Trino s3 point queries taking unusually high time on hive connector

[p-chaturvedi]

I am using Trino to access data from a hive catalog from s3. I have a use case where partitioning is not possible and files which I am trying to access are encrypted text files so other connectors like iceberg are not the correct use case. The silver-lining is that I always have a full s3 path with me before querying.

aws s3 cp command takes about 5-6 sec on my bucket while the equivalent Trino query takes about 5-6 mins. The apparent culprit is Trino doing a full ListObject even when full path is provided. If that is true, what's the thought process behind it and can that be optimised?

Are there any tips to optimise Trino queries for my use-case? I want to use Trino even when I have no partitioning because, I can still do inter-catalog joins and have a central place for decryption logic.

example trino query I use:
SELECT content FROM hive.test.data where "$path" = 's3://<bucket>/<folder>/<filename>'