qurator-spk/modstool
1
0
Fork 0
mirror of https://github.com/qurator-spk/modstool.git synced 2025-06-10 04:09:56 +02:00
Code Issues Releases Wiki Activity

⚡ Make get_struct_log faster by using precise predicates

Browse source
This commit is contained in:
Gerber, Mike 2023-12-09 11:40:45 +01:00
parent 448639b05b
commit 912e5d2b4a
1 changed files with 4 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

5
qurator/mods4pandas/mods4pandas.py
View file

@ -268,17 +268,20 @@ def pages_to_dict(mets, raise_errors=True) -> List[Dict]:
# Getting per-page/structure information is a bit different # Getting per-page/structure information is a bit different
structMap_PHYSICAL = mets.find('./mets:structMap[@TYPE="PHYSICAL"]', ns) structMap_PHYSICAL = mets.find('./mets:structMap[@TYPE="PHYSICAL"]', ns)
structMap_LOGICAL = mets.find('./mets:structMap[@TYPE="LOGICAL"]', ns) structMap_LOGICAL = mets.find('./mets:structMap[@TYPE="LOGICAL"]', ns)
fileSec = mets.find('./mets:fileSec', ns)
if structMap_PHYSICAL is None: if structMap_PHYSICAL is None:
raise ValueError("No structMap[@TYPE='PHYSICAL'] found") raise ValueError("No structMap[@TYPE='PHYSICAL'] found")
if structMap_LOGICAL is None: if structMap_LOGICAL is None:
raise ValueError("No structMap[@TYPE='LOGICAL'] found") raise ValueError("No structMap[@TYPE='LOGICAL'] found")
if fileSec is None:
raise ValueError("No fileSec found")
div_physSequence = structMap_PHYSICAL[0] div_physSequence = structMap_PHYSICAL[0]
assert div_physSequence.attrib.get("TYPE") == "physSequence" assert div_physSequence.attrib.get("TYPE") == "physSequence"
def get_mets_file(*, ID): def get_mets_file(*, ID):
if ID: if ID:
file_ = mets.find(f'.//{{{ns["mets"]}}}file[@ID="{ID}"]') file_ = fileSec.find(f'./mets:fileGrp/mets:file[@ID="{ID}"]', ns)
return file_ return file_
def get_mets_div(*, ID): def get_mets_div(*, ID):