Use valid package name in a test exception assert

[nineinchnick]

I'm seeing a lot of failures on master in a test that was recently introduced in #10613
with the following error:

tests               | 2022-01-31 12:13:37 INFO: FAILURE     /    io.trino.tests.product.hive.TestHiveCompatibility.testSmallDecimalFieldWrittenByOptimizedParquetWriterCannotBeReadByHive (Groups: storage_formats_detailed) took 0.7 seconds
tests               | 2022-01-31 12:13:37 SEVERE: Failure cause:
tests               | java.lang.AssertionError: 
tests               | Expecting message:
tests               |   <"java.io.IOException: parquet.io.ParquetDecodingException: Can not read value at 1 in block 0 in file hdfs://hadoop-master:9000/user/hive/warehouse/parquet_table_small_decimal_created_in_trino/20220131_062836_01082_j2ikr_8c200c2a-4add-4f35-a3ed-b70c4732e54b">
tests               | to match regex:
tests               |   <".* org.apache.parquet.io.ParquetDecodingException: Can not read value at 1 in block 0 in file .*">
tests               | but did not.

[kokosing]

Maybe we should not validate the package at all? Why the package got changed? Maybe different Hadoop distros are using different parquet libraries?

[nineinchnick]

The original PR doesn't have the CI run completed, so I can't tell if it changed, or was just incorrect.

[nineinchnick]

Ah this is the approach in #10847

[nineinchnick]

Duplicate of #10847

[findepi]

@nineinchnick @findinpath at some point Parquet had packages renamed, so different Hive versions have Parquet lib in either org.apache.parquet.io. or parquet.io..