add function to determine script-type for all inputs in a psbt

[rage-proof]

Currently it's only possible to set the script-type for the inputs in a psbt for wallet owned inputs. For other imported inputs(payjoin) they stay 'unknown'.
This PR adds a function to determine the inputs-types by checking if a script(in scriptpubkey or redeem-script) matches a scripts pattern.

I wanna use this function in this PR: #6804 to validate equality of input-types in the payjoin.

---

general question:
Why is the naming of the script-types different than in most other application/BIPs? If I'm right, in Electrum composed names are in the form: [inner/encapsulated-type]-outer (p2wpkh-p2sh), but in the BIP's the naming is p2sh-p2wpkh. Do I understand it correctly?

[SomberNight]

Why is the naming of the script-types different than in most other application/BIPs? If I'm right, in Electrum composed names are in the form: [inner/encapsulated-type]-outer (p2wpkh-p2sh), but in the BIP's the naming is p2sh-p2wpkh. Do I understand it correctly?

Specifically which BIPs use either naming convention?

[rage-proof]

Specifically which BIPs use either naming convention?

BIP-174(psbt) for example has Test Vectors and their description uses the later form.

[SomberNight]

Well our nomenclature predates BIP174; I think it predates any BIP that has one :)
There isn't really a fundamental reason to prefer one over the other; we use what we use for historical reasons: these are the names we arbitrarily came up with years ago.
Note that us changing them would be a breaking change, as we use these for WIF keys.

[rage-proof]

Thanks for the explanation.

[SomberNight]

Note that in many other places in transaction.py, a script_type of "p2sh" means p2sh-multisig, while a "p2wsh" means p2wsh-multisig.
I guess something will have to be done about that sooner or later anyway... Maybe it's ok to ignore that for now.
However, this is something to keep in mind if you are planning on comparing these script_type strings to determine if certain inputs use similar scripts.

[SomberNight]

Note that in many other places in transaction.py, a script_type of "p2sh" means p2sh-multisig, while a "p2wsh" means p2wsh-multisig.
I guess something will have to be done about that sooner or later anyway...

Ah, actually this might already become a problem here.

It seems to me that is_complete will consider non-ismine inputs that this PR marks as e.g. "p2sh" to be complete, and the code will try to finalise them. Even (non-ismine) inputs that are multisig can get incorrectly finalised because num_sig is not set for them correctly.

electrum/electrum/transaction.py

Lines 1334 to 1352 in d34b8d6

	def is_complete(self) -> bool:
	if self.script_sig is not None and self.witness is not None:
	return True
	if self.is_coinbase_input():
	return True
	if self.script_sig is not None and not self.is_segwit():
	return True
	signatures = list(self.part_sigs.values())
	s = len(signatures)
	# note: The 'script_type' field is currently only set by the wallet,
	# for its own addresses. This means we can only finalize inputs
	# that are related to the wallet.
	# The 'fix' would be adding extra logic that matches on templates,
	# and figures out the script_type from available fields.
	if self.script_type in ('p2pk', 'p2pkh', 'p2wpkh', 'p2wpkh-p2sh'):
	return s >= 1
	if self.script_type in ('p2sh', 'p2wsh', 'p2wsh-p2sh'):
	return s >= self.num_sig
	return False

electrum/electrum/transaction.py

Lines 1354 to 1370 in d34b8d6

	def finalize(self) -> None:
	def clear_fields_when_finalized():
	# BIP-174: "All other data except the UTXO and unknown fields in the
	# input key-value map should be cleared from the PSBT"
	self.part_sigs = {}
	self.sighash = None
	self.bip32_paths = {}
	self.redeem_script = None
	self.witness_script = None
	if self.script_sig is not None and self.witness is not None:
	clear_fields_when_finalized()
	return # already finalized
	if self.is_complete():
	self.script_sig = bfh(Transaction.input_script(self))
	self.witness = bfh(Transaction.serialize_witness(self))
	clear_fields_when_finalized()

electrum/electrum/transaction.py

Line 1126 in d34b8d6

self.num_sig = 0 # type: int # num req sigs for multisig

I think the easiest fix would be to limit the detection added in this PR to only
p2pkh, p2wpkh-p2sh, p2wpkh.

The longer term fix would be

• to split up p2sh et al to p2sh-ms and p2sh-unknown (or "p2sh-generic"? or just "p2sh"?) -- probably also change them to an Enum or similar instead of string literals
• and then adapt the detection to specifically match for multisig templates, and also set num_sig

[rage-proof]

Oh, interesting case. I agree this is a good solution for now.

I think the easiest fix would be to limit the detection added in this PR to only
p2pkh, p2wpkh-p2sh, p2wpkh.

This are the only types defined for Payjoin anyway.

[SomberNight]

Thanks.