enable redirection of iDigBio recordset uuids also; related to #199

bio-guoda · Dec 22, 2023 · 1c2d3b3 · 1c2d3b3
1 parent a10a3ba
commit 1c2d3b3
Show file tree

Hide file tree

Showing 7 changed files with 121 additions and 159 deletions.
diff --git a/preston-serve/src/main/java/bio/guoda/preston/server/RedirectingServlet.java b/preston-serve/src/main/java/bio/guoda/preston/server/RedirectingServlet.java
@@ -1,6 +1,5 @@
 package bio.guoda.preston.server;
 
-import bio.guoda.preston.HashType;
 import bio.guoda.preston.MimeTypes;
 import bio.guoda.preston.RefNodeFactory;
 import bio.guoda.preston.ResourcesHTTP;
@@ -43,8 +42,8 @@ public class RedirectingServlet extends HttpServlet {
     public static final String DOI = "doi";
     public static final String CONTENT_ID = "contentId";
     public static final String ACTIVITY = "activity";
+
     public static final String QUERY_TYPE_UUID = "uuid";
-    public static final String QUERY_TYPE_HASH = "hash";
     public static final String QUERY_TYPE_URL = "url";
     public static final String QUERY_TYPE_DOI = "doi";
 
@@ -130,9 +129,6 @@ private String parseRequestedIdOrThrow(String requestURI) throws ServletExceptio
     static String queryTypeForRequestedId(String requestURI) {
         return Stream.of(requestURI)
                 .map(req -> URN_UUID_REQUEST_PATTERN.matcher(req).matches() ? QUERY_TYPE_UUID : req)
-                .map(req -> HashType.sha1.getIRIPattern().matcher(req).matches() ? QUERY_TYPE_HASH : req)
-                .map(req -> HashType.md5.getIRIPattern().matcher(req).matches() ? QUERY_TYPE_HASH : req)
-                .map(req -> HashType.sha256.getIRIPattern().matcher(req).matches() ? QUERY_TYPE_HASH : 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)
                 .findFirst()

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
@@ -10,39 +10,37 @@ SELECT DISTINCT
   ?seenAt
   ?provenanceId
   ?activity
-WHERE  
-{
-    BIND(?_doi_iri as ?doi) 
-    BIND(<hash://sha256/5b7fa37bf8b64e7c935c4ff3389e36f8dd162f0705410dd719fd089e1ea253cd> as ?provenanceId)
-
-{
+WHERE  {
+  BIND(?_doi_iri as ?doi)
+  BIND(<hash://sha256/5b7fa37bf8b64e7c935c4ff3389e36f8dd162f0705410dd719fd089e1ea253cd> as ?provenanceId)
+  {
     SELECT ?contentId ?seenAt ?archiveUrl ?uuid ?activity
-WHERE {
-  graph ?g1 {
-   ?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" .
+    WHERE {
+      graph ?g1 {
+        ?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" .
+      }
+      graph ?activity {
+        ?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 .
+      }
+    }
   }
-  graph ?activity {
-    ?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 .  
+  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 .
     }
+  }
 
-}
-} 
-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/hash.rq b/preston-serve/src/main/resources/bio/guoda/preston/server/hash.rq
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
@@ -20,7 +20,7 @@ WHERE
 WHERE {
   graph ?g1 {
    ?uuid <http://www.w3.org/ns/prov#hadMember> ?_url_iri .
-   ?uuid <http://www.w3.org/1999/02/22-rdf-syntax-ns#seeAlso> ?doi .
+   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" .
   }
   graph ?activity {
@@ -36,11 +36,11 @@ UNION
     SELECT ?contentId ?seenAt ?uuid ?activity
 WHERE {
    ?uuid <http://www.w3.org/ns/prov#hadMember> ?_url_iri . 
-   ?uuid <http://www.w3.org/1999/02/22-rdf-syntax-ns#seeAlso> ?doi .
+   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 . 
+   ?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

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
@@ -10,38 +10,36 @@ SELECT DISTINCT
   ?seenAt
   ?provenanceId
   ?activity
-WHERE  
-{
+WHERE  {
     BIND(?_uuid_iri as ?uuid) 
     BIND(<hash://sha256/5b7fa37bf8b64e7c935c4ff3389e36f8dd162f0705410dd719fd089e1ea253cd> as ?provenanceId) 
-{
+  {
     SELECT ?contentId ?seenAt ?archiveUrl ?doi ?activity
-WHERE {
-  graph ?g1 {
-   ?_uuid_iri <http://www.w3.org/ns/prov#hadMember> ?archiveUrl .
-   ?_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" .
-  }
-  graph ?activity {
-    ?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 .  
+    WHERE {
+      graph ?g1 {
+       ?_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" .
+      }
+      graph ?activity {
+        ?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 .
+        }
     }
-
-}
-} 
-UNION
-{
+  }
+  UNION
+  {
     SELECT ?contentId ?seenAt ?archiveUrl ?doi ?activity
-WHERE {
-   ?_uuid_iri <http://www.w3.org/ns/prov#hadMember> ?archiveUrl . 
-   ?_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 . 
-}
-}
+    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
@@ -0,0 +1,62 @@
+package bio.guoda.preston.server;
+
+import bio.guoda.preston.RefNodeFactory;
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import org.hamcrest.core.Is;
+import org.junit.Test;
+
+import java.io.IOException;
+import java.net.URISyntaxException;
+import java.util.Map;
+
+import static bio.guoda.preston.server.RedirectingServlet.ACTIVITY;
+import static bio.guoda.preston.server.RedirectingServlet.CONTENT_ID;
+import static org.hamcrest.MatcherAssert.assertThat;
+
+public class RedirectingServletIT {
+
+
+    private String sparqlEndpoint = "http://localhost:7878/query";
+
+    @Test
+    public void dealiasDOI() throws IOException, URISyntaxException {
+        Map<String, String> contentId = RedirectingServlet.findMostRecentContentId(
+                RefNodeFactory.toIRI("https://doi.org/10.15468/aomfnb"),
+                RedirectingServlet.QUERY_TYPE_DOI,
+                sparqlEndpoint
+        );
+        assertThat(contentId.get(CONTENT_ID), Is.is("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"));
+        assertThat(contentId.get(ACTIVITY), Is.is("urn:uuid:77f3faf7-acd2-4f14-9c0e-4e04ef5b63c7"));
+    }
+
+    @Test
+    public void dealiasUUIDGBIF() throws IOException, URISyntaxException {
+        Map<String, String> contentId = RedirectingServlet.findMostRecentContentId(
+                RefNodeFactory.toIRI("urn:uuid:4fa7b334-ce0d-4e88-aaae-2e0c138d049e"),
+                RedirectingServlet.QUERY_TYPE_UUID,
+                sparqlEndpoint);
+        assertThat(contentId.get(CONTENT_ID), Is.is("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"));
+
+    }
+
+    @Test
+    public void dealiasUUIDiDigBio() throws IOException, URISyntaxException {
+        Map<String, String> contentId = RedirectingServlet.findMostRecentContentId(
+                RefNodeFactory.toIRI("urn:uuid:65007e62-740c-4302-ba20-260fe68da291"),
+                RedirectingServlet.QUERY_TYPE_UUID,
+                sparqlEndpoint);
+        assertThat(contentId.get(CONTENT_ID), Is.is("hash://sha256/f5d8f67c1eca34cbba1abac12f353585c78bb053bc8ce7ee7e7a78846e1bfc4a"));
+
+    }
+
+    @Test
+    public void dealiasURL() throws URISyntaxException, IOException {
+        Map<String, String> contentId = RedirectingServlet.findMostRecentContentId(
+                RefNodeFactory.toIRI("https://hosted-datasets.gbif.org/eBird/2022-eBird-dwca-1.0.zip"),
+                RedirectingServlet.QUERY_TYPE_URL,
+                sparqlEndpoint);
+        assertThat(contentId.get(CONTENT_ID), Is.is("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"));
+    }
+
+}
diff --git a/preston-serve/src/test/java/bio/guoda/preston/server/RedirectingServletTest.java b/preston-serve/src/test/java/bio/guoda/preston/server/RedirectingServletTest.java
@@ -1,61 +1,18 @@
 package bio.guoda.preston.server;
 
-import bio.guoda.preston.RefNodeFactory;
 import com.fasterxml.jackson.databind.JsonNode;
 import com.fasterxml.jackson.databind.ObjectMapper;
 import org.hamcrest.core.Is;
 import org.junit.Test;
 
 import java.io.IOException;
-import java.net.URISyntaxException;
 import java.util.Map;
 
-import static bio.guoda.preston.server.RedirectingServlet.ACTIVITY;
 import static bio.guoda.preston.server.RedirectingServlet.CONTENT_ID;
 import static org.hamcrest.MatcherAssert.assertThat;
 
 public class RedirectingServletTest {
 
-
-    @Test
-    public void dealiasDOI() throws IOException, URISyntaxException {
-        Map<String, String> contentId = RedirectingServlet.findMostRecentContentId(
-                RefNodeFactory.toIRI("https://doi.org/10.15468/aomfnb"),
-                RedirectingServlet.QUERY_TYPE_DOI, "https://lod.globalbioticinteractions.org/query"
-        );
-        assertThat(contentId.get(CONTENT_ID), Is.is("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"));
-        assertThat(contentId.get(ACTIVITY), Is.is("urn:uuid:77f3faf7-acd2-4f14-9c0e-4e04ef5b63c7"));
-    }
-
-    @Test
-    public void dealiasHash() throws IOException, URISyntaxException {
-        Map<String, String> contentId = RedirectingServlet.findMostRecentContentId(
-                RefNodeFactory.toIRI("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"),
-                RedirectingServlet.QUERY_TYPE_HASH,
-                "https://lod.globalbioticinteractions.org/query");
-        assertThat(contentId.get(CONTENT_ID), Is.is("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"));
-
-    }
-
-    @Test
-    public void dealiasUUID() throws IOException, URISyntaxException {
-        Map<String, String> contentId = RedirectingServlet.findMostRecentContentId(
-                RefNodeFactory.toIRI("urn:uuid:4fa7b334-ce0d-4e88-aaae-2e0c138d049e"),
-                RedirectingServlet.QUERY_TYPE_UUID,
-                "https://lod.globalbioticinteractions.org/query");
-        assertThat(contentId.get(CONTENT_ID), Is.is("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"));
-
-    }
-
-    @Test
-    public void dealiasURL() throws URISyntaxException, IOException {
-        Map<String, String> contentId = RedirectingServlet.findMostRecentContentId(
-                RefNodeFactory.toIRI("https://hosted-datasets.gbif.org/eBird/2022-eBird-dwca-1.0.zip"),
-                RedirectingServlet.QUERY_TYPE_URL,
-                "https://lod.globalbioticinteractions.org/query");
-        assertThat(contentId.get(CONTENT_ID), Is.is("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d"));
-    }
-
     @Test
     public void parseResult() throws IOException {
         JsonNode response = new ObjectMapper().readTree(getClass().getResourceAsStream("url-response.json"));
@@ -84,7 +41,7 @@ public void queryTypeUUID() {
     @Test
     public void queryTypeHash() {
         String s = RedirectingServlet.queryTypeForRequestedId("hash://sha256/1e2b7436fce1848f41698e5a9c193f311abaf0ee051bec1a2e48b5106d29524d");
-        assertThat(s, Is.is(RedirectingServlet.QUERY_TYPE_HASH));
+        assertThat(s, Is.is("unknown"));
     }
 
     @Test