Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Improve TWAP order mapping with activeOrderUid #1785

Merged
merged 6 commits into from
Jul 30, 2024
Merged

Conversation

iamacook
Copy link
Member

@iamacook iamacook commented Jul 24, 2024

Summary

In order to cancel a TWAP, the current order UID needs to be referenced. As we calculate each of these, it makes sense to return it as part of the transaction info (as well as at the request of the client).

This adds a new activeOrderUid to the TwapOrderTransactionInfo. On top of this, as we were previously calculating the status based on the initial 10 parts of a TWAP, or the last for those with >10, we now fetch the status from the active part. This fixes the issue whereby large part TWAPs were returning the incorrect status - final parts are not created meaning they would always throw.

Changes

  • Add activeOrderUid to TwapOrderTransactionInfo.
  • Fetch status relative to activeOrderUid.
  • Add appropriate test coverage.

@iamacook iamacook self-assigned this Jul 24, 2024
@iamacook iamacook requested a review from a team as a code owner July 24, 2024 15:01
@coveralls
Copy link

coveralls commented Jul 24, 2024

Pull Request Test Coverage Report for Build 10112332202

Details

  • 2 of 45 (4.44%) changed or added relevant lines in 3 files are covered.
  • 2 unchanged lines in 2 files lost coverage.
  • Overall coverage decreased (-0.2%) to 48.248%

Changes Missing Coverage Covered Lines Changed/Added Lines %
src/routes/transactions/entities/confirmation-view/confirmation-view.entity.ts 1 2 50.0%
src/routes/transactions/entities/swaps/twap-order-info.entity.ts 1 2 50.0%
src/routes/transactions/mappers/common/twap-order.mapper.ts 0 41 0.0%
Files with Coverage Reduction New Missed Lines %
src/routes/transactions/transactions-view.service.ts 1 15.85%
src/routes/transactions/mappers/common/twap-order.mapper.ts 1 15.24%
Totals Coverage Status
Change from base Build 10054294132: -0.2%
Covered Lines: 4328
Relevant Lines: 7215

💛 - Coveralls

@iamacook iamacook marked this pull request as draft July 24, 2024 15:34
@iamacook iamacook changed the title Mark TWAP orders as unknown if getting part orders fails Improve TWAP order mapping with activeOrderUid Jul 25, 2024
@@ -680,6 +786,7 @@ describe('TwapOrderMapper', () => {
});

it('should map the TWAP order if source apps are restricted and a part order fullAppData matches any of the allowed apps', async () => {
configurationService.set('swaps.maxNumberOfParts', 2);
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This test began failing without this.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's weird, I can't replicate the failure, maybe it's a flaky test for some reason. Or maybe the value is set on another test as the configuration is not cleaned between the tests if I'm not wrong. Anyway, I'm good with setting this explictly here.

@iamacook iamacook marked this pull request as ready for review July 25, 2024 07:48
if (args.partOrders) {
const order = args.partOrders.find((order) => order.uid === orderUid);
if (order) {
return order.status;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe I'm misunderstanding this, but if you just return the order status, wouldn't that be wrong for "fulfilled" orders? If the status is fulfilled for one part, it doesn't mean that this applies to the whole TWAP.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The logic regarding status generation is different as of #1785 (comment) which should now solve this.

const order = await this.swapsRepository.getOrder(args.chainId, orderUid);
// If the active order is fulfilled, the next one may not yet exist
// and fetching it would fail so we assume it's open
if (args.activeOrderUid && order.status === OrderStatus.Fulfilled) {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Whenever this order exists we have to return the total twap order status as Open.
Even if the activePart is still open or even if it got cancelled (as it is possible to cancel a single part without cancelling the whole order)

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

As per discussion, we now assume open, fullfilled or cancelled depending on the existince of an active/final order part as of dfea594.


// If the order was created, the TWAP is either open or fulfilled depending
// on whether the order part is active
return args.activeOrderUid ? OrderStatus.Open : OrderStatus.Fulfilled;
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think there is one more edge case we missed when we talked about this:
If the activeOrderUid is the last part.

activeOrderUid === this.gpv2OrderHelper.computeOrderUid({
        chainId: args.chainId,
        owner: args.safeAddress,
        order: args.twapParts.slice(-1)[0], // Last part's order
      });

In that case we could return Fulfilled if the order is already in a final state (fulfilled, cancelled).

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Added in 686505d.

@iamacook iamacook requested a review from schmanu July 25, 2024 11:44
});

// If active order is final one, it's fulfilled as it exists
if (args.activeOrderUid === finalOrderUid) {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we still have to fetch in order to know that it exists.
twapParts and activeOrderUid are both just computed here. So if the last twap would be active this will always be true.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Adjusted in f20a158.


// If the order exists, the TWAP is either open or fulfilled depending
// on whether the order is active
return args.activeOrderUid ? OrderStatus.Open : OrderStatus.Fulfilled;
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we could remove the above check (where I commented) and instead do something like this here:

Suggested change
return args.activeOrderUid ? OrderStatus.Open : OrderStatus.Fulfilled;
return args.activeOrderUid ? args.activeOrderUid === finalOrderUid ? OrderStatus.Fulfilled : OrderStatus.Open : OrderStatus.Fulfilled;

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Adjusted in f20a158.

@@ -841,6 +948,7 @@ describe('TwapOrderMapper', () => {
});

expect(result).toEqual({
activeOrderUid: null,
buyAmount: '51576509680023161648',
buyToken: {
address: buyToken.address,
Copy link
Member

@schmanu schmanu Jul 26, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should we add these (or similar) testcases for the Order Status mapping?

  • Cancelled Order with 3 parts where only the first part exists and the second part is currently active.
  • Cancelled ORder with 3 parts where only the first 2 parts exist and the last part is currently active (This would be a case that should currently fail)
  • Fulfilled Order with 3 parts where the last part is currently active and the repository returns fulfilled for it.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Added in f20a158.

@iamacook iamacook marked this pull request as draft July 26, 2024 09:31
Comment on lines +433 to +440
mockSwapsRepository.getOrder.mockImplementation(
async (_chainId: string, orderUid: string) => {
if (orderUid === part2.uid) {
return Promise.resolve(part2);
}
return Promise.reject(new NotFoundException());
},
);
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I encountered issues in the new test cases that were due to incorrectly calculated order UIDs. As such, I adjusted these mocks to be more specific.

@iamacook iamacook requested a review from schmanu July 26, 2024 13:48
@iamacook iamacook marked this pull request as ready for review July 26, 2024 13:48
Copy link
Member

@schmanu schmanu left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks correct to me :) Thanks for going through those edge cases 🥇

@@ -680,6 +786,7 @@ describe('TwapOrderMapper', () => {
});

it('should map the TWAP order if source apps are restricted and a part order fullAppData matches any of the allowed apps', async () => {
configurationService.set('swaps.maxNumberOfParts', 2);
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's weird, I can't replicate the failure, maybe it's a flaky test for some reason. Or maybe the value is set on another test as the configuration is not cleaned between the tests if I'm not wrong. Anyway, I'm good with setting this explictly here.

@iamacook iamacook merged commit 1519f67 into main Jul 30, 2024
16 checks passed
@iamacook iamacook deleted the large-twap-order-status branch July 30, 2024 15:13
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants