autoOffset fails if the target pool's balance of the lowest scored TCO2 is zero

[danceratopz]

Since 0x7f91...604c in block 31487844 (Aug-03-2022 10:11:10 PM +UTC), the lowest scored TCO2 in the NCT pool (TCO2-VCS-1052-2012) has balance zero. This state was previously created on Jul-11-2022 08:11:25 PM +UTC in 0xb402...7573 but subsequently reverted. These transactions, with the subsequent deposit of the TCO2 to the BCT pool, are listed here.

If the array of scoredTCO2s in the NCT storage has not been (manually) updated to reflect the fact that the NCT pool's balance of the lowest scored TCO2 is zero the autoOffset* functions in the OffsetHelper fail:

1. autoRedeem() returns arrays of TCO2 addresses and amounts that were redeemed as returned by autoRedeem2(). These arrays contain entries corresponding to the TCO2 with the lowest score from scoredTCO2s (even though the lowest scored TCO2 is not present in the NCT pool).
2. autoRetire() then tries to retire tokens from the lowest scored TCO2, even though the amount redeemed was zero, and the transaction is reverted.

[pheuberger]

tagging @lazaralex98 to triage

[danceratopz]

tagging @lazaralex98 to triage

Thanks Philipp, please see #32.

[mauricedesaxe]

Indeed redeemAuto2 in the pool contracts returns the scoredTCO2s as is. Meaning, it checks internally which have balance 0 and only redeems the ones that have a balance, but returns all of them instead of a filtered version of the array.

Regardless of that, it seems to me like OffetHelper. autoRedeem() shouldn't fail because of it (?). redeemAuto2 does also return an amounts array, so in the OffsetHelper if a tco2 has 0 balance, it'd attempt to retire 0.

And, indeed if we look at the retire() method of the ToucanCarbonOffsets contract, I think it shouldn't fail if we try to retire 0.

EDIT: Obviously it's not efficient or good if it attempts to retire 0, but I don't think it should cause a failure.

[danceratopz]

Indeed redeemAuto2 in the pool contracts returns the scoredTCO2s as is. Meaning, it checks internally which have balance 0 and only redeems the ones that have a balance, but returns all of them instead of a filtered version of the array.

Regardless of that, it seems to me like OffetHelper. autoRedeem() shouldn't fail because of it (?). redeemAuto2 does also return an amounts array, so in the OffsetHelper if a tco2 has 0 balance, it'd attempt to retire 0.

And, indeed if we look at the retire() method of the ToucanCarbonOffsets contract, I think it shouldn't fail if we try to retire 0.

EDIT: Obviously it's not efficient or good if it attempts to retire 0, but I don't think it should cause a failure.

I looked at this in more detail and tried to explain why the txn gets reverted here: ToucanProtocol/contracts/issues/5.