Re: Regression in statement locations

On Tue, May 20, 2025 at 02:04:20PM +0000, David Steele wrote:
> I did a careful examination of the remaining diffs (there are quite a few)
> and in every case I consider them to be beneficial, i.e. they make the
> output more targeted and readable.
>
> I did not do a real code review, but I did notice that the test table column
> is called weird_name as in our tests. I would argue that since it is missing
> the quotes and space it is not really all that weird and should maybe get a
> normal name so developers in the future don't wonder what is weird about it.

I have fixed that, as it is not a weird attribute, removed the
unnedeed DO blocks in the tests as pointed out by Anthonin, and moved
one comment as pointed out by Jian.

> Thank you for this improvement and the quick fix!

Yeah, thanks all for pointing out that sometimes my analysis of things
can go off tracks.

The fix has been applied now on HEAD. I've also checked the state of
pgaudit on branch dev-pg18, with the regression getting fixed. Things
look clear now, at least from my side.
--
Michael