TrinoHiveCatalog listTables should list all tables

[alexjo2144]

Description

Excluding non-Iceberg tables makes a schema seem empty
when it is not.

Revert of #1354

One situation that is confusing right now is dropping a schema that has only hive tables in it. DROP SCHEMA iceberg.my_schema will fail because the schema is not empty (in that case the exception comes from the HMS, not Trino because Trino thinks the schema is empty). A SHOW TABLES FROM iceberg.my_schema would come back empty.

it'll also be relevant for table redirections #11356. In order to redirect a table, the iceberg catalog needs to know it exists.

This also brings the iceberg connector in line with the Hive connector. Listing shows all available tables in the HMS.

Is this change a fix, improvement, new feature, refactoring, or other?

Fix

Is this a change to the core query engine, a connector, client library, or the SPI interfaces? (be specific)

Iceberg Connector with HMS

How would you describe this change to a non-technical end user or system administrator?

Related issues, pull requests, and links

Documentation

(x) No documentation is needed.
( ) Sufficient documentation is included in this PR.
( ) Documentation PR is available with #prnumber.
( ) Documentation issue #issuenumber is filed, and can be handled later.

Release notes

( ) No release notes entries required.
(x) Release notes entries required with the following suggested text:

# Section
* List all tables when using the Iceberg HMS catalog

[findinpath]

For context, this is the initial commit which added the filtering of Iceberg tables in the listing:

a2ea77b

[alexjo2144]

Even farther back #1354

[findinpath]

What is expected actually from a functional point of view when saying "show tables" in hive/iceberg/delta?
Are we supposed to show ALL the tables irrespective of their underlying format or are we supposed to show only the tables which can be read by the connector (or something else that I haven't thought about)?

[findepi]

Are we supposed to show ALL the tables irrespective of their underlying format or are we supposed to show only the tables which can be read by the connector

The latter is preferred, but not always possible do achieve in a performant manner. The former is acceptable.

In case of redirections, this should be "all tables which can be read by a user" => all tables (reasonably good approximation).

[findinpath]

Please create a follow-up to add eventually the same tests for hive-delta duo.

[alexjo2144]

AC and resolved merge conflicts

[findepi]

Looks like this fixes a (small) bug, right?

[alexjo2144]

Without redirections hiding the non-Iceberg tables could have been nice to have, at the cost of making drop schema sometimes fail in confusing ways

Once we have redirections, hiding these tables would certainly be a bug

[findepi]

AC and resolved merge conflicts

please rebase again, since there are new merge conflicts. sorry

[alexjo2144]

Rebased again, thanks