Add LBPool

[gerrrg]

Description

tl;dr: Create a weight shifting LBPool for V3.

Note

This PR replaces and obsoletes #971 by turning it into a PR from the add-lbpool branch from w/in this repo rather than my fork.

What

• Port over the main functionality of LBPool from V2 to V3
• Port over the gradual value change library
• Create a WeightValidation library for improved portability (could use this in standard WP if desired)
• Adhere to events for V2 LBP to improve the ease of integration/compatibility
• Make the pool contract its own hook contract and do beforeJoin hook for onlyOwner LP
• Store info for only a single weight for higher precision/lower gas since weights must sum to FP.ONE
• Add relevant tests
• Modify some inherited tests (BasePoolTest.sol) to be more generalizable/override-able

Why

Modernize a highly popular feature from v2

Type of change

• Bug fix
• New feature
• Breaking change
• Dependency changes
• Code refactor / cleanup
• Optimization: [ ] gas / [ ] bytecode
• Documentation or wording changes
• Other

Checklist:

• The diff is legible and has no extraneous changes
• Complex code has been commented, including external interfaces
• [tbd] Tests have 100% code coverage
• The base branch is either main, or there's a description of how to merge

Issue Resolution

n/a

[gerrrg]

Copying over summary comment from old PR

I've now incorporated all the changes that @EndymionJkb made in #974 and included them in this PR. I believe we can now close that one; I'll summarize the notes/comments from that review here for convenience to note anything known outstanding.

TODOs from PR #974

• Consider moving LBPool to pkg/pool-weighted/lbp or similar
• Consider checking for swapsEnabled in a minimally overridden onSwap rather than in an onBeforeSwap hook to reduce gas costs (using this hook results in a double storage read for balances)
• Port over v2 LBP tests
• Consider changing the fn name for updateWeightsGradually since it describes the action as "update" rather than "schedule".
  • pro: it's more clear 👓
  • con: it breaks previous version convention (do we care? 🤷)
• Consider renaming NormalizedWeightInvariant in WeightedPool (inherited by LBPool) to be more descriptive

Other TODOs

• LBPFactory test(s)
• LBPool test coverage

Misc comments/down-the-line TODOs

• Seedless version
• {,I}TrustedRouterProvider.sol

[gerrrg]

I'm not entirely sure why/how, but Vault Factory changes somehow got pulled into this branch. TODO: get rid of these/resolve the issue that brought them in

[gerrrg]

@EndymionJkb I'm not sure if this is a bug in lcov, or if I'm actually doing something wrong here.

For LBPoolFactory.sol, lcov is claiming that the two SLOC in the constructor (setting _poolVersion and _TRUSTED_ROUTER) are not covered, which appears to be incorrect. The constructor has to be invoked in order for the rest of the tests to run, and other tests (like getPoolVersion and the LBP's trusted router test) would undoubtedly fail if these lines were not run.

Other than that, I think I'm at 100% coverage now -- just a bit puzzled by that result

[EndymionJkb]

@EndymionJkb I'm not sure if this is a bug in lcov, or if I'm actually doing something wrong here.
Other than that, I think I'm at 100% coverage now -- just a bit puzzled by that result

We're aware of issues with coverage. It behaves differently / has holes on both hardhat and forge. Modifiers aren't handled very well, for instance (though that's not what's going on here), which is why they can be run either independently or merged.

It's good to run coverage to make sure you aren't missing anything, but some anomalies are expected, so don't worry about it if it flags something you know actually does run.

[jubeira]

Why not just pause the pool instead?

[EndymionJkb]

Pausing would also prevent the owner from adding liquidity. Often these are deployed with swaps disabled, funded at a later point (to set initial prices), the update is scheduled, and then swaps are enabled to start the sale