to_parquet: mismatch between path and projection_storage_location_template

[bucefalog]

Describe the bug

Not sure if this should be a bug or a feature request.
When using function to_parquet with partitions and custom projection_storage_location_template, the template is ignored and files are always written to the default path ( .../partition_column=value/... ).

How to Reproduce

wr.s3.to_parquet(
    df=data,
    path=s3_folder,    # 's3//bucket/prefix/'
    dataset=True,
    database=athena_db, 
    table=table_name, 
    partition_cols= ['my_column'],
    mode='append',
    glue_table_settings={
        'regular_partitions' : False
    },
    athena_partition_projection_settings={ # this is currently not working as the wrangler hardcodes the paths
        'projection_storage_location_template' : f'{s3_folder}${{my_column}}/',
    }
)

Expected behavior

Files should be written to s3 following the provided format.
In the sample code this should be to path:
's3://bucket/prefix/VALUE/HASH.parquet'
as opposed to:
's3://bucket/prefix/my_column=VALUE/HASH.parquet'

Your project

No response

Screenshots

No response

OS

mac

Python version

3.11.4

AWS SDK for pandas version

3.5.2

Additional context

No response

[kukushking]

Hi @bucefalog the template is only used for partition projection, unfortunately it does not impact how awswrangler creates partitions. Only Hive-style partitions are supported at the moment.