[Haskell-cafe] Property checks and PostgreSQL?

[amindfv at mailbox.org]

On Sun, Mar 14, 2021 at 04:06:41AM +0500, Ignat Insarov wrote:
> > Do you know where the majority of the time is spent? I.e. can you just drop the schema without re-creating the cluster? I'd guess it'd be fast but I don't know your setup.
> 
> The first step would be to keep the cluster, yes. It takes a few
> seconds, although we can use a `tmpfs` backed by RAM to cut this down
> to about one second. But initializing the data base from prepared
> `*.sql` files also takes a few seconds.
> 

I don't know your setup but if the problem is not with the creation of your schema it does feel like a waste to tear it down and re-create it from a tmpfs each time.

> > At $WORK we do this: property tests on PSQL data dropping data between tests, but we know which tables are modified so we can just truncate the affected ones.
> 
> So how exactly do you do that? Like, do you write the SQL statements
> needed to reset the data base for every check by hand? And what sort
> of statements do you use? Do you drop the table or delete the rows?
> 

Depends on the test but it's a lot more common to just call "truncate table foo" when necessary.

> It would be ideal if I could somehow detect the tables that were
> touched and reset them automatically. But I cannot think of a simple
> way to do that. I cannot simply erase all the data because some data
> needs to be in place for the application to even start.

It's possible I'm answering an X/Y question[0] right now, but:
  - Can you create a function that populates this essential data and only call that function after truncating?
  - If your data is timestamped you could always delete all "new" rows?

Tom

[0] https://meta.stackexchange.com/questions/66377/what-is-the-xy-problem