Thanks a lot Mark,
I will take a look at this and get back to you if I find anything unclear
---
Hannu
On Tue, Jul 4, 2023 at 10:14 PM Mark Dilger
<mark.dilger@enterprisedb.com> wrote:
>
> Hackers,
>
> Over in [1], Hannu Krosing asked me to create and post several Table Access Methods for testing/example purposes. I
amfairly happy to do so, but each one is large, and should be considered separately for inclusion/rejection in
contrib/,or in src/test/modules as Michael Paquier suggests. As such, I am starting this new email thread for the
firstsuch TAM. I've named it "pile", which is an English synonym of "heap", and which is also four characters in
length,making for easier side-by-side diffs with the heap code. The pile code is a deep copy of the heap code, meaning
thatpile functions do not call heap code, nor run the in-core regression tests, but rather pile's own modified copy of
theheap code, the regression tests, and even the test data. Rather than creating a bare-bones skeleton which needs to
bepopulated with an implementation and regression tests, this patch instead offers a fully fleshed out TAM which can be
pruneddown to something reasonably compact once the user changes it into whatever they want it to be. To reiterate,
thepatch is highly duplicative of in-core files.
>
> Hannu, I'm happy to post something like this three times again, for the named TAMs you request, but could you first
reviewthis patch and maybe try turning it into something else, such as the in memory temp tables, overlay tables, or
pythonbased tables that you mentioned in [1]? Anything that needs to be changed to make similar TAMs suitable for the
communityshould be discussed prior to spamming -hackers with more TAMs. Thanks.
>
>
> [1] https://www.postgresql.org/message-id/CAMT0RQQXtq8tgVPdFb0mk4v%2BcVuGvPWk1Oz9LDr0EgBfrV6e6w%40mail.gmail.com
>
>
>
> —
> Mark Dilger
> EnterpriseDB: http://www.enterprisedb.com
> The Enterprise PostgreSQL Company
>
>
>
