Expect single string URL for Aardvark dct_references_s

Why these changes are being introduced:
GeoHarvester will begin putting the external URL for MITAardvark records as a single string in the
dct_references_s field, as opposed to an array of objects.

How this addresses that need:
* Expects the external URL as the direct value for key 'http://schema.org/url'

Side effects of this change:
* None; downstream records for TIM and Opensearch will look the same

Relevant ticket(s):
* https://mitlibraries.atlassian.net/browse/GDT-68

tests/fixtures/aardvark/aardvark_record_all_fields.jsonl

@@ -1 +1 @@
-{"id": "mit:123", "dcat_bbox": "ENVELOPE(-111.1, -104.0, 45.0, 40.9)", "dcat_keyword_sm": ["Country"], "dcat_theme_sm": ["Political boundaries"], "dct_accessRights_s": "Access note", "dct_alternative_sm": ["Alternate title"], "dct_creator_sm": ["Smith, Jane", "Smith, John"], "dct_description_sm": ["A description"], "dct_format_s": "Shapefile", "dct_identifier_sm": ["abc123"], "dct_issued_s": "2003-10-23", "dct_language_sm": ["eng"], "dct_license_sm": ["http://license.license", "http://another_license.another_license"], "dct_publisher_sm": ["ML InfoMap (Firm)"], "dct_references_s": "{\"http://schema.org/downloadUrl\": [{\"label\": \"Source Metadata\", \"url\": \"https://cdn.dev1.mitlibrary.net/geo/public/GISPORTAL_GISOWNER01_BOSTONWATER95.source.fgdc.xml\"}, {\"label\": \"Aardvark Metadata\", \"url\": \"https://cdn.dev1.mitlibrary.net/geo/public/GISPORTAL_GISOWNER01_BOSTONWATER95.normalized.aardvark.json\"}, {\"label\": \"Data\", \"url\": \"https://cdn.dev1.mitlibrary.net/geo/public/GISPORTAL_GISOWNER01_BOSTONWATER95.zip\"}], \"http://schema.org/url\": [{\"label\": \"Website\", \"url\": \"https://search.libraries.mit.edu/record/gismit:GISPORTAL_GISOWNER01_BOSTONWATER95\"}]}", "dct_rights_sm": ["Some person has the rights"], "dct_rightsHolder_sm": ["The person with the rights", "Another person with the rights"], "dct_spatial_sm": ["Some city, Some country"], "dct_subject_sm": ["Geography", "Earth"], "dct_temporal_sm": ["1943", "1979"], "dct_title_s": "Test title 1", "gbl_dateRange_drsim": ["[1943 TO 1946]"], "gbl_displayNote_sm": ["Danger: This text will be displayed in a red box","Info: This text will be displayed in a blue box","Tip: This text will be displayed in a green box","Warning: This text will be displayed in a yellow box","This is text without a tag and it will be assigned default 'note' style"], "gbl_indexYear_im": [1943,1944,1945,1946], "gbl_resourceClass_sm": ["Dataset"], "gbl_resourceType_sm": ["Vector data"], "gbl_suppressed_b": false, "locn_geometry": "ENVELOPE(-111.1, -104.0, 45.0, 40.9)", "schema_provider_s": "MIT"}
+{"id": "mit:123", "dcat_bbox": "ENVELOPE(-111.1, -104.0, 45.0, 40.9)", "dcat_keyword_sm": ["Country"], "dcat_theme_sm": ["Political boundaries"], "dct_accessRights_s": "Access note", "dct_alternative_sm": ["Alternate title"], "dct_creator_sm": ["Smith, Jane", "Smith, John"], "dct_description_sm": ["A description"], "dct_format_s": "Shapefile", "dct_identifier_sm": ["abc123"], "dct_issued_s": "2003-10-23", "dct_language_sm": ["eng"], "dct_license_sm": ["http://license.license", "http://another_license.another_license"], "dct_publisher_sm": ["ML InfoMap (Firm)"], "dct_references_s": "{\"http://schema.org/downloadUrl\": [{\"label\": \"Source Metadata\", \"url\": \"https://cdn.dev1.mitlibrary.net/geo/public/GISPORTAL_GISOWNER01_BOSTONWATER95.source.fgdc.xml\"}, {\"label\": \"Aardvark Metadata\", \"url\": \"https://cdn.dev1.mitlibrary.net/geo/public/GISPORTAL_GISOWNER01_BOSTONWATER95.normalized.aardvark.json\"}, {\"label\": \"Data\", \"url\": \"https://cdn.dev1.mitlibrary.net/geo/public/GISPORTAL_GISOWNER01_BOSTONWATER95.zip\"}], \"http://schema.org/url\": \"https://geodata.libraries.mit.edu/record/gismit:GISPORTAL_GISOWNER01_BOSTONWATER95\"}", "dct_rights_sm": ["Some person has the rights"], "dct_rightsHolder_sm": ["The person with the rights", "Another person with the rights"], "dct_spatial_sm": ["Some city, Some country"], "dct_subject_sm": ["Geography", "Earth"], "dct_temporal_sm": ["1943", "1979"], "dct_title_s": "Test title 1", "gbl_dateRange_drsim": ["[1943 TO 1946]"], "gbl_displayNote_sm": ["Danger: This text will be displayed in a red box","Info: This text will be displayed in a blue box","Tip: This text will be displayed in a green box","Warning: This text will be displayed in a yellow box","This is text without a tag and it will be assigned default 'note' style"], "gbl_indexYear_im": [1943,1944,1945,1946], "gbl_resourceClass_sm": ["Dataset"], "gbl_resourceType_sm": ["Vector data"], "gbl_suppressed_b": false, "locn_geometry": "ENVELOPE(-111.1, -104.0, 45.0, 40.9)", "schema_provider_s": "MIT"}

tests/sources/json/test_aardvark.py

@@ -180,7 +180,7 @@ def test_aardvark_get_links_success(aardvark_record_all_fields):
             text="Data",
         ),
         timdex.Link(
-            url="https://search.libraries.mit.edu/record/gismit"
+            url="https://geodata.libraries.mit.edu/record/gismit"
             ":GISPORTAL_GISOWNER01_BOSTONWATER95",
             kind="Website",
             text="Website",

transmogrifier/sources/json/aardvark.py

@@ -274,7 +274,14 @@ def get_identifiers(source_record: dict) -> list[timdex.Identifier]:
 
     @staticmethod
     def get_links(source_record: dict, source_record_id: str) -> list[timdex.Link]:
-        """Get values from source record for TIMDEX links field."""
+        """Get values from source record for TIMDEX links field.
+
+        The dct_references_s is a JSON string following a particular format defined here:
+        https://opengeometadata.org/ogm-aardvark/#references.  Keys in the parsed JSON
+        object define what kind of URL it is.  This is a flat mapping of namespace:url,
+        except in the case of 'http://schema.org/downloadUrl' which will be an array of
+        complex objects.
+        """
         links = []
         links_string = source_record["dct_references_s"]
         try:
@@ -287,14 +294,8 @@ def get_links(source_record: dict, source_record_id: str) -> list[timdex.Link]:
                     for link in links_object.get("http://schema.org/downloadUrl", [])
                 ]
             )
-            links.extend(
-                [
-                    timdex.Link(
-                        url=link.get("url"), kind="Website", text=link.get("label")
-                    )
-                    for link in links_object.get("http://schema.org/url", [])
-                ]
-            )
+            if schema_url := links_object.get("http://schema.org/url"):
+                links.append(timdex.Link(url=schema_url, kind="Website", text="Website"))
         except ValueError:
             message = (
                 f"Record ID '{source_record_id}': Unable to parse "