Add support for concurrent write on Iceberg transformed column

[pajaks]

Description

Add concurrent write to scenarios like:

-- create test table in iceberg
create table all_defaults_partitioned
with (
  partitioning = array['month(shipdate)']
)
as select * from tpch.sf1000.lineitem

-- first session
update all_defaults_partitioned
set orderkey = 654
where shipdate = date '1995-01-01'

-- second session 
update all_defaults_partitioned
set orderkey = 765
where shipdate = date '1996-12-01'

Additional context and related issues

Release notes

( ) This is not user-visible or is docs only, and no release notes are required.
( ) Release notes are required. Please propose a release note for me.
(x) Release notes are required, with the following suggested text:

## Iceberg
* Add support for concurrent write on Iceberg transformed column

[findinpath]

Why do we intersect enforced with the unenforced predicate?
They are rather unrelated.

[pajaks]

My understanding is that both carriers information about predicate.
The difference is that taken from @raunaqmorarka explanation:

enforced is the part of the predicate which is guaranteed to be satisfied by the connector, so the engine will not apply it on it's side.
unenforced is the part of the predicate which connector cannot guarantee even if it is able to use it to reduce output, so the engine will apply it on the connector output

@raunaqmorarka Correct me if I'm wrong here

[findinpath]

{2:part:date=[ SortedRangeSet[type=date, ranges=1, {[2024-01-01]}] ]}

I'm checking io.trino.plugin.iceberg.TestIcebergLocalConcurrentWritesTest#testConcurrentUpdateWithOverlappingPartitionTransformation (BTW really cool that we have this battery of new concurrency tests)

It's unclear to me why the partition predicate is not an "enforced predicate" while debugging your code.

[pajaks]

Adding such predicates for transformed columns as enforced, would mean pushdowns of those values and connector would need to filter files and rows also during reading from such table as well. It's not supported right now and would require more work.

[pajaks]

They are rather unrelated.

We already join those in IcebergSplitSource

trino/plugin/trino-iceberg/src/main/java/io/trino/plugin/iceberg/IcebergSplitSource.java

Line 259 in 61a79dd

TupleDomain<IcebergColumnHandle> effectivePredicate = TupleDomain.intersect(