Search code examples
prestotrinogoogle-cloud-storage

How to use trino to get prefix path in GCS

I'm using Trino with Hive+GCS connector to access my JSON files inside GCS. If I direct the external table to the path with the JSON it works, like:

create table transactions(
    date DATE,
    transaction1 DOUBLE,
    TransactionType VARCHAR(255)
) WITH (
    external_location = 'gs://bucket/transaction_type/date',
    format = 'PARQUET'
);

But I would like to pass only until transaction_type and it access my files inside dates "folders" recursively. I know that GCS treats everything inside the bucket differently than a folder, I believe thats the problem, but I don't know how to workaround that.

has anyone done something similar?

Solution

  • You should pass hive.recursive-directories as parameter inside the catalog file hive.properties