Arbitrum support

[rkdud007]

• Note : This PR is depend on Refactor #40 *

Arbitrum Sepolia

MESSAGES_INBOX = "0x7459455BDDC9f5F53af76eFe857B2064b230E4fe"
HEADERS_PROCESSOR = "0x55b87e56F0D12471e1F1a49b1619cCe7B665F2f6"
FACTS_REGISTRY = "0x8Bf5d96bE3B1722114192293F7f25C575B2C70e5"

Sepolia

# -> ARBITRUM_SEPOLIA
L1_TO_ARBITRUM_MESSAGES_SENDER = "0x46A9C7155b2dd191aedc7E14B869E496c5fD0248"

# ParentHash Fetcher
ARBITRUM_PARENT_HASH_FETCHER = "0xEe22f06282174E7cbf7e7b864F5bF8E0b8eDDD62"

[rkdud007]

One comment around integration test

Ideally, wanted to have a full integration test to check from sending message to verify message is intended. Tried approach using foundry's createFork to switch vm in one script. However to test full logic in this way there is little bit of consideration compare to our current contract implementation :

• To make a test in this way, Somehow MessagesInbox this contract should expose a function to verify the messages is correct ( right now we don't have it :/ )
• Also another consideration : Maybe need to wait for the message to be processed in L2

---

So I also give another suggestion if the modification on MessagesInbox side is impossible.

1. Just make test on sending message only, manually check verify message is intended step

• Pro : ye it's simple.. probably in this PR case we would chose it
• Con : It still requires some manual checking and trust

2. Can have l1-l2 messaging testing framework ( NOTE : I m NOT talking abt implementing now, it's just idea -- maybe it can be useful opensource framework project )

• It would be looks like generalized mocktimism (= generalized meaning enable other L2s like Arbitrum/zkSync etc as well )
• Pro : if there is solid full e2e testing framework,,,, especially in our case that needed to deploy & test LOTS of l1<>l2 message logic contracts ,, would be beneficial
• Con : ye it will takes some time,,, and idk depends on opinion, would consider as over-engineering?

[marcellobardus]

Regarding So I also give another suggestion if the modification on MessagesInbox side is impossible.
You could do it as foundry allows you to replace bytecode and you test in fork mode so yes it could work.
Not sure if it is worth it tbh.