Fix handling of duplicate column names in parquet reader

[raunaqmorarka]

Description

Parquet files may contain duplicate column names when written by case sensitive tools.
We read the first case insensitive match from the file in this scenario.

Additional context and related issues

Fixes query failures caused by #22538

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:

# Hive
* Fixes query failures due to "Multiple entries with same key" when reading parquet files. ({issue}`23050`)

[findinpath]

Let's iterate a bit more on this, maybe this will give us a better result.

[findinpath]

nit: "Same column name" -> "Namesake"

[ebyhr]

Removed RELEASE-BLOCKER label since version 454 already shipped.

[raunaqmorarka]

I checked Apache Hive behaviour in this case. It always picks first case-insensitive match.
The current PR and Trino behaviour before recent changes matches this behaviour.
I also checked Apache Spark, it throws an error by default

Caused by: org.apache.spark.SparkRuntimeException: Found duplicate field(s) "upper_case_column": [UPPER_CASE_COLUMN, Upper_Case_Column] in case-insensitive mode.
	at org.apache.spark.sql.errors.QueryExecutionErrors$.foundDuplicateFieldInCaseInsensitiveModeError(QueryExecutionErrors.scala:1082)

Setting set spark.sql.hive.convertMetastoreParquet=false; makes it behave the same way as Apache Hive.

[raunaqmorarka]

I further confirmed that this PR is matching the current AWS Athena behaviour.
I also found https://issues.apache.org/jira/browse/HIVE-7554 where case insensitive column matching was implemented a long time ago, but it doesn't explicitly discuss this particular scenario.
With iceberg we continue fail to read a table with such file even after this PR

Caused by: java.lang.IllegalArgumentException: Multiple entries with same key: 1=optional binary upper_case_column (STRING) = 1 and 1=optional binary upper_case_column (STRING) = 1
	at com.google.common.collect.ImmutableMap.conflictException(ImmutableMap.java:382)
	at com.google.common.collect.ImmutableMap.checkNoConflict(ImmutableMap.java:376)
	at com.google.common.collect.RegularImmutableMap.checkNoConflictInKeyBucket(RegularImmutableMap.java:246)
	at com.google.common.collect.RegularImmutableMap.fromEntryArrayCheckingBucketOverflow(RegularImmutableMap.java:133)
	at com.google.common.collect.RegularImmutableMap.fromEntryArray(RegularImmutableMap.java:95)
	at com.google.common.collect.ImmutableMap$Builder.build(ImmutableMap.java:579)
	at com.google.common.collect.ImmutableMap$Builder.buildOrThrow(ImmutableMap.java:607)
	at io.trino.plugin.iceberg.IcebergPageSourceProvider.createParquetIdToFieldMapping(IcebergPageSourceProvider.java:1060)
	at io.trino.plugin.iceberg.IcebergPageSourceProvider.createParquetPageSource(IcebergPageSourceProvider.java:924)

Delta lake seems to explicitly disallow cases where column name differs only by case
https://docs.delta.io/latest/delta-batch.html#schema-validation
So I'm going ahead with landing this to match Trino behaviour with Apache Hive and AWS Athena in this case.