Query order presignature status

[nlordell]

This PR adds a SignaturePending status to order for pre-sign orders in the OrderMetadata for signalling that a pre-sign order is waiting for a pre-signature in the GPv2 settlement contract.

Please note that my SQL skillz are 👎. So please scrutinize the query modification that was made.

Test Plan

Added new unit tests for the added order status computation. Added an ignored test for the Postgres query for presignature status that can be manually run:

$ cargo test -p orderbook -- --ignored postgres_presignature_status --test-threads 1
    Finished test [unoptimized + debuginfo] target(s) in 0.14s
     Running unittests (target/debug/deps/orderbook-68be6fafd7ab0abc)

running 1 test
test database::orders::tests::postgres_presignature_status ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 59 filtered out; finished in 0.06s

Query Rational

We use a subquery in order to get the latest presignature.

nlordell=# explain SELECT
  (o.signing_scheme = 'presign' AND COALESCE((
    SELECT (NOT p.signed) as unsigned
    FROM presignature_events p
    WHERE o.uid = p.order_uid
    ORDER BY p.block_number DESC, p.log_index DESC
    LIMIT 1
  ), true)) as presignature_pending
FROM orders o
GROUP BY o.uid;
                                        QUERY PLAN                                        
------------------------------------------------------------------------------------------
 HashAggregate  (cost=12.88..260.70 rows=230 width=58)
   Group Key: o.uid
   ->  Seq Scan on orders o  (cost=0.00..12.30 rows=230 width=61)
   SubPlan 1
     ->  Limit  (cost=1.06..1.06 rows=1 width=17)
           ->  Sort  (cost=1.06..1.06 rows=1 width=17)
                 Sort Key: p.block_number DESC, p.log_index DESC
                 ->  Seq Scan on presignature_events p  (cost=0.00..1.05 rows=1 width=17)
                       Filter: (o.uid = order_uid)
(9 rows)

This seems to be more performant than a JOIN LEFT LATERAL.

[codecov-commenter]

Codecov Report

Merging #1025 (df63bd1) into main (3709b51) will decrease coverage by 0.12%.
The diff coverage is 63.86%.

@@            Coverage Diff             @@
##             main    #1025      +/-   ##
==========================================
- Coverage   64.47%   64.35%   -0.13%     
==========================================
  Files         123      123              
  Lines       23350    23471     +121     
==========================================
+ Hits        15056    15105      +49     
- Misses       8294     8366      +72     

[vkgnosis]

Could we rebase this set of branches on main so that it includes the change to SQL query to make reviewing easier?

[fedgiac]

The SQL query looks good to me.

Maybe an extra test could be that an expired order that is pending a presignature returns "expired" (and not "pending").

[nlordell]

Could we rebase this set of branches on main so that it includes the change to SQL query to make reviewing easier?

@vkgnosis should be rebased now.

[nlordell]

@anxolin @W3stside @alfetopito

Just pinging you guys so you are aware of the new order status (for an order that was accepted by the orderbook and is waiting for the on-chain presignature).

[alfetopito]

Thanks for letting us know.
I've created one task for the explorer gnosis/gp-ui#628

I believe Anxo is taking care of cowswap changes.