Merge pull request #125 from opendata-swiss/feat/harvest-qualified-re…

…lations-field

fix: Fix mapping of see_alsos

.github/workflows/test.yml

@@ -64,7 +64,8 @@ jobs:
     - name: Create ckan container
       run: |
         docker create --name test_ckan --network ${{ job.container.network }} --network-alias ckan \
-          -e "HOME=/github/home" -e GITHUB_ACTIONS=true -e CI=true -v "/var/run/docker.sock":"/var/run/docker.sock" \
+          -e "HOME=/github/home" -e GITHUB_ACTIONS=true -e CI=true -e CKAN_SITE_URL=http://test.ckan.net \
+          -v "/var/run/docker.sock":"/var/run/docker.sock" \
           -v "/home/runner/work":"/__w" -v "/home/runner/work/_temp":"/__w/_temp" \
           -v "/home/runner/work/_actions":"/__w/_actions" -v "/opt/hostedtoolcache":"/__t" \
           -v "/home/runner/work/_temp/_github_home":"/github/home" \

ckanext/geocat/tests/test_dataset_metadata.py

@@ -60,7 +60,7 @@ def test_fields(self):
             'spatial',
             'coverage',
             'accrual_periodicity',
-            'see_alsos',
+            'qualified_relations',
             'owner_org',
             'resources'
         ]
@@ -205,9 +205,15 @@ def test_fields_values(self):
         self.assertEquals('', dataset.get('accrual_periodicity'))
 
         # see alsos
-        self.assertTrue(hasattr(dataset['see_alsos'], '__iter__'))
-        self.assertEquals(1, len(dataset['see_alsos']))
-        self.assertEquals('8454f7d9-e3f2-4cc7-be6d-a82196660ccd@swisstopo', dataset['see_alsos'][0])  # noqa
+        self.assertTrue(hasattr(dataset['qualified_relations'], '__iter__'))
+        self.assertEquals(1, len(dataset['qualified_relations']))
+        self.assertEquals(
+            {
+                'relation': 'http://test.ckan.net/perma/8454f7d9-e3f2-4cc7-be6d-a82196660ccd@swisstopo',
+                'had_role': 'http://www.iana.org/assignments/relation/related',
+            },
+            dataset['qualified_relations'][0]
+        )
 
     def test_fields_values_de_only(self):
         xml = self._load_xml('only_de.xml')

ckanext/geocat/utils/csw_mapping.py

@@ -46,7 +46,7 @@
 ]
 GMD_DESCRIPTION = '//gmd:identificationInfo//gmd:abstract'
 GMD_RIGHTS = './/gmd:resourceConstraints//gmd:otherConstraints'
-GMD_SEE_ALSOS = '//gmd:identificationInfo//gmd:aggregationInfo//gmd:aggregateDataSetIdentifier/gmd:MD_Identifier/gmd:code/gco:CharacterString/text()'  # noqa
+GMD_QUALIFIED_RELATIONS = '//gmd:identificationInfo//gmd:aggregationInfo//gmd:aggregateDataSetIdentifier/gmd:MD_Identifier/gmd:code/gco:CharacterString/text()'  # noqa
 GMD_TEMPORAL_START = '//gmd:identificationInfo//gmd:extent//gmd:temporalElement//gml:TimePeriod/gml:beginPosition/text()'  # noqa
 GMD_TEMPORAL_END = '//gmd:identificationInfo//gmd:extent//gmd:temporalElement//gml:TimePeriod/gml:endPosition/text()'  # noqa
 GMD_LANGUAGE = ['//gmd:identificationInfo//gmd:language/gco:CharacterString/text()',  # noqa
@@ -123,10 +123,11 @@ def get_metadata(self, csw_record_as_string, geocat_id):
         dataset_dict['coverage'] = _map_dataset_coverage()
         dataset_dict['spatial'] = _map_dataset_spatial(node=root_node)
         dataset_dict['temporals'] = _map_dataset_temporals(node=root_node)
-        dataset_dict['see_alsos'] = \
-            _map_dataset_see_alsos(node=root_node,
-                                   organization_slug=self.organization_slug,
-                                   valid_identifiers=self.valid_identifiers)
+        dataset_dict['qualified_relations'] = \
+            _map_dataset_qualified_relations(
+                node=root_node,
+                organization_slug=self.organization_slug,
+                valid_identifiers=self.valid_identifiers)
         dataset_dict['owner_org'] = self.organization_slug
 
         rights = \
@@ -162,7 +163,7 @@ def get_metadata(self, csw_record_as_string, geocat_id):
                 dataset_dict['resources'].append(ogdch_service)
 
         # Map geocat permalink as relation
-        dataset_dict['relations'].append(ogdch_map_utils.get_permalink(
+        dataset_dict['relations'].append(ogdch_map_utils.get_geocat_permalink(
             geocat_id=geocat_id,
             geocat_perma_link=self.geocat_perma_link,
             geocat_perma_label=self.geocat_perma_label,
@@ -389,15 +390,16 @@ def _map_dataset_temporals(node):
             geocat_temporal_end)
 
 
-def _map_dataset_see_alsos(node, organization_slug, valid_identifiers):
-    geocat_see_alsos = \
+def _map_dataset_qualified_relations(node, organization_slug,
+                                     valid_identifiers):
+    geocat_qualified_relations = \
         xpath_utils.xpath_get_all_sub_nodes_for_node_and_path(
             node=node,
-            path=GMD_SEE_ALSOS)
-    if geocat_see_alsos:
+            path=GMD_QUALIFIED_RELATIONS)
+    if geocat_qualified_relations:
         return \
-            ogdch_map_utils.map_see_alsos(
-                geocat_see_alsos,
+            ogdch_map_utils.map_qualified_relations(
+                geocat_qualified_relations,
                 organization_slug,
                 valid_identifiers)
     return []

ckanext/geocat/utils/ogdch_map_utils.py

@@ -4,6 +4,7 @@
 from datetime import datetime
 from ckan.lib.munge import munge_tag
 from ckanext.geocat.utils import xpath_utils  # noqa
+import ckan.plugins.toolkit as tk
 import ckanext.geocat.utils.mapping_utils as mu
 
 ORGANIZATION_URI_BASE = 'https://opendata.swiss/organization/'
@@ -142,12 +143,17 @@ def map_language(geocat_language):
     return language_mapping.get(geocat_language, '')
 
 
-def map_see_alsos(geocat_see_alsos, organization_slug, valid_identifiers):
-    ogdch_see_alsos = [
+def map_qualified_relations(geocat_qualified_relations, organization_slug,
+                            valid_identifiers):
+    ogdch_qualified_relations = [
         map_geocat_to_ogdch_identifier(geocat_identifier, organization_slug)
-        for geocat_identifier in geocat_see_alsos]
-    return [see_also for see_also in ogdch_see_alsos
-            if see_also in valid_identifiers]
+        for geocat_identifier in geocat_qualified_relations]
+    return [
+        {'relation': get_ogdch_permalink(qualified_relation),
+         'had_role': "http://www.iana.org/assignments/relation/related"}
+        for qualified_relation in ogdch_qualified_relations
+        if qualified_relation in valid_identifiers
+    ]
 
 
 def map_temporals(geocat_temporal_start, geocat_temporal_end):
@@ -162,11 +168,16 @@ def map_temporals(geocat_temporal_start, geocat_temporal_end):
         return []
 
 
-def get_permalink(geocat_id, geocat_perma_link, geocat_perma_label):
+def get_geocat_permalink(geocat_id, geocat_perma_link, geocat_perma_label):
     permalink = geocat_perma_link + geocat_id
     return {'url': permalink, 'label': geocat_perma_label}
 
 
+def get_ogdch_permalink(identifier):
+    site_url = tk.config.get('ckan.site_url')
+    return u'{0}/perma/{1}'.format(site_url, identifier)
+
+
 def get_legal_basis_link(legal_basis_url):
     LEGAL_BASIS_LABEL = 'legal_basis'
     return {'url': legal_basis_url, 'label': LEGAL_BASIS_LABEL}