995x - Signposting fixes

doc/release-notes/9955-Signposting-updates.md

@@ -0,0 +1,7 @@
+This release fixes several issues (#9952, #9953, #9957) where the Signposting output did not match the Signposting specification. These changes introduce backward-incompatibility, but since Signposting support was added recently (in Dataverse 5.14 in PR #8981), we feel it's best to do this clean up and not support the old implementation that was not fully compliant with the spec.
+
+To fix #9952, we surround the license info with `<` and `>`.
+
+To fix #9953, we no longer wrap the response in a `{"status":"OK","data":{` JSON object. This has also been noted in the guides at https://dataverse-guide--9955.org.readthedocs.build/en/9955/api/native-api.html#retrieve-signposting-information
+
+To fix #9957, we corrected the mime/content type, changing it from `json+ld` to `ld+json`. For backward compatibility, we are still supporting the old one, for now.

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

@@ -2394,11 +2394,11 @@ Signposting involves the addition of a `Link <https://tools.ietf.org/html/rfc598
 
 Here is an example of a "Link" header:
 
-``Link: <https://doi.org/10.5072/FK2/YD5QDG>;rel="cite-as", <https://doi.org/10.5072/FK2/YD5QDG>;rel="describedby";type="application/vnd.citationstyles.csl+json",<https://demo.dataverse.org/api/datasets/export?exporter=schema.org&persistentId=doi:10.5072/FK2/YD5QDG>;rel="describedby";type="application/json+ld", <https://schema.org/AboutPage>;rel="type",<https://schema.org/Dataset>;rel="type", https://demo.dataverse.org/api/datasets/:persistentId/versions/1.0/customlicense?persistentId=doi:10.5072/FK2/YD5QDG;rel="license", <https://demo.dataverse.org/api/datasets/:persistentId/versions/1.0/linkset?persistentId=doi:10.5072/FK2/YD5QDG> ; rel="linkset";type="application/linkset+json"``
+``Link: <https://doi.org/10.5072/FK2/YD5QDG>;rel="cite-as", <https://doi.org/10.5072/FK2/YD5QDG>;rel="describedby";type="application/vnd.citationstyles.csl+json",<https://demo.dataverse.org/api/datasets/export?exporter=schema.org&persistentId=doi:10.5072/FK2/YD5QDG>;rel="describedby";type="application/ld+json", <https://schema.org/AboutPage>;rel="type",<https://schema.org/Dataset>;rel="type", <https://demo.dataverse.org/api/datasets/:persistentId/versions/1.0/customlicense?persistentId=doi:10.5072/FK2/YD5QDG>;rel="license", <https://demo.dataverse.org/api/datasets/:persistentId/versions/1.0/linkset?persistentId=doi:10.5072/FK2/YD5QDG> ; rel="linkset";type="application/linkset+json"``
 
 The URL for linkset information is discoverable under the ``rel="linkset";type="application/linkset+json`` entry in the "Link" header, such as in the example above.
 
-The reponse includes a JSON object conforming to the `Signposting <https://signposting.org>`__ specification.
+The reponse includes a JSON object conforming to the `Signposting <https://signposting.org>`__ specification. As part of this conformance, unlike most Dataverse API responses, the output is not wrapped in a ``{"status":"OK","data":{`` object.
 Signposting is not supported for draft dataset versions.
 
 .. code-block:: bash

src/main/java/edu/harvard/iq/dataverse/DatasetFieldServiceBean.java

@@ -500,7 +500,8 @@ public void process(HttpResponse response, HttpContext context) throws HttpExcep
                         .setRetryHandler(new DefaultHttpRequestRetryHandler(3, false))
                         .build()) {
                     HttpGet httpGet = new HttpGet(retrievalUri);
-                    httpGet.addHeader("Accept", "application/json+ld, application/json");
+                    //application/json+ld is for backward compatibility
+                    httpGet.addHeader("Accept", "application/ld+json, application/json+ld, application/json");
 
                     HttpResponse response = httpClient.execute(httpGet);
                     String data = EntityUtils.toString(response.getEntity(), StandardCharsets.UTF_8);

src/main/java/edu/harvard/iq/dataverse/DatasetPage.java

@@ -2880,6 +2880,12 @@ public void sort() {
     public String refresh() {
         logger.fine("refreshing");
 
+        //In v5.14, versionId was null here. In 6.0, it appears not to be.
+        //This check is to handle the null if it reappears/occurs under other circumstances
+        if(versionId==null) {
+            logger.warning("versionId was null in refresh");
+            versionId = workingVersion.getId();
+        }
         //dataset = datasetService.find(dataset.getId());
         dataset = null;
         workingVersion = null; 
@@ -2889,10 +2895,9 @@ public String refresh() {
         DatasetVersionServiceBean.RetrieveDatasetVersionResponse retrieveDatasetVersionResponse = null;
 
         if (versionId != null) {
-            // versionId must have been set by now, in the init() method, 
-            // regardless of how the page was originally called - by the dataset
-            // database id, by the persistent identifier, or by the db id of
-            // the version. 
+            // versionId must have been set by now (see null check above), in the init()
+            // method, regardless of how the page was originally called - by the dataset
+            // database id, by the persistent identifier, or by the db id of the version.
             this.workingVersion = datasetVersionService.findDeep(versionId);
             dataset = workingVersion.getDataset();
         } 
@@ -6282,7 +6287,10 @@ public String getWebloaderUrlForDataset(Dataset d) {
     String signpostingLinkHeader = null;
 
     public String getSignpostingLinkHeader() {
-        if (!workingVersion.isReleased()) {
+        if ((workingVersion==null) || (!workingVersion.isReleased())) {
+            if(workingVersion==null) {
+                logger.warning("workingVersion was null in getSignpostingLinkHeader");
+            }
             return null;
         }
         if (signpostingLinkHeader == null) {

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

@@ -277,7 +277,7 @@ public Response getDataset(@Context ContainerRequestContext crc, @PathParam("id"
 
     @GET
     @Path("/export")
-    @Produces({"application/xml", "application/json", "application/html" })
+    @Produces({"application/xml", "application/json", "application/html", "application/ld+json" })
     public Response exportDataset(@QueryParam("persistentId") String persistentId, @QueryParam("exporter") String exporter, @Context UriInfo uriInfo, @Context HttpHeaders headers, @Context HttpServletResponse response) {
 
         try {
@@ -650,24 +650,26 @@ public Response getVersionMetadataBlock(@Context ContainerRequestContext crc,
     @GET
     @AuthRequired
     @Path("{id}/versions/{versionId}/linkset")
-    public Response getLinkset(@Context ContainerRequestContext crc, @PathParam("id") String datasetId, @PathParam("versionId") String versionId, @Context UriInfo uriInfo, @Context HttpHeaders headers) {
+    public Response getLinkset(@Context ContainerRequestContext crc, @PathParam("id") String datasetId, @PathParam("versionId") String versionId, 
+           @Context UriInfo uriInfo, @Context HttpHeaders headers) {
         if (DS_VERSION_DRAFT.equals(versionId)) {
             return badRequest("Signposting is not supported on the " + DS_VERSION_DRAFT + " version");
         }
-        User user = getRequestUser(crc);
-        return response(req -> {
+        DataverseRequest req = createDataverseRequest(getRequestUser(crc));
+        try {
             DatasetVersion dsv = getDatasetVersionOrDie(req, versionId, findDatasetOrDie(datasetId), uriInfo, headers);
-            return ok(Json.createObjectBuilder().add(
-                    "linkset",
-                    new SignpostingResources(
-                            systemConfig,
-                            dsv,
-                            JvmSettings.SIGNPOSTING_LEVEL1_AUTHOR_LIMIT.lookupOptional().orElse(""),
-                            JvmSettings.SIGNPOSTING_LEVEL1_ITEM_LIMIT.lookupOptional().orElse("")
-                    ).getJsonLinkset()
-            )
-            );
-        }, user);
+            return Response
+                    .ok(Json.createObjectBuilder()
+                            .add("linkset",
+                                    new SignpostingResources(systemConfig, dsv,
+                                            JvmSettings.SIGNPOSTING_LEVEL1_AUTHOR_LIMIT.lookupOptional().orElse(""),
+                                            JvmSettings.SIGNPOSTING_LEVEL1_ITEM_LIMIT.lookupOptional().orElse(""))
+                                                    .getJsonLinkset())
+                            .build())
+                    .type(MediaType.APPLICATION_JSON).build();
+        } catch (WrappedResponse wr) {
+            return wr.getResponse();
+        }
     }
 
     @GET

src/main/java/edu/harvard/iq/dataverse/util/SignpostingResources.java

@@ -71,14 +71,14 @@ public String getLinks() {
 
         String describedby = "<" + ds.getGlobalId().asURL().toString() + ">;rel=\"describedby\"" + ";type=\"" + "application/vnd.citationstyles.csl+json\"";
         describedby += ",<" + systemConfig.getDataverseSiteUrl() + "/api/datasets/export?exporter=schema.org&persistentId="
-                + ds.getProtocol() + ":" + ds.getAuthority() + "/" + ds.getIdentifier() + ">;rel=\"describedby\"" + ";type=\"application/json+ld\"";
+                + ds.getProtocol() + ":" + ds.getAuthority() + "/" + ds.getIdentifier() + ">;rel=\"describedby\"" + ";type=\"application/ld+json\"";
         valueList.add(describedby);
 
         String type = "<https://schema.org/AboutPage>;rel=\"type\"";
         type = "<https://schema.org/AboutPage>;rel=\"type\",<" + defaultFileTypeValue + ">;rel=\"type\"";
         valueList.add(type);
 
-        String licenseString = DatasetUtil.getLicenseURI(workingDatasetVersion) + ";rel=\"license\"";
+        String licenseString = "<" + DatasetUtil.getLicenseURI(workingDatasetVersion) + ">;rel=\"license\"";
         valueList.add(licenseString);
 
         String linkset = "<" + systemConfig.getDataverseSiteUrl() + "/api/datasets/:persistentId/versions/"
@@ -116,7 +116,7 @@ public JsonArrayBuilder getJsonLinkset() {
                         systemConfig.getDataverseSiteUrl() + "/api/datasets/export?exporter=schema.org&persistentId=" + ds.getProtocol() + ":" + ds.getAuthority() + "/" + ds.getIdentifier()
                 ).add(
                         "type",
-                        "application/json+ld"
+                        "application/ld+json"
                 )
         );
         JsonArrayBuilder linksetJsonObj = Json.createArrayBuilder();

src/test/java/edu/harvard/iq/dataverse/api/SignpostingIT.java

@@ -80,6 +80,7 @@ public void testSignposting() {
         assertTrue(linkHeader.contains(datasetPid));
         assertTrue(linkHeader.contains("cite-as"));
         assertTrue(linkHeader.contains("describedby"));
+        assertTrue(linkHeader.contains("<http://creativecommons.org/publicdomain/zero/1.0>;rel=\"license\""));
 
         Pattern pattern = Pattern.compile("<([^<]*)> ; rel=\"linkset\";type=\"application\\/linkset\\+json\"");
         Matcher matcher = pattern.matcher(linkHeader);
@@ -92,7 +93,7 @@ public void testSignposting() {
 
         String responseString = linksetResponse.getBody().asString();
 
-        JsonObject data = JsonUtil.getJsonObject(responseString).getJsonObject("data");
+        JsonObject data = JsonUtil.getJsonObject(responseString);
         JsonObject lso = data.getJsonArray("linkset").getJsonObject(0);
         System.out.println("Linkset: " + lso.toString());
 
@@ -101,6 +102,16 @@ public void testSignposting() {
         assertTrue(lso.getString("anchor").indexOf("/dataset.xhtml?persistentId=" + datasetPid) > 0);
         assertTrue(lso.containsKey("describedby"));
 
+        // Test export URL from link header
+        // regex inspired by https://stackoverflow.com/questions/68860255/how-to-match-the-closest-opening-and-closing-brackets
+        Pattern exporterPattern = Pattern.compile("[<\\[][^()\\[\\]]*?exporter=schema.org[^()\\[\\]]*[>\\]]");
+        Matcher exporterMatcher = exporterPattern.matcher(linkHeader);
+        exporterMatcher.find();
+
+        Response exportDataset = UtilIT.exportDataset(datasetPid, "schema.org");
+        exportDataset.prettyPrint();
+        exportDataset.then().assertThat().statusCode(OK.getStatusCode());
+
     }
 
 }