Hi. When I use the

gcs_stage

method using

target-bigquery

(@alexander_butler variant), it only works if the bucket is a parent directory. i.e

bucket: top_level_folder

. I receive a loader error when trying to specify a sub-directory like

bucket: top_level_folder/sub_folder

I would like to place the staging contents in a particular location and this doesn't seem possible? Looking at the code in

constants.py

line 70, I wonder why this is?

DEFAULT_BUCKET_PATH = (
    "gs://{bucket}/target_bigquery/{dataset}/{table}/extracted_date={date}/{batch_id}.jsonl.gz"
)

I guess this has something to do with the notion of buckets and objects in GCS?

Hi @M! What's the specific error you're seeing? At a glance the code doesn't seem like it'd prevent you from specifying a subdirectory 🤔

Hi @Edgar Ramírez (Arch.dev) Thanks for the reply. It just mentioned the loader had failed. It's ok though I think it's down to GCS definition of a bucket. GCS doesn't really have 'directories'.