qurator-spk/modstool
1
0
Fork 0
mirror of https://github.com/qurator-spk/modstool.git synced 2025-11-04 11:24:14 +01:00
Code Issues Releases Wiki Activity

Merge branch 'master' of https://github.com/qurator-spk/mods4pandas

Browse source
This commit is contained in:
Mike Gerber 2023-11-10 17:57:46 +01:00
commit 5238c0600b
2 changed files with 20 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

8
qurator/mods4pandas/mods4pandas.py
View file

@ -95,7 +95,13 @@ def mods_to_dict(mods, raise_errors=True):
elif tag == '{http://www.loc.gov/mods/v3}recordInfo':
value['recordInfo'] = TagGroup(tag, group).is_singleton().has_no_attributes().descend(raise_errors)
elif tag == '{http://www.loc.gov/mods/v3}recordIdentifier':
value['recordIdentifier'] = TagGroup(tag, group).is_singleton().has_attributes({'source': 'gbv-ppn'}).text()
# By default we assume source="gbv-ppn" mods:recordIdentifiers (= PPNs),
# however, in mods:relatedItems, there may be source="dnb-ppns",
# which we need to distinguish by using a separate field name.
try:
value['recordIdentifier'] = TagGroup(tag, group).is_singleton().has_attributes({'source': 'gbv-ppn'}).text()
except ValueError:
value['recordIdentifier-dnb-ppn'] = TagGroup(tag, group).is_singleton().has_attributes({'source': 'dnb-ppn'}).text()
elif tag == '{http://www.loc.gov/mods/v3}identifier':
for e in group:
if len(e.attrib) != 1:

13
qurator/mods4pandas/tests/test_modstool.py
View file

@ -138,3 +138,16 @@ def test_relatedItem():
""")
assert d['relatedItem-original_recordInfo_recordIdentifier'] == 'PPN167755803'
# mods:relatedItem may also have source="dnb-ppn" recordIdentifiers:
d = dict_fromstring("""
<mods:mods xmlns:mods="http://www.loc.gov/mods/v3">
<mods:relatedItem type="original">
<mods:recordInfo>
<mods:recordIdentifier source="dnb-ppn">1236513355</mods:recordIdentifier>
</mods:recordInfo>
</mods:relatedItem>
</mods:mods>
""")
assert d['relatedItem-original_recordInfo_recordIdentifier-dnb-ppn'] == '1236513355'