feat: eliminate sender flag if account

[zjesko]

closes #10632

Primary Enhancement

Make --sender optional when --account is specified. Foundry should automatically derive the sender address from the account's private key.

Desired Usage

# Simple case - sender automatically derived from account
forge script script/Deploy.s.sol --rpc-url unichain --account deployer --broadcast

Implementation Logic

1. When --account is provided without --sender:

• Load the specified account from the keystore
• Derive the address from the private key
• Use that address as the sender for transactions
• Set msg.sender appropriately in the script context

2. When both --account and --sender are provided:

• Validate that the sender address matches the account's derived address
• Throw a clear error if there's a mismatch
• This maintains backward compatibility for existing scripts

3. When only --sender is provided (current behavior):

• Continue to work as before for raw private key workflows
• Maintain full backward compatibility

[zjesko]

solves #10632

please have a look @zerosnacks @grandizzy @mattsse

[mshakeg]

nice! lgtm

[zjesko]

any comments?
@zerosnacks @grandizzy @mattsse @DaniPopes

[grandizzy]

hm, don't really understand how this is different from existing maybe_load_private_key ?

foundry/crates/script/src/lib.rs

Lines 352 to 359 in 643c7fc

	fn maybe_load_private_key(&self) -> Result<Option<Address>> {
	let maybe_sender = self
	.wallets
	.private_keys()?
	.filter(|pks| pks.len() == 1)
	.map(|pks| pks.first().unwrap().address());
	Ok(maybe_sender)
	}

[mattsse]

I dont fully understand this just yet, because this is currently using privat_keys not the --account arg as @grandizzy pointed out.

we also want a new test for this because all of this isn't trivial

[mattsse]

while this works, I'd prefer if we check if there is a sender value provided in the ScriptArgs instead

or since this is already done tin the maybe_derive_sender_from_account, we can skip the zero check here and use an else if

[mattsse]

shoulnd't those use the --account arg as described in the issue and pr?