Re: Propagate sanity checks of ProcessUtility() to standard_ProcessUtility()? - Mailing list pgsql-hackers

On Thu, Feb 29, 2024 at 3:21 PM Michael Paquier <michael@paquier.xyz> wrote:
>
> Hi all,
>
> It's been brought to me that an extension may finish by breaking the
> assumptions ProcessUtility() relies on when calling
> standard_ProcessUtility(), causing breakages when passing down data to
> cascading utility hooks.
>
> Isn't the state of the arguments given something we should check not
> only in the main entry point ProcessUtility() but also in
> standard_ProcessUtility(), to prevent issues if an extension
> incorrectly manipulates the arguments it needs to pass down to other
> modules that use the utility hook, like using a NULL query string?
>
> See the attached for the idea.

why not just shovel these to standard_ProcessUtility.
so ProcessUtility will looking consistent with (in format)
 * ExecutorStart()
 * ExecutorRun()
 * ExecutorFinish()
 * ExecutorEnd()