qurator-spk/modstool
1
0
Fork 0
mirror of https://github.com/qurator-spk/modstool.git synced 2025-06-25 19:49:54 +02:00
Code Issues Releases Wiki Activity

✨ page_info: Use boolean for indicator variable, str for hrefs

Browse source
This commit is contained in:
Mike Gerber 2025-06-11 20:41:13 +02:00
parent 64ed7298da
commit d685454c52
3 changed files with 15 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

4
src/mods4pandas/lib.py
View file

@ -378,10 +378,12 @@ def convert_db_to_parquet(con, table, index_col, output_file):
df[c] = df[c].astype("Int64")
elif column_type == "float64":
df[c] = df[c].astype("Float64")
elif column_type == "bool":
df[c] = df[c].map({"True": True, "False": False}).astype("boolean")
elif column_type == "set":
# TODO WIP
continue
else:
raise NotImplementedError(f"Column type {column_type} not implemented yet.")
raise NotImplementedError(f"Column {c}: type {column_type} not implemented yet.")
df.to_parquet(output_file)

4
src/mods4pandas/mods4pandas.py
View file

@ -327,6 +327,8 @@ def pages_to_dict(mets, raise_errors=True) -> List[Dict]:
assert file_ is not None
fileGrp_USE = file_.getparent().attrib.get("USE")
file_FLocat_href = (file_.xpath('mets:FLocat/@xlink:href', namespaces=ns) or [None])[0]
if file_FLocat_href is not None:
file_FLocat_href = str(file_FLocat_href)
page_dict[f"fileGrp_{fileGrp_USE}_file_FLocat_href"] = file_FLocat_href
def get_struct_log(*, to_phys):
@ -368,7 +370,7 @@ def pages_to_dict(mets, raise_errors=True) -> List[Dict]:
for struct_div in struct_divs:
type_ = struct_div.attrib.get("TYPE").lower()
assert type_
page_dict[f"structMap-LOGICAL_TYPE_{type_}"] = 1
page_dict[f"structMap-LOGICAL_TYPE_{type_}"] = True
result.append(page_dict)