From 5773b9c9b146a39e45d73a36fca6d918a885a29e Mon Sep 17 00:00:00 2001
From: "Gerber, Mike" <mike.gerber@sbb.spk-berlin.de>
Date: Thu, 29 Aug 2019 15:42:13 +0200
Subject: [PATCH] =?UTF-8?q?=F0=9F=90=9B=20modstool:=20Handle=20multiple=20?=
 =?UTF-8?q?scriptTerms=20per=20language=20correctly?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 qurator/modstool/modstool.py            | 15 ++++++++++++---
 qurator/modstool/tests/test_modstool.py |  6 +++++-
 2 files changed, 17 insertions(+), 4 deletions(-)

diff --git a/qurator/modstool/modstool.py b/qurator/modstool/modstool.py
index e98e3c0..eecc5b8 100755
--- a/qurator/modstool/modstool.py
+++ b/qurator/modstool/modstool.py
@@ -298,17 +298,26 @@ def mods_to_dict(mods, raise_errors=True):
             sub_dicts = [mods_to_dict(e) for e in group]
             sub_tags = {k for d in sub_dicts for k in d.keys()}
             for sub_tag in sub_tags:
-                value['language_{}'.format(sub_tag)] = {d.get(sub_tag) for d in sub_dicts if d.get(sub_tag)}
+                s = set()
+                for d in sub_dicts:
+                    v = d.get(sub_tag)
+                    if v:
+                        # There could be multiple scriptTerms in one language element, e.g. Antiqua and Fraktur in a
+                        # German language document.
+                        if isinstance(v, set):
+                            s.update(v)
+                        else:
+                            s.add(v)
+                value['language_{}'.format(sub_tag)] = s
         elif tag == '{http://www.loc.gov/mods/v3}languageTerm':
             value['languageTerm'] = TagGroup(tag, group) \
                 .is_singleton().has_attributes({'authority': 'iso639-2b', 'type': 'code'}) \
                 .text()
         elif tag == '{http://www.loc.gov/mods/v3}scriptTerm':
             value['scriptTerm'] = TagGroup(tag, group) \
-                .is_singleton() \
                 .fix_script_term() \
                 .has_attributes({'authority': 'iso15924', 'type': 'code'}) \
-                .text()
+                .text_set()
         elif tag == '{http://www.loc.gov/mods/v3}relatedItem':
             pass
         elif tag == '{http://www.loc.gov/mods/v3}name':
diff --git a/qurator/modstool/tests/test_modstool.py b/qurator/modstool/tests/test_modstool.py
index 1b990a8..87a0b94 100644
--- a/qurator/modstool/tests/test_modstool.py
+++ b/qurator/modstool/tests/test_modstool.py
@@ -33,9 +33,13 @@ def test_scriptTerm():
             <mods:scriptTerm authority="iso15924" type="code">215</mods:scriptTerm>
             <mods:scriptTerm authority="iso15924" type="code">217</mods:scriptTerm>
         </mods:language>
+        <mods:language>
+            <mods:languageTerm authority="iso639-2b" type="code">lat</mods:languageTerm>
+            <mods:scriptTerm authority="iso15924" type="code">216</mods:scriptTerm>
+        </mods:language>
     </mods:mods>
     """)
-    assert d['language_scriptTerm'] == {'215', '217'}
+    assert d['language_scriptTerm'] == {'215', '216', '217'}
 
 def test_recordInfo():
     d = dict_fromstring("""