add support for multilingual RDF

README.md

@@ -805,7 +805,40 @@ This makes it very easy to display the fields in the current language.
 To disable this behavior, you can set the following config value in your ini file (default: True):
 
     ckanext.dcat.translate_keys = False
+
+## Multilingual RDF support
+To add multilingual values from RDF to ckan, the function `_object_value` can be called with optional parameter `multilang=true` (defaults to `false`)). 
+If `_object_value` is called with the `multilang=true`-parameter, but no language-attribute is found, the value will be added as Literal with the default language (en).
+
+Example RDF:
+```xml
+<dct:title xml:lang="en">Dataset Title (EN)</dct:title>
+<dct:title xml:lang="de">Dataset Title (DE)</dct:title>
+<dct:title xml:lang="fr">Dataset Title (FR)</dct:title>
+```
+```json
+{
+    "title":
+        {
+            "en": "Dataset Title (EN)",
+            "de": "Dataset Title (DE)",
+            "fr": "Dataset Title (FR)"
+        }
+}
+```
 
+Example with missing language in RDF:
+```xml
+<dct:title>Dataset Title</dct:title>
+```
+```json
+{
+    "title":
+        {
+            "en": "Dataset Title"
+        }
+}
+```
 
 ## Structured data and Google Dataset Search indexing
 

ckanext/dcat/profiles.py

@@ -18,7 +18,7 @@
 from ckan.lib.munge import munge_tag
 from ckan.lib.helpers import url_for
 
-from ckanext.dcat.utils import resource_uri, publisher_uri_from_dataset_dict, DCAT_EXPOSE_SUBCATALOGS, DCAT_CLEAN_TAGS
+from ckanext.dcat.utils import resource_uri, publisher_uri_from_dataset_dict, DCAT_EXPOSE_SUBCATALOGS, DCAT_CLEAN_TAGS, get_langs
 
 DCT = Namespace("http://purl.org/dc/terms/")
 DCAT = Namespace("http://www.w3.org/ns/dcat#")
@@ -133,16 +133,27 @@ def _object(self, subject, predicate):
             return _object
         return None
 
-    def _object_value(self, subject, predicate):
+    def _object_value(self, subject, predicate, multilang=False):
         '''
         Given a subject and a predicate, returns the value of the object
-
         Both subject and predicate must be rdflib URIRef or BNode objects
-
         If found, the unicode representation is returned, else an empty string
         '''
+        default_lang = config.get('ckan.locale_default', 'en')
+        lang_dict = {}
         for o in self.g.objects(subject, predicate):
-            return unicode(o)
+            if multilang and o.language:
+                lang_dict[o.language] = unicode(o)
+            elif multilang:
+                lang_dict[default_lang] = unicode(o)
+            else:
+                return unicode(o)
+        if multilang:
+            # when translation does not exist, create an empty one
+            for lang in get_langs():
+                if lang not in lang_dict:
+                    lang_dict[lang] = ''
+            return lang_dict
         return ''
 
     def _object_value_int(self, subject, predicate):
@@ -489,19 +500,22 @@ def _add_list_triples_from_dict(self, _dict, subject, items):
 
     def _add_triples_from_dict(self, _dict, subject, items,
                                list_value=False,
-                               date_value=False):
+                               date_value=False,
+                               multilang=False):
         for item in items:
             key, predicate, fallbacks, _type = item
             self._add_triple_from_dict(_dict, subject, predicate, key,
                                        fallbacks=fallbacks,
                                        list_value=list_value,
                                        date_value=date_value,
+                                       multilang=multilang,
                                        _type=_type)
 
     def _add_triple_from_dict(self, _dict, subject, predicate, key,
                               fallbacks=None,
                               list_value=False,
                               date_value=False,
+                              multilang=False,
                               _type=Literal,
                               value_modifier=None):
         '''
@@ -535,13 +549,25 @@ def _add_triple_from_dict(self, _dict, subject, predicate, key,
             self._add_list_triple(subject, predicate, value, _type)
         elif value and date_value:
             self._add_date_triple(subject, predicate, value, _type)
+        elif value and multilang:
+            self._add_multilang_triple(subject, predicate, value)
         elif value:
             # Normal text value
             # ensure URIRef items are preprocessed (space removal/url encoding)
             if _type == URIRef:
                 _type = CleanedURIRef
             self.g.add((subject, predicate, _type(value)))
 
+    def _add_multilang_triple(self, subject, predicate, multilang_values):  # noqa
+         for key, values in multilang_values.iteritems():
+            if values:
+                # the values can be either a multilang-dict or they are
+                # nested in another iterable (e.g. keywords)
+                if not hasattr(values, '__iter__'):
+                    values = [values]
+                for value in values:
+                    self.g.add((subject, predicate, Literal(value, lang=key)))  # noqa
+
     def _add_list_triple(self, subject, predicate, value, _type=Literal):
         '''
         Adds as many triples to the graph as values

ckanext/dcat/tests/test_base_profile.py

@@ -149,6 +149,35 @@ def test_object_int_wrong_value(self):
 
         eq_(value, None)
 
+    def test_object_value_multilang(self):
+
+        p = RDFProfile(_default_graph())
+
+        p.g.add((URIRef('http://example.org/datasets/1'),
+                 DCT.title, Literal('Test Datensatz 1', lang='de')))
+
+        p.g.add((URIRef('http://example.org/datasets/1'),
+                 DCT.title, Literal('Test Dataset 1', lang='en')))
+
+        value = p._object_value(URIRef('http://example.org/datasets/1'),
+                                DCT.title,
+                                multilang=True)
+
+        assert isinstance(value, dict)
+        eq_(value.get('de'), u'Test Datensatz 1')
+        eq_(value.get('en'), u'Test Dataset 1')
+
+    def test_object_value_multilang_missing_lang_param(self):
+
+        p = RDFProfile(_default_graph())
+
+        value = p._object_value(URIRef('http://example.org/datasets/1'),
+                                DCT.title,
+                                multilang=True)
+
+        assert isinstance(value, dict)
+        eq_(value.get('en'), u'Test Dataset 1')
+
     def test_object_list(self):
 
         p = RDFProfile(_default_graph())

ckanext/dcat/utils.py

@@ -337,3 +337,8 @@ def parse_accept_header(accept_header=''):
                 return accepted_media_types_wildcard[_type]
 
     return None
+
+
+def get_langs():
+    language_priorities = config.get('ckan.locales_offered', '').split()
+    return language_priorities