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

⚡ Make gett_struct_log faster by using find(all) instead of xpath()

Browse source
This commit is contained in:
Gerber, Mike 2023-12-09 10:36:55 +01:00
parent 90c60ebb80
commit 1dac77a2f5
1 changed files with 4 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

6
qurator/mods4pandas/mods4pandas.py
View file

@ -308,12 +308,14 @@ def pages_to_dict(mets, raise_errors=True) -> List[Dict]:
# This is all XLink, there might be a more generic way to traverse the links. However, currently,
# it suffices to do this the old-fashioned way.
sm_links = mets.xpath(f'//mets:structLink/mets:smLink[@xlink:to="{to_phys}"]', namespaces=ns)
sm_links = mets.findall(
f'//mets:structLink/mets:smLink[@xlink:to="{to_phys}"]', ns
)
targets = []
for sm_link in sm_links:
xlink_from = sm_link.attrib.get(f"{{{ns['xlink']}}}from")
targets.extend(mets.xpath(f'//mets:div[@ID="{xlink_from}"]', namespaces=ns))
targets.extend(mets.findall(f'//mets:div[@ID="{xlink_from}"]', ns))
return targets
struct_divs = set(get_struct_log(to_phys=page_dict["ID"]))