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

Make Layout_Page_WIDTH/HEIGHT integer values

This commit is contained in:
Mike Gerber 2025-06-11 19:13:38 +02:00
parent a20c979351
commit 64ed7298da
2 changed files with 7 additions and 2 deletions

View file

@ -47,8 +47,7 @@ EXPECTED_TYPES = {
r"Layout_Page_//alto:String/@WC-.*": ("Float64", None),
r"alto_xmlns": ("object", ["str", "NoneType"]),
# XXX r"Layout_Page_(WIDTH|HEIGHT)": ("Int64", None),
r"Layout_Page_(WIDTH|HEIGHT)": ("object", ["str", "NoneType"]),
r"Layout_Page_(WIDTH|HEIGHT)": ("Int64", None),
}
def expected_types(c):
for r, types in EXPECTED_TYPES.items():

View file

@ -89,6 +89,12 @@ def alto_to_dict(alto, raise_errors=True):
elif localname == 'Page':
value[localname] = {}
value[localname].update(TagGroup(tag, group).is_singleton().attributes())
for attr in ("WIDTH", "HEIGHT"):
if attr in value[localname]:
try:
value[localname][attr] = int(value[localname][attr])
except ValueError:
del value[localname][attr]
value[localname].update(TagGroup(tag, group).subelement_counts())
value[localname].update(TagGroup(tag, group).xpath_statistics("//alto:String/@WC", namespaces))