MDEV-35369: Add is_deprecated column to system_variables table

[HNOONa-0]

• The Jira issue number for this PR is: MDEV-35369

Description

I added a tinyint(1) column to the system_variables table.
I assume system_variables is a virtual table that does not need any migration scripts but I could be wrong.
The author explicitly stated on JIRA that is_deprecated should be boolean column, however, sql_i_s.h doesn't contain a definition for the boolean datatype, only STiny type is defined so that is what i used. I could add a separate definition and handler for the boolean type if that is desired.
I added the column at the end of the table so that no other changes are necessary inside of fill_sysvars or other areas of code.

here are a couple of queries with is_deprecated filter:

Release Notes

Exposed the IS_DEPRECATED column in INFORMATION_SCHEMA.SYSTEM_VARIABLES, enabling automated checks for deprecated configuration settings.

How can this PR be tested?

Basing the PR against the correct MariaDB version

• This is a new feature or a refactoring, and the PR is based against the main branch.
• This is a bug fix, and the PR is based against the earliest maintained branch in which the bug can be reproduced.

PR quality check

• I checked the CODING_STANDARDS.md file and my PR conforms to this where appropriate.
• For any trivial modifications to the PR, I am ok with the reviewer making the changes themselves.

[CLAassistant]

All committers have signed the CLA.

[HNOONa-0]

rebased on main and updated the failed test

[grooverdan]

I assume system_variables is a virtual table that does not need any migration scripts but I could be wrong.

You are correct. No migration need.

The author explicitly stated on JIRA that is_deprecated should be boolean column, however, sql_i_s.h doesn't contain a definition for the boolean datatype, only STiny type is defined so that is what i used. I could add a separate definition and handler for the boolean type if that is desired.

STiny is ok. Per JIRA, I think the replacement variable would be more useful, however lets get the author to comment there before implementing.

I added the column at the end of the table so that no other changes are necessary inside of fill_sysvars or other areas of code.

Correct.

So in summary:

• good so far
• Lets see if JIRA issue author just wants the replacement variable
• replace_regex shouldn't be there for the tests, most won't need to change
• re-record tests or adjust the SELECT * to include less fields. I think that just a re-recording of the test case and committing the result should be fine in all of these cases.

[gkodinov]

This is a preliminary review. Please keep working with Daniel on the full review.

The only extra remark I have is that you need to get your commit message to be according to CODING_STANDARDS.md

[HNOONa-0]

Hi @grooverdan, I played with cmake flags and updated test results for all tests that query system_variables. I also added deprecated_replacement variable as per discussion on JIRA. I did not add a separate definition for Boolean variable. Turns out, STiny(1) is an alias for the boolean datatype. I previously assumed that a boolean column is its own dedicated type. PTAL!

[grooverdan]

Code is good. Thanks for following and adapting to requested changes and JIRA discussion from @hartmut-mariadb.

Can I get you to amend and push a more verbose in the commit message. Explicitly:

[MDEV-35369 Add deprecation columns to i_s.system_variables

Adds columns:
* IS_DEPRECATED, as  a boolean if the value is deprecated
* DEPRECATED_REPLACEMENT, as how the server replaces this variable

Note this may not be merged immediately, there's a QA/testing process to go through for new features. It will be used.

Many thanks for the contribution.