From cf6358736874010fd421bfc6ffec6b100378ccb6 Mon Sep 17 00:00:00 2001
From: Ian Nesbitt <ian.nesbitt@gmail.com>
Date: Sat, 17 Aug 2024 07:57:02 -0700
Subject: [PATCH] Adding queries to address #100

---
 .../application-context-schema-org.xml        | 104 +++++++++++++++
 .../geocoordinates/geocoordinates.jsonld      | 123 ++++++++++++++++++
 .../json-ld/geocoordinates/systemmetadata.xml |  26 ++++
 3 files changed, 253 insertions(+)
 create mode 100644 src/test/resources/org/dataone/cn/index/resources/d1_testdocs/json-ld/geocoordinates/geocoordinates.jsonld
 create mode 100644 src/test/resources/org/dataone/cn/index/resources/d1_testdocs/json-ld/geocoordinates/systemmetadata.xml

diff --git a/src/main/resources/application-context-schema-org.xml b/src/main/resources/application-context-schema-org.xml
index 448f306e..02609ff9 100644
--- a/src/main/resources/application-context-schema-org.xml
+++ b/src/main/resources/application-context-schema-org.xml
@@ -576,6 +576,110 @@
         <property name="converter" ref="solrLongitudeConverter" />
     </bean>
 
+<!-- Extract south bounding coordinate from a 'SO:spatialCoverage' property of type 'GeoCoordinates'. -->
+<bean id="schema_org_geoCoordinates_south" class="org.dataone.cn.indexer.annotation.SparqlField">
+    <constructor-arg name="name" value="southBoundCoord" />
+    <constructor-arg name="query">
+        <value>
+            <![CDATA[
+                PREFIX rdf:  <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+                PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+                PREFIX SO:   <http://schema.org/>
+
+                SELECT ?southBoundCoord
+                WHERE {
+                    ?datasetId rdf:type           SO:Dataset .
+                    ?datasetId SO:spatialCoverage ?spatial .
+                    ?spatial   rdf:type           SO:Place .
+                    ?spatial   SO:geo             ?geo .
+                    ?geo       rdf:type           SO:GeoCoordinates .
+                    ?geo       SO:latitude        ?southBoundCoord .
+                }
+                LIMIT 1
+            ]]>
+        </value>
+    </constructor-arg>
+    <property name="converter" ref="solrLatitudeConverter" />
+</bean>
+
+<!-- Extract west bounding coordinate from a 'SO:spatialCoverage' property of type 'GeoCoordinates'. -->
+<bean id="schema_org_geoCoordinates_west" class="org.dataone.cn.indexer.annotation.SparqlField">
+    <constructor-arg name="name" value="westBoundCoord" />
+    <constructor-arg name="query">
+        <value>
+            <![CDATA[
+                PREFIX rdf:  <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+                PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+                PREFIX SO:   <http://schema.org/>
+
+                SELECT ?westBoundCoord
+                WHERE {
+                    ?datasetId rdf:type           SO:Dataset .
+                    ?datasetId SO:spatialCoverage ?spatial .
+                    ?spatial   rdf:type           SO:Place .
+                    ?spatial   SO:geo             ?geo .
+                    ?geo       rdf:type           SO:GeoCoordinates .
+                    ?geo       SO:longitude       ?westBoundCoord .
+                }
+                LIMIT 1
+            ]]>
+        </value>
+    </constructor-arg>
+    <property name="converter" ref="solrLongitudeConverter" />
+</bean>
+
+<!-- Extract north bounding coordinate from a 'SO:spatialCoverage' property of type 'GeoCoordinates'. -->
+<bean id="schema_org_geoCoordinates_north" class="org.dataone.cn.indexer.annotation.SparqlField">
+    <constructor-arg name="name" value="northBoundCoord" />
+    <constructor-arg name="query">
+        <value>
+            <![CDATA[
+                PREFIX rdf:  <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+                PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+                PREFIX SO:   <http://schema.org/>
+
+                SELECT ?northBoundCoord
+                WHERE {
+                    ?datasetId rdf:type           SO:Dataset .
+                    ?datasetId SO:spatialCoverage ?spatial .
+                    ?spatial   rdf:type           SO:Place .
+                    ?spatial   SO:geo             ?geo .
+                    ?geo       rdf:type           SO:GeoCoordinates .
+                    ?geo       SO:latitude        ?northBoundCoord .
+                }
+                LIMIT 1
+            ]]>
+        </value>
+    </constructor-arg>
+    <property name="converter" ref="solrLatitudeConverter" />
+</bean>
+
+<!-- Extract east bounding coordinate from a 'SO:spatialCoverage' property of type 'GeoCoordinates'. -->
+<bean id="schema_org_geoCoordinates_east" class="org.dataone.cn.indexer.annotation.SparqlField">
+    <constructor-arg name="name" value="eastBoundCoord" />
+    <constructor-arg name="query">
+        <value>
+            <![CDATA[
+                PREFIX rdf:  <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+                PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+                PREFIX SO:   <http://schema.org/>
+
+                SELECT ?eastBoundCoord
+                WHERE {
+                    ?datasetId rdf:type           SO:Dataset .
+                    ?datasetId SO:spatialCoverage ?spatial .
+                    ?spatial   rdf:type           SO:Place .
+                    ?spatial   SO:geo             ?geo .
+                    ?geo       rdf:type           SO:GeoCoordinates .
+                    ?geo       SO:longitude       ?eastBoundCoord .
+                }
+                LIMIT 1
+            ]]>
+        </value>
+    </constructor-arg>
+    <property name="converter" ref="solrLongitudeConverter" />
+</bean>
+
     <!-- Extract 'SO:spatialCoverage'. This property can have several different forms, for different
          types of spatial coverage representations. if the value is simple text (i.e. not SO:Place),
          then it should be assigned to Solr 'namedLocation' field,
diff --git a/src/test/resources/org/dataone/cn/index/resources/d1_testdocs/json-ld/geocoordinates/geocoordinates.jsonld b/src/test/resources/org/dataone/cn/index/resources/d1_testdocs/json-ld/geocoordinates/geocoordinates.jsonld
new file mode 100644
index 00000000..0fa1e8e5
--- /dev/null
+++ b/src/test/resources/org/dataone/cn/index/resources/d1_testdocs/json-ld/geocoordinates/geocoordinates.jsonld
@@ -0,0 +1,123 @@
+{
+    "@context": {
+      "@vocab": "https://schema.org/",
+      "datacite": "http://purl.org/spar/datacite/"
+    },
+    "@id": "https://doi.org/10.26022/IEDA/111561",
+    "@type": "Dataset",
+    "author": {
+      "@list": [
+        {
+          "@type": "Role",
+          "author": [
+            {
+              "@type": "Person",
+              "familyName": "Carlin",
+              "givenName": "Joseph",
+              "name": "Joseph A Carlin"
+            }
+          ],
+          "roleName": "Lead Author"
+        },
+        {}
+      ]
+    },
+    "citation": [
+      "https://doi.org/10.3389/feart.2019.00113 "
+    ],
+    "dateCreated": "2020-06-01",
+    "description": "This dataset contains sedimentological and geochronological data from 5 sediment cores collected from Monterey Bay, CA, USA. The purpose of this data was to investigate changes in sedimentation on the Monterey Bay shelf over decadal and centennial time scales. All cores were collected from continental shelf areas, in water depths of ~100 m or less, and the cores were collected in 2014 and 2017. Four of the cores were collected using a multi-corer and were less than ~35 cm in length, the fifth core was collected using a gravity corer and was ~90 cm in length. The data include grain size data, 210Pb activities, 137Cs activities, and 14C dates.  ",
+    "distribution": {
+      "@type": "DataDownload",
+      "contentUrl": "https://ecl.earthchem.org/view.php?id=1561",
+      "datePublished": "2020-06-01 00:00:00",
+      "encodingFormat": "application/vnd.ms-excel"
+    },
+    "funding": [
+      {
+        "@id": "http://www.nsf.gov/awardsearch/showAward.do?AwardNumber=9530299",
+        "@type": "MonetaryGrant",
+        "funder": {
+          "@type": "Organization",
+          "name": "National Science Foundation"
+        },
+        "identifier": "9530299",
+        "url": "http://www.nsf.gov/awardsearch/showAward.do?AwardNumber=9530299"
+      },
+      {
+        "@type": "MonetaryGrant",
+        "funder": {
+          "@type": "Organization",
+          "name": "American Chemical Society Petroleum Research Fund"
+        },
+        "identifier": "57363-UNI8"
+      }
+    ],
+    "identifier": {
+      "@type": "PropertyValue",
+      "propertyID": "https://registry.identifiers.org/registry/doi",
+      "url": "https://doi.org/10.26022/IEDA/111561",
+      "value": "doi:10.26022/IEDA/111561"
+    },
+    "inLanguage": "English",
+    "isAccessibleForFree": true,
+    "isBasedOn": [],
+    "keywords": [
+      "Monterey Bay",
+      "California Coast",
+      "Pacific Ocean",
+      "Regional (Continents, Oceans)",
+      "sediment",
+      "continental shelf",
+      "grain size",
+      "210Pb",
+      "137Cs",
+      "14C"
+    ],
+    "license": "https://spdx.org/licenses/CC-BY-SA-4.0",
+    "name": "Variability in Shelf Sedimentation in Response to Fluvial Sediment Supply and Coastal Erosion over the Past 1,000 Years in Monterey Bay, CA, USA. ",
+    "provider": {
+      "@type": "Organization",
+      "name": "EarthChem Library"
+    },
+    "publisher": {
+      "@id": "https://www.earthchem.org",
+      "@type": "Organization",
+      "contactPoint": {
+        "@type": "ContactPoint",
+        "contactType": "Customer Service",
+        "email": "info@earthchem.org",
+        "name": "Information Desk",
+        "url": "https://www.earthchem.org/contact/"
+      },
+      "name": "EarthChem Library",
+      "url": "https://www.earthchem.org/library"
+    },
+    "sameAs": "https://ecl.earthchem.org/view.php?id=1561",
+    "spatialCoverage": {
+      "@type": "Place",
+      "geo": [
+        {
+          "@type": "GeoCoordinates",
+          "latitude": "36.7016",
+          "longitude": "-121.90504"
+        },
+        {
+          "@type": "GeoCoordinates",
+          "latitude": "36.837017",
+          "longitude": "-121.88215"
+        },
+        {
+          "@type": "GeoCoordinates",
+          "latitude": "36.873",
+          "longitude": "-122.03484"
+        },
+        {
+          "@type": "GeoCoordinates",
+          "latitude": "37.04653",
+          "longitude": "-122.404"
+        }
+      ]
+    },
+    "url": "https://doi.org/10.26022/IEDA/111561"
+}
\ No newline at end of file
diff --git a/src/test/resources/org/dataone/cn/index/resources/d1_testdocs/json-ld/geocoordinates/systemmetadata.xml b/src/test/resources/org/dataone/cn/index/resources/d1_testdocs/json-ld/geocoordinates/systemmetadata.xml
new file mode 100644
index 00000000..4da62bf1
--- /dev/null
+++ b/src/test/resources/org/dataone/cn/index/resources/d1_testdocs/json-ld/geocoordinates/systemmetadata.xml
@@ -0,0 +1,26 @@
+<?xml version='1.0' encoding='UTF-8'?>
+<ns1:systemMetadata xmlns:ns1="http://ns.dataone.org/service/types/v2.0">
+    <serialVersion>1</serialVersion>
+  <identifier>geocoordinates.jsonld</identifier>
+  <formatId>science-on-schema.org/Dataset;ld+json</formatId>
+  <size>4207</size>
+  <checksum algorithm="MD5"></checksum>
+  <submitter>dataone_integration_test_user</submitter>
+  <rightsHolder>dataone_integration_test_user</rightsHolder>
+  <accessPolicy>
+    <allow>
+      <subject>dataone_public_user</subject>
+      <permission>read</permission>
+    </allow>
+    <allow>
+      <subject>dataone_integration_test_user</subject>
+      <permission>write</permission>
+    </allow>
+  </accessPolicy>
+  <replicationPolicy replicationAllowed="true"/>
+  <dateUploaded>2024-08-17T14:59:47.171874</dateUploaded>
+  <dateSysMetadataModified>2024-08-17T14:59:47.173344</dateSysMetadataModified>
+  <originMemberNode>test_documents</originMemberNode>
+  <authoritativeMemberNode>test_documents</authoritativeMemberNode>
+</ns1:systemMetadata>
+