Add better support for "state simulation"

[rkalis]

See comments in #98

We can add an OP_CODESEPARATOR between the constructor args and the bytecode. This should make it easier to make "state simulated" contracts. Since then OP_ACTIVEBYTECODE would always be the "base" bytecode, and we could add the "new" state in front of that.

Example syntax (for a contract with name ExampleContract):

bytes34 output = new OutputP2SH(bytes8(1000), hash160(new ExampleContract(tx.locktime + 100, 1000)));

Then the compiled output for new ExampleContract(tx.locktime + 100, 1000) will be something like this:

<VarInt> <1000> OP_CAT <VarInt> <tx.locktime + 100> OP_CAT OP_CAT OP_ACTIVEBYTECODE OP_CAT

We just need to figure out whether we can dynamically determine the VarInts and how to do that.

[rkalis]

After discussion on Telegram I realised that OP_INPUTINDEX OP_UTXOBYTECODE only pushes the locking script (not the redeem script). While OP_ACTIVEBYTECODE pushes the redeem script. So I'll need to include OP_ACTIVEBYTECODE support in the upcoming next release. This makes the construction above slightly more tricky since I'll still have to account for the case with and without constructor args.

[rkalis]

Now that CashTokens are live, I think the best way to do most state simulation is through NFT commitments. So I'll close this issue.