re-introduce provenance query by content id; related to #199

bio-guoda · Dec 27, 2023 · 0b95792 · 0b95792
1 parent 1e530af
commit 0b95792
Show file tree

Hide file tree

Showing 6 changed files with 71 additions and 37 deletions.
diff --git a/preston-serve/src/main/java/bio/guoda/preston/server/ProvUtil.java b/preston-serve/src/main/java/bio/guoda/preston/server/ProvUtil.java
@@ -16,6 +16,7 @@
 import java.net.URISyntaxException;
 import java.nio.charset.StandardCharsets;
 import java.util.Arrays;
+import java.util.List;
 import java.util.Map;
 import java.util.TreeMap;
 import java.util.regex.Pattern;
@@ -25,8 +26,10 @@ public class ProvUtil {
     public static final Pattern URN_UUID_REQUEST_PATTERN
             = Pattern.compile("^urn:uuid:" + UUIDUtil.UUID_PATTERN_PART + "$");
     public static final String QUERY_TYPE_UUID = "uuid";
+    public static final String QUERY_TYPE_CONTENT_ID = "contentId";
     public static final String QUERY_TYPE_URL = "url";
     public static final String QUERY_TYPE_DOI = "doi";
+    public static final List<String> QUERIES_SUPPORTED = Arrays.asList(QUERY_TYPE_DOI, QUERY_TYPE_UUID, QUERY_TYPE_URL, QUERY_TYPE_CONTENT_ID);
 
     public static Map<String, String> findMostRecentContentId(IRI iri, String paramName, String sparqlEndpoint) throws IOException, URISyntaxException {
         String response = findProvenance(iri, paramName, sparqlEndpoint);
@@ -75,7 +78,8 @@ static String queryTypeForRequestedId(String requestURI) {
                 .map(req -> URN_UUID_REQUEST_PATTERN.matcher(req).matches() ? QUERY_TYPE_UUID : req)
                 .map(req -> Pattern.compile("^(10[.])([^/]+)/(.*)$").matcher(req).matches() ? QUERY_TYPE_DOI : req)
                 .map(req -> Pattern.compile("^http[s]{0,1}://[^ ]+").matcher(req).matches() ? QUERY_TYPE_URL : req)
-                .filter(type -> Arrays.asList(QUERY_TYPE_DOI, QUERY_TYPE_UUID, QUERY_TYPE_URL).contains(type))
+                .map(req -> Pattern.compile("^hash://[a-zA-Z0-9]+/[a-f0-9]+$").matcher(req).matches() ? QUERY_TYPE_CONTENT_ID : req)
+                .filter(QUERIES_SUPPORTED::contains)
                 .findFirst()
                 .orElse("unknown");
     }

diff --git a/preston-serve/src/main/resources/bio/guoda/preston/server/contentId.rq b/preston-serve/src/main/resources/bio/guoda/preston/server/contentId.rq
@@ -0,0 +1,45 @@
+#+ endpoint: https://lod.globalbioticinteractions.org/query
+#+ defaults:
+#+   - uuid: urn:uuid:4fa7b334-ce0d-4e88-aaae-2e0c138d049e
+
+SELECT DISTINCT
+  ?uuid
+  ?doi
+  ?contentId
+  ?archiveUrl
+  ?seenAt
+  ?provenanceId
+  ?activity
+  WHERE {
+            {
+                SELECT DISTINCT
+                  ?contentId
+                  ?archiveUrl
+                  ?seenAt
+                  ?provenanceId
+                  ?activity
+                WHERE  {
+                    BIND(?_contentId_iri as ?contentId)
+                    BIND(<hash://sha256/5b7fa37bf8b64e7c935c4ff3389e36f8dd162f0705410dd719fd089e1ea253cd> as ?provenanceId)
+                  {
+                    SELECT ?uuid ?seenAt ?archiveUrl ?doi ?activity
+                    WHERE {
+                      graph ?activity {
+                        ?activity <http://www.w3.org/ns/prov#used> ?archiveUrl .
+                        ?activity <http://www.w3.org/ns/prov#generatedAtTime> ?seenAt .
+                        ?_contentId_iri <http://www.w3.org/ns/prov#qualifiedGeneration> ?activity .
+                      }
+                    }
+                  }
+                } ORDER BY DESC(?seenAt) LIMIT 2
+            }
+            {
+                graph ?g1 {
+                  ?uuid <http://www.w3.org/ns/prov#hadMember> ?archiveUrl .
+                  ?uuid <http://www.w3.org/1999/02/22-rdf-syntax-ns#seeAlso> ?doi .
+                  ?archiveUrl <http://purl.org/dc/elements/1.1/format> "application/dwca" .
+                }
+            }
+
+
+  }
diff --git a/preston-serve/src/main/resources/bio/guoda/preston/server/doi.rq b/preston-serve/src/main/resources/bio/guoda/preston/server/doi.rq
@@ -28,18 +28,6 @@ WHERE  {
       }
     }
   }
-  UNION
-  {
-    SELECT ?contentId ?seenAt ?archiveUrl ?uuid ?activity
-    WHERE {
-       ?uuid <http://www.w3.org/ns/prov#hadMember> ?archiveUrl .
-       ?uuid <http://www.w3.org/1999/02/22-rdf-syntax-ns#seeAlso> ?_doi_iri .
-       ?archiveUrl <http://purl.org/dc/elements/1.1/format> "application/dwca" .
-       ?activity <http://www.w3.org/ns/prov#used> ?archiveUrl .
-       ?activity <http://www.w3.org/ns/prov#generatedAtTime> ?seenAt .
-       ?contentId <http://www.w3.org/ns/prov#qualifiedGeneration> ?activity .
-    }
-  }
 
   # exclude blank nodes that represent unavailable or non-content
   FILTER(!CONTAINS(str(?contentId), ".well-known/genid" ) )

diff --git a/preston-serve/src/main/resources/bio/guoda/preston/server/url.rq b/preston-serve/src/main/resources/bio/guoda/preston/server/url.rq
@@ -30,18 +30,6 @@ WHERE {
   }
 
 }
-} 
-UNION
-{
-    SELECT ?contentId ?seenAt ?uuid ?activity
-WHERE {
-   ?uuid <http://www.w3.org/ns/prov#hadMember> ?_url_iri . 
-   OPTIONAL { ?uuid <http://www.w3.org/1999/02/22-rdf-syntax-ns#seeAlso> ?doi . }
-   ?_url_iri <http://purl.org/dc/elements/1.1/format> "application/dwca" .
-   ?activity <http://www.w3.org/ns/prov#used> ?_url_iri .
-   ?activity <http://www.w3.org/ns/prov#generatedAtTime> ?seenAt .
-   ?contentId <http://www.w3.org/ns/prov#qualifiedGeneration> ?activity .
-}
 }
   # exclude blank nodes that represent unavailable or non-content
   FILTER(!CONTAINS(str(?contentId), ".well-known/genid" ) )

diff --git a/preston-serve/src/main/resources/bio/guoda/preston/server/uuid.rq b/preston-serve/src/main/resources/bio/guoda/preston/server/uuid.rq
@@ -28,18 +28,6 @@ WHERE  {
         }
     }
   }
-  UNION
-  {
-    SELECT ?contentId ?seenAt ?archiveUrl ?doi ?activity
-    WHERE {
-       ?_uuid_iri <http://www.w3.org/ns/prov#hadMember> ?archiveUrl .
-       OPTIONAL { ?_uuid_iri <http://www.w3.org/1999/02/22-rdf-syntax-ns#seeAlso> ?doi . }
-       ?archiveUrl <http://purl.org/dc/elements/1.1/format> "application/dwca" .
-       ?activity <http://www.w3.org/ns/prov#used> ?archiveUrl .
-       ?activity <http://www.w3.org/ns/prov#generatedAtTime> ?seenAt .
-       ?contentId <http://www.w3.org/ns/prov#qualifiedGeneration> ?activity .
-    }
-  }
   # exclude blank nodes that represent unavailable or non-content
   FILTER(!CONTAINS(str(?contentId), ".well-known/genid" ) )
 

diff --git a/preston-serve/src/test/java/bio/guoda/preston/server/RedirectingServletIT.java b/preston-serve/src/test/java/bio/guoda/preston/server/RedirectingServletIT.java
@@ -9,7 +9,12 @@
 import java.util.Map;
 
 import static bio.guoda.preston.server.RedirectingServlet.ACTIVITY;
+import static bio.guoda.preston.server.RedirectingServlet.ARCHIVE_URL;
 import static bio.guoda.preston.server.RedirectingServlet.CONTENT_ID;
+import static bio.guoda.preston.server.RedirectingServlet.DOI;
+import static bio.guoda.preston.server.RedirectingServlet.PROVENANCE_ID;
+import static bio.guoda.preston.server.RedirectingServlet.SEEN_AT;
+import static bio.guoda.preston.server.RedirectingServlet.UUID;
 import static org.hamcrest.MatcherAssert.assertThat;
 
 public class RedirectingServletIT {
@@ -38,6 +43,22 @@ public void dealiasUUIDGBIF() throws IOException, URISyntaxException {
 
     }
 
+    @Test
+    public void dealiasContentIdGBIF() throws IOException, URISyntaxException {
+        Map<String, String> contentId = ProvUtil.findMostRecentContentId(
+                RefNodeFactory.toIRI("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"),
+                ProvUtil.QUERY_TYPE_CONTENT_ID,
+                sparqlEndpoint);
+        assertThat(contentId.get(CONTENT_ID), Is.is("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"));
+        assertThat(contentId.get(ACTIVITY), Is.is("urn:uuid:77f3faf7-acd2-4f14-9c0e-4e04ef5b63c7"));
+        assertThat(contentId.get(PROVENANCE_ID), Is.is("hash://sha256/5b7fa37bf8b64e7c935c4ff3389e36f8dd162f0705410dd719fd089e1ea253cd"));
+        assertThat(contentId.get(SEEN_AT), Is.is("2023-12-02T16:05:25.261Z"));
+        assertThat(contentId.get(ARCHIVE_URL), Is.is("https://hosted-datasets.gbif.org/eBird/2022-eBird-dwca-1.0.zip"));
+        assertThat(contentId.get(UUID), Is.is("urn:uuid:4fa7b334-ce0d-4e88-aaae-2e0c138d049e"));
+        assertThat(contentId.get(DOI), Is.is("https://doi.org/10.15468/aomfnb"));
+
+    }
+
     @Test
     public void dealiasUUIDiDigBio() throws IOException, URISyntaxException {
         Map<String, String> contentId = ProvUtil.findMostRecentContentId(