datacite: fix a raw name constraint violation

It was possible that contribs got added which had no raw name. One
example would be a name consisting of whitespace only.

This fix adds a final check for this case.

4 files changed, 86 insertions, 1 deletions

diff --git a/python/fatcat_tools/importers/datacite.py b/python/fatcat_tools/importers/datacite.py
index 81f00876..244984f5 100644
--- a/python/fatcat_tools/importers/datacite.py
+++ b/python/fatcat_tools/importers/datacite.py
@@ -758,6 +758,14 @@ class DataciteImporter(EntityImporter):
                     given_name = clean(given_name)
                 if surname:
                     surname = clean(surname)
+
+                # Perform a final assertion that name does not reduce to zero
+                # (e.g. whitespace only name).
+                if name:
+                    name = name.strip()
+                if not name:
+                    continue
+
                 if raw_affiliation == '':
                     continue
 
diff --git a/python/tests/files/datacite/datacite_doc_31.json b/python/tests/files/datacite/datacite_doc_31.json
new file mode 100644
index 00000000..83af3e4d
--- /dev/null
+++ b/python/tests/files/datacite/datacite_doc_31.json
@@ -0,0 +1,53 @@
+{
+  "id": "10.17912/micropub.biology.000143",
+  "type": "dois",
+  "attributes": {
+    "doi": "10.17912/micropub.biology.000143",
+    "identifiers": null,
+    "creators": [
+      {
+        "raw_name": " ",
+        "givenName": "",
+        "familyName": "",
+        "affiliation": [],
+        "role": "author"
+      }
+    ],
+    "titles": [
+      {
+        "title": "Sample"
+      }
+    ],
+    "publisher": "microPublication Biology",
+    "publicationYear": 2019,
+    "types": {
+      "resourceTypeGeneral": "DataPaper"
+    },
+    "relatedIdentifiers": [],
+    "sizes": [],
+    "formats": [],
+    "version": null,
+    "rightsList": [],
+    "descriptions": [
+      {
+        "description": "Biological liquid-liquid phase separation",
+        "descriptionType": "Abstract"
+      }
+    ],
+    "geoLocations": [],
+    "fundingReferences": [],
+    "url": "https://www.micropublication.org/journals/biology/micropub.biology.000143",
+    "created": "2019-08-19T14:43:08.000Z",
+    "registered": "2019-08-19T14:43:09.000Z",
+    "published": "2019",
+    "updated": "2019-11-09T12:32:02.000Z"
+  },
+  "relationships": {
+    "client": {
+      "data": {
+        "id": "caltech.micropub",
+        "type": "clients"
+      }
+    }
+  }
+}
diff --git a/python/tests/files/datacite/datacite_result_31.json b/python/tests/files/datacite/datacite_result_31.json
new file mode 100644
index 00000000..193104b0
--- /dev/null
+++ b/python/tests/files/datacite/datacite_result_31.json
@@ -0,0 +1,24 @@
+{
+  "abstracts": [
+    {
+      "content": "Biological liquid-liquid phase separation",
+      "lang": "fr",
+      "mimetype": "text/plain"
+    }
+  ],
+  "contribs": [],
+  "ext_ids": {
+    "doi": "10.17912/micropub.biology.000143"
+  },
+  "extra": {
+    "datacite": {
+      "resourceTypeGeneral": "DataPaper"
+    },
+    "container_name": "microPublication Biology"
+  },
+  "refs": [],
+  "release_stage": "published",
+  "release_year": 2019,
+  "publisher": "microPublication Biology",
+  "title": "Sample"
+}
diff --git a/python/tests/import_datacite.py b/python/tests/import_datacite.py
index 15650375..7fdd8230 100644
--- a/python/tests/import_datacite.py
+++ b/python/tests/import_datacite.py
@@ -287,7 +287,7 @@ def test_datacite_conversions(datacite_importer):
     for now.
     """
     datacite_importer.debug = True
-    for i in range(31):
+    for i in range(32):
         src = 'tests/files/datacite/datacite_doc_{0:02d}.json'.format(i)
         dst = 'tests/files/datacite/datacite_result_{0:02d}.json'.format(i)
         with open(src, 'r') as f: