FIP-0076 fix: s/SectorDeals/SectorDealIDs in market state to disambiguate naming

[rvagg]

SectorDeals is also used in BatchActivateDeals but has a different form there. This change simply renames the struct to disambiguate it. This also reflects how it's implemented in builtin-actors.

---

While I'm in here, can I ask for clarity on why this is a struct in the first place? It's introduced new complexity now that it's adopted "transparent" encoding in builtin-actors which has required new cbor-gen adjustments to make it work with go-state-types too. Why are we containing it in a struct in the first place and not just doing SectorDealIDs []DealID or omitting SectorDealIDs entirely and saying HAMT[Address]HAMT[SectorNumber][]DealID, which would both result in the same forms as they are currently implemented in both of the proposed builtin-actors and go-state-types but may afford a little bit of simplification in the code by removing an intermediary?

(Edit: here's what such simplification could look like: filecoin-project/builtin-actors#1510)

[jennijuju]

The fip change lgtm to reflect the implementation.

(Cc @Stebalien @anorth for the question asked in PR description

[Stebalien]

I agree with the change in filecoin-project/builtin-actors#1510, honestly. Wrapping in a struct makes some sense for method parameters and return values, but doesn't really add anything here (IMO).

[anorth]

Why are we containing it in a struct in the first place and not just doing SectorDealIDs []DealID or omitting SectorDealIDs entirely and saying HAMT[Address]HAMT[SectorNumber][]DealID, which would both result in the same forms as they are currently implemented in both of the proposed builtin-actors and go-state-types but may afford a little bit of simplification in the code by removing an intermediary?

Sure, they're all fine. I don't understand the problem. Pseudocode in the FIP isn't code code, implementations should represent it however makes sense to them. I would happily also switch to an IPLD-CBOR schema language if that was going to be clearly understandable to readers (it might! I'd love to!).