pindexer(dex_ex): index batch swap execution traces

[erwanor]

Describe your changes

This PR adds a dex_ex_batch_swap_traces table that tracks batch swap execution traces.

We do not (yet) track individual position ids for each subtrace hops, so this is left empty.

Notably, the schema contains a price float, but no input/output amount floats for now. This is good to have, but we can add it later and it's not immediately useful to the intended consumer of this data.

The schema:

-- This table tracks individual execution traces for a directed batch swap.
CREATE TABLE IF NOT EXISTS dex_ex_batch_swap_traces (
  -- Primary key
  rowid SERIAL PRIMARY KEY,

  -- The height of the block the batch swap was included in.
  height INTEGER NOT NULL,
  -- The time the batch swap was included in a block.
  time TIMESTAMPTZ NOT NULL,

  -- The amount of asset 1 consumed by the micro execution in raw denom.
  input NUMERIC(39) NOT NULL,
  -- The amount of asset 2 produced by the micro execution in raw denom.
  output NUMERIC(39) NOT NULL,

  -- The amount of asset 1 consumed by the macro execution.
  batch_input NUMERIC(39) NOT NULL,
  -- The amount of asset 2 produced by the macro execution.
  batch_output NUMERIC(39) NOT NULL,
  -- The price (output/input) as a float
  price_float DOUBLE PRECISION NOT NULL,


  -- The directed start asset of the batch swap.
  asset_start BYTEA NOT NULL,
  -- The directed end asset of the batch swap.
  asset_end BYTEA NOT NULL,

  -- Each hop in the list contains an asset id.
  asset_hops BYTEA[] NOT NULL,
  -- Each hop in the list contains an amount.
  amount_hops NUMERIC(39)[] NOT NULL,
  -- Each hop in the list contains a position ID.
  position_id_hops BYTEA[] NOT NULL
);

CREATE INDEX ON dex_ex_batch_swap_traces (time, height);
CREATE INDEX ON dex_ex_batch_swap_traces (asset_start, asset_end);

Checklist before requesting a review

• I have added guiding text to explain how a reviewer should test these changes.

• If this code contains consensus-breaking changes, I have added the "consensus-breaking" label. Otherwise, I declare my belief that there are not consensus-breaking changes, for the following reason:

  pindexer changes

[conorsch]

To review, I've kicked off a reindex of a mainnet db using this binary. As long as that completes without error, I'll call it good enough, and get it in.

[conorsch]

Encountered an error:

Error: error returned from database: syntax error at or near ")"

Caused by:
    syntax error at or near ")"

Investigating...

[erwanor]

Thanks, downloading the mainnet db, it's taking a second

[cronokirby]

Suggested change

	asset_start,
	asset_end,
	price_float,
	price_float,
	asset_start,
	asset_end,

to match the order stuff gets passed in later.

[erwanor]

Thanks!

[cronokirby]

I think there's currently a bug in the data insertion method.

[cronokirby]

Suggested change

	position_id_hops,
	position_id_hops

I think this should fix the SQL error (no trailing commas allowed :()

[conorsch]

Tested again locally on e079188 (the most recent commit), and this time I was able to complete the reindex based on mainnet events. So LGTM, with the changes that have been added.

[conorsch]

Intentionally not merging until @cronokirby has had a chance to look closely, and updates their requested-changes status.

[cronokirby]

All good for me!