feat: #3 integrate with CKAN

_http/ckan.http

@@ -2,7 +2,7 @@
 #
 # SPDX-License-Identifier: Apache-2.0
 
-GET https://ckan-test.healthdata.nl/api/3/action/package_search?facet.field=["organization","theme","conforms_to","has_version","access_rights","language","publisher_name","res_format","provenance"]&rows=1&fq=id:e1b3eff9-13eb-48b0-b180-7ecb76b84454
+GET https://ckan-test.healthdata.nl/api/3/action/package_search?fq=&sort=score+desc%2C+metadata_modified+desc&rows=10&start=0&facet.field=%5B%22organization%22%2C%22theme%22%2C%22conforms_to%22%2C%22has_version%22%2C%22access_rights%22%2C%22language%22%2C%22publisher_name%22%2C%22res_format%22%2C%22provenance%22%5D
 
 ###
-GET https://ckan-test.healthdata.nl/api/3/action/package_show?id=e1b3eff9-13eb-48b0-b180-7ecb76b84454
+GET https://ckan-test.healthdata.nl/api/3/action/package_show?id=e1b3eff9-13eb-48b0-b180-7ecb76b84454

_http/discovery.http

@@ -0,0 +1,26 @@
+# SPDX-FileCopyrightText: 2024 PNED G.I.E.
+#
+# SPDX-License-Identifier: Apache-2.0
+
+POST http://localhost:8080/api/v1/datasets/search
+Content-Type: application/json
+
+{
+    "query": "COVID",
+    "facets": [{
+        "facetGroup": "ckan",
+        "facet": "theme",
+        "value": "http://purl.bioontology.org/ontology/ICD10CM/U07.1"
+    }, {
+        "facetGroup": "ckan",
+        "facet": "theme",
+        "value": "http://purl.org/zonmw/covid19/10003"
+    }, {
+        "facetGroup": "ckan",
+        "facet": "tags",
+        "value": "COVID-19"
+    }]
+}
+
+###
+GET http://localhost:8080/api/v1/datasets/dummy

src/main/java/io/github/genomicdatainfrastructure/discovery/api/DatasetQueryApiImpl.java

@@ -0,0 +1,39 @@
+// SPDX-FileCopyrightText: 2024 PNED G.I.E.
+//
+// SPDX-License-Identifier: Apache-2.0
+
+package io.github.genomicdatainfrastructure.discovery.api;
+
+import io.github.genomicdatainfrastructure.discovery.model.DatasetSearchQuery;
+import io.github.genomicdatainfrastructure.discovery.model.DatasetsSearchResponse;
+import io.github.genomicdatainfrastructure.discovery.model.RetrievedDataset;
+import io.github.genomicdatainfrastructure.discovery.services.DatasetsSearchService;
+import io.quarkus.oidc.runtime.OidcJwtCallerPrincipal;
+import io.quarkus.security.identity.SecurityIdentity;
+import lombok.RequiredArgsConstructor;
+
+@RequiredArgsConstructor
+public class DatasetQueryApiImpl implements DatasetQueryApi {
+
+    private final SecurityIdentity identity;
+    private final DatasetsSearchService datasetsSearchService;
+
+    @Override
+    public DatasetsSearchResponse datasetSearch(DatasetSearchQuery datasetSearchQuery) {
+        return datasetsSearchService.search(accessToken(), datasetSearchQuery);
+    }
+
+    @Override
+    public RetrievedDataset retrieveDataset(String id) {
+        return RetrievedDataset.builder()
+                .build();
+    }
+
+    private String accessToken() {
+        if (identity.isAnonymous()) {
+            return null;
+        }
+        var principal = (OidcJwtCallerPrincipal) identity.getPrincipal();
+        return principal.getRawToken();
+    }
+}

src/main/java/io/github/genomicdatainfrastructure/discovery/api/GlobalExceptionMapper.java

@@ -0,0 +1,37 @@
+// SPDX-FileCopyrightText: 2024 PNED G.I.E.
+//
+// SPDX-License-Identifier: Apache-2.0
+
+package io.github.genomicdatainfrastructure.discovery.api;
+
+import static jakarta.ws.rs.core.Response.Status.INTERNAL_SERVER_ERROR;
+
+import io.github.genomicdatainfrastructure.discovery.model.ErrorResponse;
+import jakarta.ws.rs.core.MediaType;
+import jakarta.ws.rs.core.Response;
+import jakarta.ws.rs.ext.ExceptionMapper;
+import jakarta.ws.rs.ext.Provider;
+import lombok.extern.java.Log;
+import java.util.logging.Level;
+
+@Log
+@Provider
+public class GlobalExceptionMapper implements ExceptionMapper<Exception> {
+
+    @Override
+    public Response toResponse(Exception exception) {
+        log.log(Level.SEVERE, exception, exception::getMessage);
+
+        var errorResponse = ErrorResponse.builder()
+                .title("Not expected exception")
+                .status(INTERNAL_SERVER_ERROR.getStatusCode())
+                .detail(exception.getMessage())
+                .build();
+
+        return Response
+                .status(INTERNAL_SERVER_ERROR)
+                .entity(errorResponse)
+                .type(MediaType.APPLICATION_JSON)
+                .build();
+    }
+}

src/main/java/io/github/genomicdatainfrastructure/discovery/services/CkanFacetsQueryBuilder.java

@@ -0,0 +1,55 @@
+// SPDX-FileCopyrightText: 2024 PNED G.I.E.
+//
+// SPDX-License-Identifier: Apache-2.0
+
+package io.github.genomicdatainfrastructure.discovery.services;
+
+import java.util.List;
+import java.util.Objects;
+
+import io.github.genomicdatainfrastructure.discovery.model.DatasetSearchQueryFacet;
+
+import static java.util.Objects.nonNull;
+import static java.util.Optional.ofNullable;
+import static java.util.stream.Collectors.groupingBy;
+import static java.util.stream.Collectors.joining;
+
+public class CkanFacetsQueryBuilder {
+
+    private static final String CKAN_FACET_GROUP = "ckan";
+    private static final String QUOTED_VALUE = "\"%s\"";
+    private static final String FACET_PATTERN = "%s:(%s)";
+    private static final String AND = " AND ";
+
+    private CkanFacetsQueryBuilder() {
+        // utility class
+    }
+
+    public static String buildFacetQuery(List<DatasetSearchQueryFacet> facets) {
+        var nonNullFacets = ofNullable(facets)
+                .orElseGet(List::of)
+                .stream()
+                .filter(CkanFacetsQueryBuilder::isCkanGroupAndFacetIsNotBlank)
+                .collect(groupingBy(DatasetSearchQueryFacet::getFacet));
+
+        return nonNullFacets.entrySet().stream()
+                .map(entry -> getFacetQuery(entry.getKey(), entry.getValue()))
+                .collect(joining(AND));
+    }
+
+    private static Boolean isCkanGroupAndFacetIsNotBlank(DatasetSearchQueryFacet facet) {
+        return Objects.equals(CKAN_FACET_GROUP, facet.getFacetGroup()) &&
+                nonNull(facet.getFacet()) &&
+                !facet.getFacet().isBlank() &&
+                nonNull(facet.getValue()) &&
+                !facet.getValue().isBlank();
+    }
+
+    private static String getFacetQuery(String key, List<DatasetSearchQueryFacet> facets) {
+        var values = facets.stream()
+                .map(facet -> QUOTED_VALUE.formatted(facet.getValue()))
+                .collect(joining(AND));
+
+        return FACET_PATTERN.formatted(key, values);
+    }
+}

src/main/java/io/github/genomicdatainfrastructure/discovery/services/DatasetsSearchService.java

@@ -0,0 +1,41 @@
+// SPDX-FileCopyrightText: 2024 PNED G.I.E.
+//
+// SPDX-License-Identifier: Apache-2.0
+
+package io.github.genomicdatainfrastructure.discovery.services;
+
+import org.eclipse.microprofile.rest.client.inject.RestClient;
+
+import io.github.genomicdatainfrastructure.discovery.model.DatasetSearchQuery;
+import io.github.genomicdatainfrastructure.discovery.model.DatasetsSearchResponse;
+import io.github.genomicdatainfrastructure.discovery.remote.ckan.api.CkanQueryApi;
+import jakarta.enterprise.context.ApplicationScoped;
+import jakarta.inject.Inject;
+
+@ApplicationScoped
+public class DatasetsSearchService {
+
+    private static final String FACETS_QUERY = "[\"access_rights\",\"theme\",\"tags\",\"spatial_uri\",\"organization\",\"publisher_name\",\"res_format\"]";
+
+    private final CkanQueryApi ckanQueryApi;
+
+    @Inject
+    public DatasetsSearchService(
+            @RestClient CkanQueryApi ckanQueryApi
+    ) {
+        this.ckanQueryApi = ckanQueryApi;
+    }
+
+    public DatasetsSearchResponse search(String accessToken, DatasetSearchQuery query) {
+        var response = ckanQueryApi.packageSearch(
+                query.getQuery(),
+                CkanFacetsQueryBuilder.buildFacetQuery(query.getFacets()),
+                query.getSort(),
+                query.getRows(),
+                query.getStart(),
+                FACETS_QUERY,
+                accessToken
+        );
+        return PackagesSearchResponseMapper.from(response);
+    }
+}

src/main/java/io/github/genomicdatainfrastructure/discovery/services/PackagesSearchResponseMapper.java

@@ -0,0 +1,126 @@
+// SPDX-FileCopyrightText: 2024 PNED G.I.E.
+//
+// SPDX-License-Identifier: Apache-2.0
+
+package io.github.genomicdatainfrastructure.discovery.services;
+
+import io.github.genomicdatainfrastructure.discovery.model.DatasetsSearchResponse;
+import io.github.genomicdatainfrastructure.discovery.model.Facet;
+import io.github.genomicdatainfrastructure.discovery.model.FacetGroup;
+import io.github.genomicdatainfrastructure.discovery.model.SearchedDataset;
+import io.github.genomicdatainfrastructure.discovery.model.ValueLabel;
+import io.github.genomicdatainfrastructure.discovery.remote.ckan.model.CkanFacet;
+import io.github.genomicdatainfrastructure.discovery.remote.ckan.model.CkanOrganization;
+import io.github.genomicdatainfrastructure.discovery.remote.ckan.model.CkanPackage;
+import io.github.genomicdatainfrastructure.discovery.remote.ckan.model.PackagesSearchResponse;
+import io.github.genomicdatainfrastructure.discovery.remote.ckan.model.PackagesSearchResult;
+import java.util.List;
+import java.util.Map;
+import java.time.format.DateTimeFormatter;
+import java.time.LocalDateTime;
+
+import static java.util.Optional.ofNullable;
+
+public class PackagesSearchResponseMapper {
+
+    private static final DateTimeFormatter DATE_FORMATTER = DateTimeFormatter.ofPattern(
+            "yyyy-MM-dd'T'HH:mm:ss.SSSSSS"
+    );
+
+    private PackagesSearchResponseMapper() {
+        // Utility class
+    }
+
+    public static DatasetsSearchResponse from(PackagesSearchResponse response) {
+        return DatasetsSearchResponse.builder()
+                .count(count(response.getResult()))
+                .facetGroups(facetGroups(response.getResult()))
+                .results(results(response.getResult()))
+                .build();
+    }
+
+    private static Integer count(PackagesSearchResult result) {
+        return ofNullable(result)
+                .map(PackagesSearchResult::getCount)
+                .orElse(null);
+    }
+
+    private static List<FacetGroup> facetGroups(PackagesSearchResult result) {
+        return ofNullable(result)
+                .map(PackagesSearchResult::getSearchFacets)
+                .map(PackagesSearchResponseMapper::facetGroup)
+                .map(List::of)
+                .orElseGet(List::of);
+    }
+
+    private static FacetGroup facetGroup(Map<String, CkanFacet> facets) {
+        return FacetGroup.builder()
+                .key("ckan")
+                .label("Metadata")
+                .facets(facets.entrySet().stream()
+                        .map(PackagesSearchResponseMapper::facet)
+                        .toList())
+                .build();
+    }
+
+    private static Facet facet(Map.Entry<String, CkanFacet> entry) {
+        var key = entry.getKey();
+        var facet = entry.getValue();
+        var values = ofNullable(facet.getItems())
+                .orElseGet(List::of)
+                .stream().map(value -> ValueLabel.builder()
+                        .value(value.getName())
+                        .label(value.getDisplayName())
+                        .build())
+                .toList();
+
+        return Facet.builder()
+                .key(key)
+                .label(facet.getTitle())
+                .values(values)
+                .build();
+    }
+
+    private static List<SearchedDataset> results(PackagesSearchResult result) {
+        var nonNullDatasets = ofNullable(result)
+                .map(PackagesSearchResult::getResults)
+                .orElseGet(List::of);
+
+        return nonNullDatasets.stream()
+                .map(PackagesSearchResponseMapper::result)
+                .toList();
+    }
+
+    private static SearchedDataset result(CkanPackage dataset) {
+        var catalogue = ofNullable(dataset.getOrganization())
+                .map(CkanOrganization::getTitle)
+                .orElse(null);
+
+        return SearchedDataset.builder()
+                .id(dataset.getId())
+                .identifier(dataset.getIdentifier())
+                .title(dataset.getName())
+                .description(dataset.getNotes())
+                .themes(values(dataset.getTheme()))
+                .catalogue(catalogue)
+                .modifiedAt(parse(dataset.getMetadataModified()))
+                .build();
+    }
+
+    private static LocalDateTime parse(String date) {
+        return ofNullable(date)
+                .map(it -> LocalDateTime.parse(it, DATE_FORMATTER))
+                .orElse(null);
+    }
+
+    private static List<ValueLabel> values(List<String> values) {
+        return ofNullable(values)
+                .orElseGet(List::of)
+                .stream()
+                .map(it -> ValueLabel.builder()
+                        .value(it)
+                        .label(it)
+                        .build())
+                .toList();
+    }
+}

src/main/openapi/ckan.yaml

@@ -102,11 +102,19 @@ paths:
 components:
   schemas:
     PackagesSearchResponse:
+      type: object
+      properties:
+        help:
+          type: string
+        success:
+          type: boolean
+        result:
+          $ref: "#/components/schemas/PackagesSearchResult"
+    PackagesSearchResult:
       type: object
       properties:
         count:
           type: integer
-          description: The number of results found
         results:
           type: array
           items:
@@ -115,7 +123,6 @@ components:
           type: object
           additionalProperties:
             $ref: "#/components/schemas/CkanFacet"
-          description: Aggregated information about facet
     CkanFacet:
       type: object
       properties:
@@ -159,10 +166,8 @@ components:
           $ref: "#/components/schemas/CkanOrganization"
         metadata_created:
           type: string
-          format: date-time
         metadata_modified:
           type: string
-          format: date-time
         url:
           type: string
         language:
@@ -220,10 +225,8 @@ components:
           type: string
         created:
           type: string
-          format: date-time
         last_modified:
           type: string
-          format: date-time
       required:
         - id
         - name