Multiple refactors (SegmentWithDelta, broker.account, range unpack, order of params)

[PaulRBerg]

Description

Closes #328, #329, #330, and #333.

I have implemented #333 rather grudgingly - as I said in #310, I don't like it that there are now multiple ways to access the times: one via the Range struct (e.g. in getRange) and another via the LockupLinear.Stream struct (e.g. in getStream). But I agree that 20k gas is not insignificant.

Regarding #328 - it would be good if @razgraf could also take a look at the new API implemented by this PR, before we merge it.

Changelog

• chore: improve wording in code comments
• feat: add "SegmentWithDelta" struct
• perf: mark helper functions as "pure"
• refactor: delete "SegmentArrayCountsNotEqual" error
• refactor: rename "broker.addr" to "broker.account"
• refactor: reorder params in "create" functions in pro contract
• refactor: unpack the range structs
• refactor: use "SegmentWithDelta" struct in "createWithDeltas" function
• test: add "DEFAULT_SEGMENTS_WITH_DELTAS" constant
• test: add "getSegmentsWithMilestones" util in "Utils.sol"
• test: add overloads for "fuzzSegmentAmountsAndCalculateCreateAmounts"
• test: delete tests related to "SegmentArrayCountsNotEqual" error
• test: delete unused "sdUint128" and "sdUint40" functions
• test: move "getBlockTimestamp" to "Utils.sol"
• test: rename "vars.amounts" to "vars.createAmounts"

[razgraf]

[andreivladbrg]

The deltas var can be deleted here:
https://github.com/sablierhq/v2-core/blob/98c0d5efd0c924b0bb3cc929d2de5affce6862d6/test/invariant/handlers/LockupProCreateHandler.t.sol#L74

You forgot to update the notice natspec for Range structs, they are not used anymore in the Stream
https://github.com/sablierhq/v2-core/blob/98c0d5efd0c924b0bb3cc929d2de5affce6862d6/src/types/DataTypes.sol#L62
https://github.com/sablierhq/v2-core/blob/98c0d5efd0c924b0bb3cc929d2de5affce6862d6/src/types/DataTypes.sol#L96

It is not about this PR, but I noticed it now. In the BootstrapProtocol the first stream isn't depleted because the end time can't be equal to the start time.

https://github.com/sablierhq/v2-core/blob/98c0d5efd0c924b0bb3cc929d2de5affce6862d6/script/bootstrap/BootstrapProtocol.s.sol#L76-L77

[PaulRBerg]

Thanks for reviewing this, @andreivladbrg. I agree with all of your proposed changes.

In the BootstrapProtocol the first stream isn't depleted because the end time can't be equal to the start time.

Good catch, though this isn't strictly the case. It is possible that Forge will process the stream creation and the stream withdrawal in two different blocks, which would deplete the stream when withdrawMax is called.

Nonetheless, I agree that we should remove that call. The solution is to use cast after the script is executed - or write another script.

[PaulRBerg]

@andreivladbrg I will merge this PR now.

The end-to-end tests have failed, but this is because the fuzzing campaign used an address that is blacklisted by USDC (#317).