documents: export records in Dublic Core format

This PR configures OAI server to export documents records in Dublin Core formats.

* Installs `dcxml` library to generate the output in Dublic CORE XML.
* Creates and configures the serializer for `oai_dc` format.
* Dumps `mimetype` from file object to record's file metadata.
* Adds a serializer schema for Dublic Core.
* Adds `mimetype` property to all resources that use files.
* Avoids to display issue and pages if volume is not defined in `partOf` property.
* Adds a custom serializers for dumping DC objects.
* Closes #325.

Co-Authored-by: Sébastien Délèze <[email protected]>

pyproject.toml

@@ -30,6 +30,7 @@ flask-cors = ">3.0.8"
 nbconvert = {version = ">=5.6.1,<6.0.0", extras = ["execute"]}
 cryptography = ">=3.2"
 netaddr = "*"
+dcxml = "*"
 
 [tool.poetry.dev-dependencies]
 Flask-Debugtoolbar = ">=0.10.1"

sonar/config.py

@@ -665,3 +665,13 @@ def _(x):
 OAISERVER_ID_PREFIX = 'oai:sonar.ch:'
 #: XSL file
 OAISERVER_XSL_URL = 'static/xsl/oai2.xsl'
+#: Export formats
+OAISERVER_METADATA_FORMATS = {
+    'oai_dc': {
+        'namespace': 'http://www.openarchives.org/OAI/2.0/oai_dc/',
+        'schema': 'http://www.openarchives.org/OAI/2.0/oai_dc.xsd',
+        'serializer': 'sonar.modules.documents.serializers.oaipmh_oai_dc',
+    }
+}
+#: Number of records to return per page in OAI-PMH results.
+OAISERVER_PAGE_SIZE = 100

sonar/modules/api.py

@@ -31,14 +31,32 @@
 from invenio_jsonschemas import current_jsonschemas
 from invenio_pidstore.errors import PIDDoesNotExistError
 from invenio_pidstore.models import PersistentIdentifier, PIDStatus
-from invenio_records_files.api import FilesMixin, Record
+from invenio_records_files.api import FileObject as InvenioFileObjet
+from invenio_records_files.api import FilesMixin as InvenioFilesMixin
+from invenio_records_files.api import Record
 from invenio_records_files.models import RecordsBuckets
 from invenio_records_rest.utils import obj_or_import_string
 from invenio_search import current_search
 from invenio_search.api import RecordsSearch
 from sqlalchemy.orm.exc import NoResultFound
 
 
+class FileObject(InvenioFileObjet):
+    """Wrapper for files."""
+
+    def dumps(self):
+        """Create a dump of the metadata associated to the record."""
+        super(FileObject, self).dumps()
+        self.data.update({'mimetype': self.obj.mimetype})
+        return self.data
+
+
+class FilesMixin(InvenioFilesMixin):
+    """Implement files attribute for Record models."""
+
+    file_cls = FileObject
+
+
 class SonarRecord(Record, FilesMixin):
     """SONAR Record."""
 

sonar/modules/deposits/jsonschemas/deposits/deposit-v1.0.0_src.json

@@ -50,6 +50,11 @@
             "type": "string",
             "minLength": 1
           },
+          "mimetype": {
+            "title": "MIME type",
+            "type": "string",
+            "minLength": 1
+          },
           "checksum": {
             "title": "Checksum",
             "description": "MD5 checksum of the file.",

sonar/modules/deposits/mappings/v7/deposits/deposit-v1.0.0.json

@@ -27,6 +27,9 @@
           "key": {
             "type": "keyword"
           },
+          "mimetype": {
+            "type": "keyword"
+          },
           "checksum": {
             "type": "keyword"
           },

sonar/modules/documents/jsonschemas/documents/document-v1.0.0_src.json

@@ -49,6 +49,11 @@
             "type": "string",
             "minLength": 1
           },
+          "mimetype": {
+            "title": "MIME type",
+            "type": "string",
+            "minLength": 1
+          },
           "checksum": {
             "title": "Checksum",
             "description": "MD5 checksum of the file.",

sonar/modules/documents/mappings/v7/documents/document-v1.0.0.json

@@ -37,6 +37,9 @@
           "key": {
             "type": "keyword"
           },
+          "mimetype": {
+            "type": "keyword"
+          },
           "checksum": {
             "type": "keyword"
           },

sonar/modules/documents/marshmallow/json.py

@@ -56,6 +56,7 @@ class Meta:
     file_id = SanitizedUnicode()
     version_id = SanitizedUnicode()
     key = SanitizedUnicode()
+    mimetype = SanitizedUnicode()
     checksum = SanitizedUnicode()
     size = fields.Number()
     label = SanitizedUnicode()

sonar/modules/documents/serializers/__init__.py

@@ -25,6 +25,8 @@
 from invenio_records_rest.serializers.response import record_responsify, \
     search_responsify
 
+from sonar.modules.documents.serializers.dc import SonarDublinCoreSerializer
+from sonar.modules.documents.serializers.schemas.dc import DublinCoreV1
 from sonar.modules.organisations.api import OrganisationRecord
 from sonar.modules.serializers import JSONSerializer as _JSONSerializer
 from sonar.modules.users.api import current_user_record
@@ -72,6 +74,8 @@ def post_process_serialize_search(self, results, pid_fetcher):
 # ===========
 #: JSON serializer definition.
 json_v1 = JSONSerializer(DocumentSchemaV1)
+#: Dublin Core serializer
+dc_v1 = SonarDublinCoreSerializer(DublinCoreV1, replace_refs=True)
 
 # Records-REST serializers
 # ========================
@@ -85,3 +89,8 @@ def post_process_serialize_search(self, results, pid_fetcher):
     'json_v1_response',
     'json_v1_search',
 )
+
+# OAI-PMH record serializers.
+# ===========================
+#: OAI-PMH OAI Dublin Core record serializer.
+oaipmh_oai_dc = dc_v1.serialize_oaipmh

sonar/modules/documents/serializers/dc.py

@@ -0,0 +1,32 @@
+# -*- coding: utf-8 -*-
+#
+# Swiss Open Access Repository
+# Copyright (C) 2019 RERO
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as published by
+# the Free Software Foundation, version 3 of the License.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU Affero General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+"""Dublin Core serializer."""
+
+from invenio_records_rest.serializers.dc import DublinCoreSerializer
+
+
+class SonarDublinCoreSerializer(DublinCoreSerializer):
+    """Marshmallow based DublinCore serializer for records."""
+
+    def dump(self, obj, context=None):
+        """Serialize object with schema.
+
+        Mandatory to override this method, as invenio-records-rest does not
+        use the right way to dump objects (compatible with marshmallow 3.9).
+        """
+        return self.schema_class(context=context).dump(obj)