qurator-spk/modstool
1
0
Fork 0
mirror of https://github.com/qurator-spk/modstool.git synced 2025-11-03 19:04:13 +01:00
Code Issues Releases Wiki Activity

🐛 modstool: Handle multiple scriptTerms per language correctly

Browse source
This commit is contained in:
Gerber, Mike 2019-08-29 15:42:13 +02:00
parent 2259d01843
commit 5773b9c9b1
2 changed files with 17 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

15
qurator/modstool/modstool.py
View file

@ -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':

6
qurator/modstool/tests/test_modstool.py
View file

@ -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("""