From 8eab14ac12b17f4965ec5569853bf885ca6aacd2 Mon Sep 17 00:00:00 2001
From: Martin Czygan <martin.czygan@gmail.com>
Date: Thu, 9 Jan 2020 02:08:08 +0100
Subject: datacite: add 'Unknown' to blacklist

---
 python/fatcat_tools/importers/datacite.py           | 6 +++++-
 python/tests/files/datacite/datacite_result_10.json | 8 +-------
 2 files changed, 6 insertions(+), 8 deletions(-)

diff --git a/python/fatcat_tools/importers/datacite.py b/python/fatcat_tools/importers/datacite.py
index 7878ebfa..ed8b0906 100644
--- a/python/fatcat_tools/importers/datacite.py
+++ b/python/fatcat_tools/importers/datacite.py
@@ -148,8 +148,12 @@ UNKNOWN_MARKERS = set(DATACITE_UNKNOWN_MARKERS).union(set((
     'NN',
     'n.a.',
     '[s.n.]',
+    'Unknown',
 )))
 
+# UNKNOWN_MARKERS_LOWER are lowercase version of UNKNOWN blacklist.
+UNKNOWN_MARKERS_LOWER = set((v.lower() for v in UNKNOWN_MARKERS))
+
 # TODO(martin): merge this with other maps, maybe.
 LICENSE_SLUG_MAP = {
     "//creativecommons.org/licenses/by/2.0/": "CC-BY",
@@ -736,7 +740,7 @@ class DataciteImporter(EntityImporter):
                     continue
                 if name in name_blacklist:
                     continue
-                if name.lower() in UNKNOWN_MARKERS:
+                if name.lower() in UNKNOWN_MARKERS_LOWER:
                     continue
                 # Unpack name, if we have an index form (e.g. 'Razis, Panos A') into 'Panos A razis'.
                 if name:
diff --git a/python/tests/files/datacite/datacite_result_10.json b/python/tests/files/datacite/datacite_result_10.json
index 1bb70be6..ed10fe01 100644
--- a/python/tests/files/datacite/datacite_result_10.json
+++ b/python/tests/files/datacite/datacite_result_10.json
@@ -1,12 +1,6 @@
 {
   "abstracts": [],
-  "contribs": [
-    {
-      "index": 0,
-      "raw_name": "Unknown",
-      "role": "author"
-    }
-  ],
+  "contribs": [],
   "ext_ids": {
     "doi": "10.25549/wpacards-m6171"
   },
-- 
cgit v1.2.3