Merge pull request #10246 from IQSS/10240-file-citation

file citation via API

Author: sekmiller

doc/release-notes/10240-file-citation.md

@@ -0,0 +1,5 @@
+## Get file citation as JSON
+
+It is now possible to retrieve via API the file citation as it appears on the file landing page. It is formatted in HTML and encoded in JSON.
+
+This API is not for downloading various citation formats such as EndNote XML, RIS, or BibTeX. This functionality has been requested in https://github.com/IQSS/dataverse/issues/3140 and https://github.com/IQSS/dataverse/issues/9994

doc/sphinx-guides/source/api/native-api.rst

@@ -845,7 +845,12 @@ Datasets
 
 **Note** Creation of new datasets is done with a ``POST`` onto a Dataverse collection. See the Dataverse Collections section above.
 
-**Note** In all commands below, dataset versions can be referred to as:
+.. _dataset-version-specifiers:
+
+Dataset Version Specifiers
+~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+In all commands below, dataset versions can be referred to as:
 
 * ``:draft``  the draft version, if any
 * ``:latest`` either a draft (if exists) or the latest published version.
@@ -2712,6 +2717,8 @@ The fully expanded example above (without environment variables) looks like this
 Files
 -----
 
+.. _get-json-rep-of-file:
+
 Get JSON Representation of a File
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
@@ -3499,6 +3506,55 @@ The fully expanded example above (without environment variables) looks like this
 
 You can download :download:`dct.xml <../../../../src/test/resources/xml/dct.xml>` from the example above to see what the XML looks like.
 
+Get File Citation as JSON
+~~~~~~~~~~~~~~~~~~~~~~~~~
+
+This API is for getting the file citation as it appears on the file landing page. It is formatted in HTML and encoded in JSON.
+
+To specify the version, you can use ``:latest-published`` or ``:draft`` or ``1.0`` or any other style listed under :ref:`dataset-version-specifiers`.
+
+When the dataset version is published, authentication is not required:
+
+.. code-block:: bash
+
+  export SERVER_URL=https://demo.dataverse.org
+  export FILE_ID=42
+  export DATASET_VERSION=:latest-published
+
+  curl "$SERVER_URL/api/files/$FILE_ID/versions/$DATASET_VERSION/citation"
+
+The fully expanded example above (without environment variables) looks like this:
+
+.. code-block:: bash
+
+  curl "https://demo.dataverse.org/api/files/42/versions/:latest-published/citation"
+
+When the dataset version is a draft or deaccessioned, authentication is required.
+
+By default, deaccessioned dataset versions are not included in the search when applying the :latest or :latest-published identifiers. Additionally, when filtering by a specific version tag, you will get a "unauthorized" error if the version is deaccessioned and you do not enable the ``includeDeaccessioned`` option described below.
+
+If you want to include deaccessioned dataset versions, you must set ``includeDeaccessioned`` query parameter to ``true``.
+
+.. code-block:: bash
+
+  export API_TOKEN=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
+  export SERVER_URL=https://demo.dataverse.org
+  export FILE_ID=42
+  export DATASET_VERSION=:draft
+  export INCLUDE_DEACCESSIONED=true
+
+  curl -H "X-Dataverse-key:$API_TOKEN" "$SERVER_URL/api/files/$FILE_ID/versions/$DATASET_VERSION/citation?includeDeaccessioned=$INCLUDE_DEACCESSIONED"
+
+The fully expanded example above (without environment variables) looks like this:
+
+.. code-block:: bash
+
+  curl -H "X-Dataverse-key:xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx" "https://demo.dataverse.org/api/files/42/versions/:draft/citation?includeDeaccessioned=true"
+
+If your file has a persistent identifier (PID, such as a DOI), you can pass it using the technique described under :ref:`get-json-rep-of-file`.
+
+This API is not for downloading various citation formats such as EndNote XML, RIS, or BibTeX. This functionality has been requested in https://github.com/IQSS/dataverse/issues/3140 and https://github.com/IQSS/dataverse/issues/9994.
+
 Provenance
 ~~~~~~~~~~
 

src/main/java/edu/harvard/iq/dataverse/api/AbstractApiBean.java

@@ -2,6 +2,7 @@
 
 import edu.harvard.iq.dataverse.*;
 import edu.harvard.iq.dataverse.actionlogging.ActionLogServiceBean;
+import static edu.harvard.iq.dataverse.api.Datasets.handleVersion;
 import edu.harvard.iq.dataverse.authorization.AuthenticationServiceBean;
 import edu.harvard.iq.dataverse.authorization.DataverseRole;
 import edu.harvard.iq.dataverse.authorization.RoleAssignee;
@@ -15,6 +16,10 @@
 import edu.harvard.iq.dataverse.engine.command.exception.CommandException;
 import edu.harvard.iq.dataverse.engine.command.exception.IllegalCommandException;
 import edu.harvard.iq.dataverse.engine.command.exception.PermissionException;
+import edu.harvard.iq.dataverse.engine.command.impl.GetDraftDatasetVersionCommand;
+import edu.harvard.iq.dataverse.engine.command.impl.GetLatestAccessibleDatasetVersionCommand;
+import edu.harvard.iq.dataverse.engine.command.impl.GetLatestPublishedDatasetVersionCommand;
+import edu.harvard.iq.dataverse.engine.command.impl.GetSpecificPublishedDatasetVersionCommand;
 import edu.harvard.iq.dataverse.externaltools.ExternalToolServiceBean;
 import edu.harvard.iq.dataverse.license.LicenseServiceBean;
 import edu.harvard.iq.dataverse.locality.StorageSiteServiceBean;
@@ -390,6 +395,32 @@ protected Dataset findDatasetOrDie(String id) throws WrappedResponse {
             }
         }
     }
+
+    protected DatasetVersion findDatasetVersionOrDie(final DataverseRequest req, String versionNumber, final Dataset ds, boolean includeDeaccessioned, boolean checkPermsWhenDeaccessioned) throws WrappedResponse {
+        DatasetVersion dsv = execCommand(handleVersion(versionNumber, new Datasets.DsVersionHandler<Command<DatasetVersion>>() {
+
+            @Override
+            public Command<DatasetVersion> handleLatest() {
+                return new GetLatestAccessibleDatasetVersionCommand(req, ds, includeDeaccessioned, checkPermsWhenDeaccessioned);
+            }
+
+            @Override
+            public Command<DatasetVersion> handleDraft() {
+                return new GetDraftDatasetVersionCommand(req, ds);
+            }
+
+            @Override
+            public Command<DatasetVersion> handleSpecific(long major, long minor) {
+                return new GetSpecificPublishedDatasetVersionCommand(req, ds, major, minor, includeDeaccessioned, checkPermsWhenDeaccessioned);
+            }
+
+            @Override
+            public Command<DatasetVersion> handleLatestPublished() {
+                return new GetLatestPublishedDatasetVersionCommand(req, ds, includeDeaccessioned, checkPermsWhenDeaccessioned);
+            }
+        }));
+        return dsv;
+    }
 
     protected DataFile findDataFileOrDie(String id) throws WrappedResponse {
 

src/main/java/edu/harvard/iq/dataverse/api/Datasets.java

@@ -2727,28 +2727,7 @@ private DatasetVersion getDatasetVersionOrDie(final DataverseRequest req, String
      * Will allow to define when the permissions should be checked when a deaccesioned dataset is requested. If the user doesn't have edit permissions will result in an error.
      */
     private DatasetVersion getDatasetVersionOrDie(final DataverseRequest req, String versionNumber, final Dataset ds, UriInfo uriInfo, HttpHeaders headers, boolean includeDeaccessioned, boolean checkPermsWhenDeaccessioned) throws WrappedResponse {
-        DatasetVersion dsv = execCommand(handleVersion(versionNumber, new DsVersionHandler<Command<DatasetVersion>>() {
-
-            @Override
-            public Command<DatasetVersion> handleLatest() {
-                return new GetLatestAccessibleDatasetVersionCommand(req, ds, includeDeaccessioned, checkPermsWhenDeaccessioned);
-            }
-
-            @Override
-            public Command<DatasetVersion> handleDraft() {
-                return new GetDraftDatasetVersionCommand(req, ds);
-            }
-
-            @Override
-            public Command<DatasetVersion> handleSpecific(long major, long minor) {
-                return new GetSpecificPublishedDatasetVersionCommand(req, ds, major, minor, includeDeaccessioned, checkPermsWhenDeaccessioned);
-            }
-
-            @Override
-            public Command<DatasetVersion> handleLatestPublished() {
-                return new GetLatestPublishedDatasetVersionCommand(req, ds, includeDeaccessioned, checkPermsWhenDeaccessioned);
-            }
-        }));
+        DatasetVersion dsv = findDatasetVersionOrDie(req, versionNumber, ds, includeDeaccessioned, checkPermsWhenDeaccessioned);
         if (dsv == null || dsv.getId() == null) {
             throw new WrappedResponse(notFound("Dataset version " + versionNumber + " of dataset " + ds.getId() + " not found"));
         }

src/main/java/edu/harvard/iq/dataverse/api/Files.java

@@ -2,6 +2,7 @@
 
 import com.google.gson.Gson;
 import com.google.gson.JsonObject;
+import edu.harvard.iq.dataverse.DataCitation;
 import edu.harvard.iq.dataverse.DataFile;
 import edu.harvard.iq.dataverse.DataFileServiceBean;
 import edu.harvard.iq.dataverse.DataFileTag;
@@ -27,6 +28,7 @@
 import edu.harvard.iq.dataverse.datasetutility.DataFileTagException;
 import edu.harvard.iq.dataverse.datasetutility.NoFilesException;
 import edu.harvard.iq.dataverse.datasetutility.OptionalFileParams;
+import edu.harvard.iq.dataverse.engine.command.Command;
 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.IllegalCommandException;
@@ -931,4 +933,37 @@ public Response getHasBeenDeleted(@Context ContainerRequestContext crc, @PathPar
             return ok(dataFileServiceBean.hasBeenDeleted(dataFile));
         }, getRequestUser(crc));
     }
+
+    /**
+     * @param fileIdOrPersistentId Database ID or PID of the data file.
+     * @param versionNumber The version of the dataset, such as 1.0, :draft,
+     * :latest-published, etc.
+     * @param includeDeaccessioned Defaults to false.
+     */
+    @GET
+    @AuthRequired
+    @Path("{id}/versions/{dsVersionString}/citation")
+    public Response getFileCitationByVersion(@Context ContainerRequestContext crc, @PathParam("id") String fileIdOrPersistentId, @PathParam("dsVersionString") String versionNumber, @QueryParam("includeDeaccessioned") boolean includeDeaccessioned) {
+        try {
+            DataverseRequest req = createDataverseRequest(getRequestUser(crc));
+            final DataFile df = execCommand(new GetDataFileCommand(req, findDataFileOrDie(fileIdOrPersistentId)));
+            Dataset ds = df.getOwner();
+            DatasetVersion dsv = findDatasetVersionOrDie(req, versionNumber, ds, includeDeaccessioned, true);
+            if (dsv == null) {
+                return unauthorized(BundleUtil.getStringFromBundle("files.api.no.draftOrUnauth"));
+            }
+
+            Long getDatasetVersionID = dsv.getId();
+            FileMetadata fm = dataFileServiceBean.findFileMetadataByDatasetVersionIdAndDataFileId(getDatasetVersionID, df.getId());
+            if (fm == null) {
+                return notFound(BundleUtil.getStringFromBundle("files.api.fileNotFound"));
+            }
+            boolean direct = df.isIdentifierRegistered();
+            DataCitation citation = new DataCitation(fm, direct);
+            return ok(citation.toString(true));
+        } catch (WrappedResponse ex) {
+            return ex.getResponse();
+        }
+    }
+
 }

src/main/java/propertyFiles/Bundle.properties

@@ -2633,7 +2633,9 @@ admin.api.deleteUser.success=Authenticated User {0} deleted.
 #Files.java
 files.api.metadata.update.duplicateFile=Filename already exists at {0}
 files.api.no.draft=No draft available for this file
+files.api.no.draftOrUnauth=Dataset version cannot be found or unauthorized.
 files.api.only.tabular.supported=This operation is only available for tabular files.
+files.api.fileNotFound=File could not be found.
 
 #Datasets.java
 datasets.api.updatePIDMetadata.failure.dataset.must.be.released=Modify Registration Metadata must be run on a published dataset.

src/test/java/edu/harvard/iq/dataverse/DataCitationTest.java

@@ -378,6 +378,36 @@ public void testTitleWithQuotes() throws ParseException {
 
     }
 
+    @Test
+    public void testFileCitationToStringHtml() throws ParseException {
+        DatasetVersion dsv = createATestDatasetVersion("Dataset Title", true);
+        FileMetadata fileMetadata = new FileMetadata();
+        fileMetadata.setLabel("foo.txt");
+        fileMetadata.setDataFile(new DataFile());
+        dsv.setVersionState(DatasetVersion.VersionState.RELEASED);
+        fileMetadata.setDatasetVersion(dsv);
+        dsv.setDataset(dsv.getDataset());
+        DataCitation fileCitation = new DataCitation(fileMetadata, false);
+        assertEquals("First Last, 1955, \"Dataset Title\", <a href=\"https://doi.org/10.5072/FK2/LK0D1H\" target=\"_blank\">https://doi.org/10.5072/FK2/LK0D1H</a>, LibraScholar, V1; foo.txt [fileName]", fileCitation.toString(true));
+    }
+
+    @Test
+    public void testFileCitationToStringHtmlFilePid() throws ParseException {
+        DatasetVersion dsv = createATestDatasetVersion("Dataset Title", true);
+        FileMetadata fileMetadata = new FileMetadata();
+        fileMetadata.setLabel("foo.txt");
+        DataFile dataFile = new DataFile();
+        dataFile.setProtocol("doi");
+        dataFile.setAuthority("10.42");
+        dataFile.setIdentifier("myFilePid");
+        fileMetadata.setDataFile(dataFile);
+        dsv.setVersionState(DatasetVersion.VersionState.RELEASED);
+        fileMetadata.setDatasetVersion(dsv);
+        dsv.setDataset(dsv.getDataset());
+        DataCitation fileCitation = new DataCitation(fileMetadata, true);
+        assertEquals("First Last, 1955, \"foo.txt\", <em>Dataset Title</em>, <a href=\"https://doi.org/10.42/myFilePid\" target=\"_blank\">https://doi.org/10.42/myFilePid</a>, LibraScholar, V1", fileCitation.toString(true));
+    }
+
     private DatasetVersion createATestDatasetVersion(String withTitle, boolean withAuthor) throws ParseException {
 
         Dataverse dataverse = new Dataverse();
@@ -400,6 +430,7 @@ private DatasetVersion createATestDatasetVersion(String withTitle, boolean withA
             fields.add(createTitleField(withTitle));
         }
         if (withAuthor) {
+            // TODO: "Last, First" would make more sense.
             fields.add(createAuthorField("First Last"));
         }