feat: new syntax to interact with session variables (define/undefine/show variables)

[spena]

Description

This is part of KLIP-38 - Variable Substitution
Variable substitution will have a follow-up PR.

It adds a syntax to interact with session variables:

• DEFINE = '';
• UNDEFINE ;
• (LIST | SHOW) VARIABLES
  ^ The KLIP proposed the use of DEFINE to list values. But I saw the SHOW more appropriate as we have SHOW and LIST syntax for other entities. I will change the KLIP after variable substitution is merged.

The syntax is added on the server-side and CLI-side.

For the server-side, the variable scope is seen per-request. i.e.

Request 1 will return the variables `x` and `y`.
{ 
  statement: "
    DEFINE x = '1';
    DEFINE y = '2';
    SHOW VARIABLES;
  "
}

Request 2 will return the variables `z`.
{ 
  statement: "
    DEFINE z = '3';
    SHOW VARIABLES;
  "
}

For the CLI-side, the variable scope is during the session of the CLI. i.e.

ksql> DEFINE x = '1';
ksql> DEFINE y = '2';
ksql> SHOW VARIABLES;
 Variable Name | Value 
-----------------------
 x             | 1     
 y             | 2     
-----------------------
ksql> DEFINE z = '3';
ksql> SHOW VARIABLES;
 Variable Name | Value 
-----------------------
 x             | 1     
 y             | 2     
 z             | 3     
-----------------------

Testing done

Added unit tests
Verified manually (see above examples)

Reviewer checklist

• Ensure docs are updated if necessary. (eg. if a user visible feature is being added or changed).
• Ensure relevant issues are linked (description should include text like "Fixes #")

[agavra]

LGTM!

[agavra]

is this used anywhere for this PR?

[spena]

No in this PR. I used it initially, but after splitting the PR I left it by accident.

I will leave it here. The next PR will use it anyway, so doesn't hurt.

[agavra]

do you need to actually run through configure etc... or can you mock out the configured statements?

[spena]

I copied most of the code from ListTopicsExecutorTest and ListPropertiesExecutorTest. But you're right, we don't need this engine. I'll change it with a mock.

[agavra]

should we add a test on adding duplicates? we should also check to make sure that case insensitive duplicates are stilled rejected

[spena]

Variables will be overridden (no rejected) if an existing case insensitive variable exists. I'll add a test case that verifies that.

It will allow us to do this:

DEFINE var1 = '1';
DEFINE VAR1 = '2';
DEFINE vAr1 = '3'; // latest update