feat: Map full media_type onto resource, not just subtype

opendata-swiss · Dec 4, 2023 · 138f989 · 138f989
1 parent 236bd96
commit 138f989
Show file tree

Hide file tree

Showing 3 changed files with 4 additions and 4 deletions.
diff --git a/ckanext/dcatapchharvest/profiles.py b/ckanext/dcatapchharvest/profiles.py
@@ -252,10 +252,10 @@ def _get_iana_media_type(self, subject):
         if isinstance(media_type_value_raw, dict):
             log.debug("The media type object is a dictionary type.")
         else:
-            pattern = r'[^/]+$'  # Match characters that are not '/'
+            pattern = r'http://www.iana.org/assignments/(.+/.+)$'
             media_type_value_re = re.search(pattern, media_type_value_raw)
             if media_type_value_re:
-                media_type_value = media_type_value_re.group(0)
+                media_type_value = media_type_value_re.group(1)
             else:
                 media_type_value = media_type_value_raw
 

diff --git a/ckanext/dcatapchharvest/tests/fixtures/dataset.json b/ckanext/dcatapchharvest/tests/fixtures/dataset.json
@@ -138,7 +138,7 @@
       "temporal_resolution":"P1D",
       "rights": "http://dcat-ap.ch/vocabulary/licenses/terms_by_ask",
       "license": "http://dcat-ap.ch/vocabulary/licenses/cc-by/4.0",
-      "media_type": "1d-interleaved-parityfec"
+      "media_type": "application/1d-interleaved-parityfec"
     }
   ],
   "extras": [

diff --git a/ckanext/dcatapchharvest/tests/test_dcatap_ch_parse.py b/ckanext/dcatapchharvest/tests/test_dcatap_ch_parse.py
@@ -144,7 +144,7 @@ def test_dataset_all_fields(self):
         assert all(l in resource['description'] for l in ['de', 'fr', 'it', 'en']), "resource description contains all languages"
         eq_(resource['description']['de'], u'')
         eq_(resource['format'], u'html')
-        eq_(resource['media_type'], u'html')
+        eq_(resource['media_type'], u'text/html')
         eq_(resource['identifier'], u'346265-fr@bundesamt-fur-statistik-bfs')
         eq_(resource['rights'], u'NonCommercialAllowed-CommercialAllowed-ReferenceRequired')
         eq_(resource['license'], u'Creative Commons CC Zero License (cc-zero)')