Mariadb – Does any MariaDB engine have the option to declaratively or configuratively abstract highly-redundant data

I'd suggest you split this large column off into its own table. While you're there take out any other columns which are only used once in a blue moon. (The smaller your row the faster your queries will run.)

I'll take your current table as MyData. Rename it to, say, MyData_Crucial. Let's call the new table MyData_Sometimes. Create this view:

create view MyData
select
    <all the columns>
from MyData_Crucial
inner join MyData_Sometimes
    on <primary key> = <primary key>

Now, assuming your optimizer is at least a little bit clever, you will not need to change your application code at all.

The !! <command> syntax lets you shell out.

The external command is handled synchronously (i.e. the next line of the SQL script will not execute until this command finishes). This is why the documentation has a note in a few places stating:

When you use the Query Editor in SQLCMD Mode, you must be careful not to execute interactive statements. The Query Editor cannot respond to operating system prompts.

and:

...because the Query Editor does not have the ability to respond to operating system prompts, you must be careful not to execute interactive statements.

You can test this rather easily by creating a SQL script of nothing more than:

!!pause

If you run this in SSMS, you will have to cancel the query else it will just keep running. Once it is canceled, you will see the following in the "Messages" tab:

Press any key to continue . . . Query was cancelled by user.

Please see this MSDN page for more information.

However, this introduces an external dependency on your build process, not to mention that SQLCMD mode is not enabled by default in SSMS and cannot be programmatically enabled (though it is possible to change the default setting to always be enabled). Hence you might be better off encapsulating the JSON functionality in a SQLCLR function or stored procedure.