feat(optimism): Implement new L1 cost function for Fjord

[BrianBland]

• Add new Optimism Fjord hardfork spec
• Implement the Fjord L1 Cost function according to this spec

Note 1: This adds a length-only FastLZ compression implementation under the optimism module. One alternative approach would be to depend on a battle-tested full FastLZ compression implementation, but this would be more computationally expensive and may require the use of C bindings. This could also be a test-only dependency for testing behavioral equivalence.

Note 2: The FastLZ size estimator is called in L1BlockInfo.data_gas, which itself is called within L1BlockInfo.calculate_tx_l1_cost. Callers may often need to perform this pseudo-compression twice per transaction unless we refactor these methods to reuse this computation.

[BrianBland]

cc @danyalprout as this is an adaptation of your parity fuzz test

[mattsse]

@clabby mind taking a look at this?

is the fast_lz function maintained anywhere?

[gakonst]

This shoudl be using alloy cc @DaniPopes

alloy_sol_types::sol! {
    interface FaztLz {
       function fastLz(bytes) external view returns (uint256);
   }
}

And then fastLzCall::abi_encode

https://docs.rs/alloy-sol-types/latest/alloy_sol_types/

[BrianBland]

Thanks I'll apply that refactor to the examples as well if this is the preferred method for interacting with solidity types
Edit: I moved the examples change to a separate PR

[rakita]

left few nits

[rakita]

lgtm

addressed in different PR

[rakita]

Should FJORD be after or before PRAGUE? In secp256r1 precompile PR it was before.

[rakita]

I have put FJORD before PRAGUE as Prague is still not activated