Implement Uniswap V2-style fees-in-reserves: the full swap_amount_in is
deposited into the reserve (growing k = reserve_a * reserve_b), while
only the fee-adjusted effective_amount_in is used to compute the output
amount. This means LPs earn fees proportionally on every removal via
k-growth rather than through a separate vault surplus.
- swap_logic: add fee_bps parameter; compute effective_amount_in for
output formula only; return full swap_amount_in as the reserve deposit
- Fix all integration test fixture values to match fees-in-reserves math
- Remove dead-code vault_a/b_init_zero helpers from unit tests
Remove the `active: bool` field from `PoolDefinition` and replace it with an
implicit invariant: a pool is considered active when
`liquidity_pool_supply >= MINIMUM_LIQUIDITY`.
BREAKING CHANGE: `PoolDefinition` Borsh serialization format has changed.
Existing on-chain pool accounts encoded with the `active` field are
incompatible with this version.
Closes#25
- accept a supported fee tier in pool creation
- store fee tiers in AMM pool state and validate them
- update AMM tests and IDL for the new pool creation argument
Renames the Swap instruction and its guest handler to SwapExactInput to
distinguish it from the newly added SwapExactOutput, and to make the
intent of each variant explicit at the call site.
BREAKING CHANGE: the Swap instruction variant and swap() function are
renamed to SwapExactInput and swap_exact_input(). Callers must update
instruction construction and any IDL-generated bindings.
Adds SwapExactOutput to the AMM, allowing callers to specify the exact
desired output amount while the protocol computes the required input
(ceiling division to prevent rounding in the protocol's favour).
The swap-exact-output success tests now use a dedicated small-pool
fixture (reserve_a=1_000, reserve_b=500) rather than the shared
pool_definition_init, which had its reserves bumped to 5_000/2_500 in a
later commit to satisfy the MINIMUM_LIQUIDITY invariant introduced for
new_definition. Using a dedicated fixture keeps each test self-contained
and avoids hardcoded expected values silently breaking when shared
baselines change.
