Error on reading partitioned Parquets from an Azure Blob Storage

[juergend7lytix]

When reading a hive partitioned Parquet from an Azure Blob Storage, I get a "json exception".

The query is the following,

INSTALL azure;
LOAD azure;

CREATE SECRET azure_secret (
    TYPE AZURE,
    CONNECTION_STRING 'DefaultEndpointsProtocol=https;AccountName=xxx;AccountKey=xxx==;EndpointSuffix=core.windows.net'
  );

SET azure_transport_option_type = 'curl';

SELECT distinct(some_column)
FROM read_parquet('abfss://xxx/xxx/xxx.pq/**');

and the exact error,

Invalid Error:
[json.exception.type_error.302] type must be string, but is null

What did I try?

• Reading the same partitioned parquet from the local disk works flawlessly,
• using the CLI and the Python installation of DuckDB,
• downgrading DuckDB to 1.1.4 and 1.2.0.

Unfortunately, all of this without avail. I also don't know how to narrow the source of the error more down. But as it works when reading the files from the disk, I guess it has something to do with the Azure binding.