select_all swallows exceptions for invalid queries or server errors

[ricardovj]

Is there a way to expose the exceptions and allow callers to capture and handle appropriately? Right now running a select_all on a rogue query quietly fails saying it got 0 results, as opposed to fail because of a logic error:

Simple steps to reproduce from the rails console:

>> # This should cause an exception as the function ccccccc() doesn't exist
>> ClickHouse.connection.select_all("SELECT ccccccc() from cccccccxxxxxx")

SQL (Total: 104MS, CH: 0MS)� SELECT ccccccc() from cccccccxxxxxx;�
Read: 0 rows, 0B. Written: 0 rows, 0B
[]

>> ClickHouse.connection.select_all("SELECT now()")

SQL (Total: 171MS, CH: 1MS)� SELECT now();�
Read: 1 rows, 1.0B. Written: 0 rows, 0B
[{"now()"=>2024-04-04 17:25:04 -0400}]

The exception is there in the faraday.body, but when converting to a ResultSet in the ClickHouse::Response::Factory.response() method that exception is never checked.

[ricardovj]

Any updates on this @shlima?