feat: Add Db2 z/OS support

[nj1973]

Description of changes

Adds support for Db2 z/OS and makes the distinction that our existing support was for Db2 LUW.

In the most part Db2 z/OS shares code with LUW. We've added a new Backend because there are some differences, but not a huge amount. The underlying ibm_db/ibm_db_sa drivers are used for both.

Notable differences:

• Sorting by rand() is not supported, we needed a specific random row sample query that selects rand and sorts by the column alias.
• LOB & XML data types are not as flexible on z/OS and needed extra restrictions.
• Catalog SQL for data type detection was slightly different.
• On Db2 z/OS we saw two hidden/auto generated columns being returned by the IBM driver, db2_generated_docid_for_xml and db2_generated_rowid_for_lob. We added code to exclude these from the Ibis tables because we could not influence the SQL used by the driver.
• Db2 z/OS ROWID data type is unsupported and causes an exception. The customer who needs this support did not use ROWIDs explicitly and therefore this was not a blocker.
• Generated SQL statements do not support bind variables. A few functions had to be changed to explicitly include tokens that would otherwise be changed to variables, for example the replacement text used with coalesce.

As a side task I moved some Impala/Hive code out of operations.py into dedicated files.

Unfortunately we have no access to a Db2 z/OS test system. For the most part I've been able to rely on testing code using LUW. I have also benefited from Db2 z/OS testing by the customer who requested this work.

Issues to be closed

Closes #1630

Checklist

• I have followed the CONTRIBUTING Guide.
• I have commented my code, particularly in hard-to-understand areas
• I have updated any relevant documentation to reflect my changes, if applicable
• I have added unit and/or integration tests relevant to my change as needed
• I have already checked locally that all unit tests and linting are passing (use the tests/local_check.sh script)
• I have manually executed end-to-end testing (E2E) with the affected databases/engines

[nj1973]

/gcbrun

[nj1973]

/gcbrun

[sundar-mudupalli-work]

I think it is a bad idea for us to set the default value for this port - this is true for all dbs. The underlying driver sets the default port, so we don't have to. If there is every a change in the default in the future, you need to update this.

[nj1973]

That's not actually a line I changed. Although I did copy that line into the z/OS version of the same method.

I've removed the port default from both implementations.

[sundar-mudupalli-work]

LGTM. Unfortunately, we can't test these change nightly.

[nj1973]

/gcbrun

[nj1973]

/gcbrun

[sundar-mudupalli-work]

/gcbrun