creates a new instance of the command.
     * @param theDataset The dataset we want to import.
     * @param aRequest Request context for the command.
     */
    public ImportDatasetCommand(Dataset theDataset, DataverseRequest aRequest) {
        super(theDataset, aRequest);
    }

    /**
     * Validate that the PID of the dataset, if any, exists.
     * @param ctxt
     * @throws CommandException 
     */
    @Override
    protected void additionalParameterTests(CommandContext ctxt) throws CommandException {
        
        if ( ! getUser().isSuperuser() ) {
            throw new PermissionException("ImportDatasetCommand can only be issued by a super-user.", this, Collections.emptySet(), getDataset());
        }
        
        Dataset ds = getDataset();
        
        if ( isEmpty(ds.getIdentifier()) ) {
            throw new IllegalCommandException("Imported datasets must have a persistent global identifier.", this);
        }
        
        if ( ! ctxt.dvObjects().isGlobalIdLocallyUnique(ds.getGlobalId()) ) {
            throw new IllegalCommandException("Persistent identifier " + ds.getGlobalId().asString() + " already exists in this Dataverse installation.", this);
        }
        
        String pid = ds.getPersistentURL();
        GetMethod httpGet = new GetMethod(pid); 
        httpGet.setFollowRedirects(false);

        HttpClient client = new HttpClient();

        try {
            int responseStatus = client.executeMethod(httpGet);

            if (responseStatus == 404) {
                /*
                 * Using test DOIs from DataCite, we'll get a 404 when trying to resolve the DOI
                 * to a landing page, but the DOI may already exist. An extra check here allows
                 * use of DataCite test DOIs. It also changes import slightly in allowing PIDs
                 * that exist (and accessible in the PID provider account configured in
                 * Dataverse) but aren't findable to be used. That could be the case if, for
                 * example, someone was importing a draft dataset from elsewhere.
                 */
                PidProvider pidProvider = PidUtil.getPidProvider(ds.getGlobalId().getProviderId());
                if (pidProvider != null) {
                    if (pidProvider.alreadyRegistered(ds.getGlobalId(), true)) {
                        return;
                    }
                }
                throw new CommandExecutionException(
                        "Provided PID does not exist. Status code for GET '" + pid + "' is 404.", this);
            }

        } catch (Exception ex) {
            logger.log(Level.WARNING,
                    "Error while validating PID at '" + pid + "' for an imported dataset: " + ex.getMessage(), ex);
            throw new CommandExecutionException("Cannot validate PID due to an error: " + ex.getMessage(), this);
        }

    }
    
    @Override
    protected void handlePid(Dataset theDataset, CommandContext ctxt) {
        theDataset.setGlobalIdCreateTime( getTimestamp() );
    }

    @Override
    protected void postPersist(Dataset theDataset, CommandContext ctxt) throws CommandException {
    }
    
    
    
}

1	package edu.harvard.iq.dataverse.engine.command.impl;
2
3	import edu.harvard.iq.dataverse.Dataset;
4	import edu.harvard.iq.dataverse.engine.command.CommandContext;
5	import edu.harvard.iq.dataverse.engine.command.DataverseRequest;
6	import edu.harvard.iq.dataverse.engine.command.exception.CommandException;
7	import edu.harvard.iq.dataverse.engine.command.exception.CommandExecutionException;
8	import edu.harvard.iq.dataverse.engine.command.exception.IllegalCommandException;
9	import edu.harvard.iq.dataverse.engine.command.exception.PermissionException;
10	import edu.harvard.iq.dataverse.pidproviders.PidProvider;
11	import edu.harvard.iq.dataverse.pidproviders.PidUtil;
12
13	import java.io.IOException;
14	import java.util.Collections;
15	import java.util.logging.Level;
16	import java.util.logging.Logger;
17	import org.apache.commons.httpclient.HttpClient;
18	import org.apache.commons.httpclient.methods.GetMethod;
19	import static org.apache.commons.lang3.StringUtils.isEmpty;
20
21	/**
22	* Imports a dataset from a different system. This command validates that the PID
23	* of the new dataset exists, and then inserts the new dataset into the database.
24	*
25	* <b>NOTE:</b> At the moment, this command only supports a single version in the
26	* dataset, and was tested with package files only.
27	*
28	* @author michael
29	*/
30	public class ImportDatasetCommand extends AbstractCreateDatasetCommand {
31
32	private static final Logger logger = Logger.getLogger(ImportDatasetCommand.class.getName());	×
33
34	/**
35	* Creates a new instance of the command.
36	* @param theDataset The dataset we want to import.
37	* @param aRequest Request context for the command.
38	*/
39	public ImportDatasetCommand(Dataset theDataset, DataverseRequest aRequest) {
40	super(theDataset, aRequest);	×
41	}	×
42
43	/**
44	* Validate that the PID of the dataset, if any, exists.
45	* @param ctxt
46	* @throws CommandException
47	*/
48	@Override
49	protected void additionalParameterTests(CommandContext ctxt) throws CommandException {
50
51	if ( ! getUser().isSuperuser() ) {	×
52	throw new PermissionException("ImportDatasetCommand can only be issued by a super-user.", this, Collections.emptySet(), getDataset());	×
53	}
54
55	Dataset ds = getDataset();	×
56
57	if ( isEmpty(ds.getIdentifier()) ) {	×
58	throw new IllegalCommandException("Imported datasets must have a persistent global identifier.", this);	×
59	}
60
61	if ( ! ctxt.dvObjects().isGlobalIdLocallyUnique(ds.getGlobalId()) ) {	×
62	throw new IllegalCommandException("Persistent identifier " + ds.getGlobalId().asString() + " already exists in this Dataverse installation.", this);	×
63	}
64
65	String pid = ds.getPersistentURL();	×
66	GetMethod httpGet = new GetMethod(pid);	×
67	httpGet.setFollowRedirects(false);	×
68
69	HttpClient client = new HttpClient();	×
70
71	try {
72	int responseStatus = client.executeMethod(httpGet);	×
73
74	if (responseStatus == 404) {	×
75	/*
76	* Using test DOIs from DataCite, we'll get a 404 when trying to resolve the DOI
77	* to a landing page, but the DOI may already exist. An extra check here allows
78	* use of DataCite test DOIs. It also changes import slightly in allowing PIDs
79	* that exist (and accessible in the PID provider account configured in
80	* Dataverse) but aren't findable to be used. That could be the case if, for
81	* example, someone was importing a draft dataset from elsewhere.
82	*/
NEW 83	PidProvider pidProvider = PidUtil.getPidProvider(ds.getGlobalId().getProviderId());	×
NEW 84	if (pidProvider != null) {	×
NEW 85	if (pidProvider.alreadyRegistered(ds.getGlobalId(), true)) {	×
UNCOV 86	return;	×
87	}
88	}
89	throw new CommandExecutionException(	×
90	"Provided PID does not exist. Status code for GET '" + pid + "' is 404.", this);
91	}
92
93	} catch (Exception ex) {	×
94	logger.log(Level.WARNING,	×
95	"Error while validating PID at '" + pid + "' for an imported dataset: " + ex.getMessage(), ex);	×
96	throw new CommandExecutionException("Cannot validate PID due to an error: " + ex.getMessage(), this);	×
97	}	×
98
99	}	×
100
101	@Override
102	protected void handlePid(Dataset theDataset, CommandContext ctxt) {
103	theDataset.setGlobalIdCreateTime( getTimestamp() );	×
104	}	×
105
106	@Override
107	protected void postPersist(Dataset theDataset, CommandContext ctxt) throws CommandException {
108	}	×
109
110
111
112	}

IQSS / dataverse / #22002

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous