Partially filled orders support in simulation by m-sz · Pull Request #4306 · cowprotocol/services

m-sz · 2026-04-01T16:58:25Z

Description

The order simulation endpoint so far always tried to fill the order with the original 100% fill amount. This leads to reverts if an order was already partially filled.

Changes

Fetches the order's fill amount using an RPC call at the given block and deducts that from the full order amount. This will cause the simulation to always fill the remaining executable amount.

How to test

E2E test simulating a partially filled order.

Contact and Token addresses now are private structs containing Address. Implemented Deref for them and updated usages.

Adds dependency on simulator for solver

These can be added as post-processing step at the callsite. Reverted invalid contracts changes Addressed comments

Co-authored-by: José Duarte <duarte.gmj@gmail.com>

AryanGodara · 2026-04-06T10:33:29Z

crates/orderbook/openapi.yml

          description: >
            Block number to simulate the order at. If not specified, the
            simulation uses the latest block.
+        - in: query


(Spent some time with claude, brushing up on openAPI spec)

The CI is failing because required: false (L819 below) on the blockNumber property inside the SimulationRequest schema isn't valid OpenAPI 3.0.

In Schema Objects, required is only valid as an array on the parent object.

That one should probably fix it 👀

Since you went through all that, you could use the suggest feature in the GitHub to provide a more concrete change

jmg-duarte · 2026-04-06T14:22:15Z

crates/orderbook/openapi.yml

          description: >
            Block number to simulate the order at. If not specified, the
            simulation uses the latest block.
+        - in: query


Since you went through all that, you could use the suggest feature in the GitHub to provide a more concrete change

jmg-duarte · 2026-04-06T14:23:34Z

crates/orderbook/src/api/debug_simulation.rs

+    /// Override for how much of the order has already been filled, expressed
+    /// in the order's fill token (sell token for sell orders, buy token for
+    /// buy orders). When absent, the current on-chain fill state from the
+    /// order metadata is used.
+    #[serde_as(as = "Option<HexOrDecimalU256>")]
+    pub executed_amount: Option<U256>,


I don't like this, are we adding all parameters in the query from here on? Feels like a slippery slope to me

This is a GET endpoint, the POST is used to simulate a custom order and has the respective fields. I do not foresee more query parameters than the simulation block number and executed amounts so I would be for keeping it this way.

What's the reason for making the user provide this value in the first place? We could use an RPC call to automatically figure out the remaining fill amount on the given block. (order.amount - settlementContract.filledAmount(orderUid)).

I will change the implementation to not allow specifying filled amount by the user and issue an RPC call to figure out the executed amount.

crates/orderbook/src/order_simulator.rs

jmg-duarte · 2026-04-06T14:38:20Z

crates/orderbook/src/order_simulator.rs

        &self,
        order: &Order,
        wrappers: Vec<WrapperCall>,
+        executed_amount: Option<U256>,


I'm not sure if we should move this up and split the function in two: with and without executed amount — which would eventually be "with and without overrides"

I'd say that for now it can stay but I'm afraid it will gather more parameters if we're not careful

Let's keep it as-is, and reiterate if the number of overrides grows.

crates/e2e/tests/e2e/order_simulation.rs

…tion

jmg-duarte · 2026-04-07T13:30:56Z

crates/orderbook/src/order_simulator.rs

+        // Clearing prices represent the limit price of the order; both order kinds
+        // produce the same ratio: [buy_amount, sell_amount] for [sell_token,
+        // buy_token].
+        let clearing_prices = vec![order.data.buy_amount, order.data.sell_amount];


if they represent the limit prices, why isn't this called "limit_prices" instead, it's much cleaner and there's no confusion with the whole concept of (uniform) clearing prices

It is called clearing prices everywhere where quoting and orders are involved. I would keep it as-is.

…tion

crates/orderbook/src/orderbook.rs

MartinquaXD · 2026-04-07T14:59:23Z

crates/orderbook/src/orderbook.rs

        let (_, wrappers) = self
            .parse_interactions_and_wrappers(
                order.metadata.full_app_data.as_deref().unwrap_or_default(),
            )
            .map_err(OrderSimulationError::Other)?;


Seems a bit strange to have some of the appdata handled outside of encode_order and some of it inside that function. Is there a reason why it was done this way?

Some of the parts of app data go directly into the "Order" struct, and the wrappers do not. I can change the method to just take the appdata extracted wrappers to keep it in one place.

There is no app data handling inside encode order, the method takes what is provided by the calller (Order struct with filled in interactions) and wrappers that it then encodes.

crates/orderbook/openapi.yml

MartinquaXD · 2026-04-07T15:00:34Z

crates/orderbook/openapi.yml

  "/api/v1/debug/simulation":
    post:
      operationId: debugSimulationPost
      summary: Simulate an arbitrary order.


Is the PR correctly rebased? IIRC there was another PR that introduced simulations for arbitrary orders, right?

It must have been lost during some of the rebases that I have done on the PRs. Let's keep it here then so it is in the openapi spec. Sorry for that.

m-sz and others added 30 commits March 12, 2026 18:02

Move common eth domain types to separate crate

69ca6c1

tombi format

c87657c

Split up acess_list and token_amount

fd4a8d6

fix doc-test

d2687f8

Merge branch 'main' into eth-domain-types

b3253a5

fmt

43510bf

Fully qualify ops

8ba638c

Merge branch 'main' into eth-domain-types

6bd3879

Move autopilot specific, non fundamental types back

c89ea6a

Remove WethAddress in favour of WrappedNativeToken

ef5c1af

Hide internals of {Contact,Token}Address

225f55c

Contact and Token addresses now are private structs containing Address. Implemented Deref for them and updated usages.

Add Deref to WrappedNativeToken

050a35a

Migrate driver to eth-domain-types

607683e

Hide internals of {Contact,Token}Address

8bb68a1

Contact and Token addresses now are private structs containing Address. Implemented Deref for them and updated usages.

Remove driver/src/domain/eth

59cc93a

Add domain/blockchain and flashloan files for driver

f995199

fmt + doc test fix

5e277c5

Factor out code-fetching to separate crate

1d90ff6

Simulator

37bc9f7

De-duplicate settlement_encoding

eb1c81f

Adds dependency on simulator for solver

Merge branch 'main' into simulator-crate

9f8994b

Removed debug tracing calls

43ea033

Small fix

da7aeaf

TOML fmt

612fd75

Remove unused files

a922bb7

Fix copy-and-paste error

07da382

Remove interactions from Query

a050340

These can be added as post-processing step at the callsite. Reverted invalid contracts changes Addressed comments

Merge branch 'main' into simulator-crate

d57160e

Update crates/simulator/src/swap_simulator.rs

3e127b0

Co-authored-by: José Duarte <duarte.gmj@gmail.com>

Add docs

ec6fc0f

m-sz added 2 commits April 3, 2026 17:24

Merge simulation e2e tests where applicable

d213690

clippy + openapi

c416121

AryanGodara reviewed Apr 6, 2026

View reviewed changes

jmg-duarte reviewed Apr 6, 2026

View reviewed changes

m-sz added 9 commits April 7, 2026 12:52

Fix docs, change E2E test request type to the dto

87175a5

Fix OpenAPI spec, add error message coverage in e2e tests

e8f9e1e

Merge branch 'main' into custom-order-simulation

c2f1c38

Merge branch 'custom-order-simulation' into partial-fill-order-simula…

1fe8497

…tion

Use EthUnit in E2E test cases for fill amounts

60b8bd9

Factor out remaining amounts to function, check E2E error for "reverted"

55eb740

Clippy

96fe72b

Merge branch 'custom-order-simulation' into partial-fill-order-simula…

ebdf792

…tion

Merge branch 'main' into custom-order-simulation

91a54fe

jmg-duarte reviewed Apr 7, 2026

View reviewed changes

m-sz added 2 commits April 7, 2026 15:51

Fix native price estimation config in E2E test

5577531

Merge branch 'custom-order-simulation' into partial-fill-order-simula…

d3606c4

…tion

Base automatically changed from custom-order-simulation to main April 7, 2026 14:12

Remove executed amount from the query, calculate it from RPC call

258a30a

MartinquaXD reviewed Apr 7, 2026

View reviewed changes

m-sz and others added 4 commits April 7, 2026 17:04

Merge branch 'main' into partial-fill-order-simulation

e72a85c

Fix partial fill e2e test case

9a5939a

Style fix in orderbook related to Ok()

9d6d931

Remove code that is not needed with new approach anymore

901714b

MartinquaXD approved these changes Apr 8, 2026

View reviewed changes

Merge branch 'main' into partial-fill-order-simulation

65bdb5e

MartinquaXD added this pull request to the merge queue Apr 8, 2026

Merged via the queue into main with commit 7b16a3d Apr 8, 2026
19 checks passed

MartinquaXD deleted the partial-fill-order-simulation branch April 8, 2026 07:24

github-actions bot locked and limited conversation to collaborators Apr 8, 2026

Conversation

m-sz commented Apr 1, 2026 • edited by MartinquaXD Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Changes

How to test

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

m-sz commented Apr 1, 2026 •

edited by MartinquaXD

Loading