lbry-sdk/lbry/wallet/script.py
2022-04-08 09:57:15 -04:00

593 lines
18 KiB
Python

from typing import List
from itertools import chain
from binascii import hexlify
from collections import namedtuple
from .bcd_data_stream import BCDataStream
from .util import subclass_tuple
# bitcoin opcodes
OP_0 = 0x00
OP_1 = 0x51
OP_16 = 0x60
OP_VERIFY = 0x69
OP_DUP = 0x76
OP_HASH160 = 0xa9
OP_EQUALVERIFY = 0x88
OP_CHECKSIG = 0xac
OP_CHECKMULTISIG = 0xae
OP_CHECKLOCKTIMEVERIFY = 0xb1
OP_EQUAL = 0x87
OP_PUSHDATA1 = 0x4c
OP_PUSHDATA2 = 0x4d
OP_PUSHDATA4 = 0x4e
OP_RETURN = 0x6a
OP_2DROP = 0x6d
OP_DROP = 0x75
# lbry custom opcodes
# checks
OP_PRICECHECK = 0xb0 # checks that the BUY output is >= SELL price
# tx types
OP_CLAIM_NAME = 0xb5
OP_SUPPORT_CLAIM = 0xb6
OP_UPDATE_CLAIM = 0xb7
OP_SELL_CLAIM = 0xb8
OP_BUY_CLAIM = 0xb9
# template matching opcodes (not real opcodes)
# base class for PUSH_DATA related opcodes
# pylint: disable=invalid-name
PUSH_DATA_OP = namedtuple('PUSH_DATA_OP', 'name')
# opcode for variable length strings
# pylint: disable=invalid-name
PUSH_SINGLE = subclass_tuple('PUSH_SINGLE', PUSH_DATA_OP)
# opcode for variable size integers
# pylint: disable=invalid-name
PUSH_INTEGER = subclass_tuple('PUSH_INTEGER', PUSH_DATA_OP)
# opcode for variable number of variable length strings
# pylint: disable=invalid-name
PUSH_MANY = subclass_tuple('PUSH_MANY', PUSH_DATA_OP)
# opcode with embedded subscript parsing
# pylint: disable=invalid-name
PUSH_SUBSCRIPT = namedtuple('PUSH_SUBSCRIPT', 'name template')
def is_push_data_opcode(opcode):
return isinstance(opcode, (PUSH_DATA_OP, PUSH_SUBSCRIPT))
def is_push_data_token(token):
return 1 <= token <= OP_PUSHDATA4
def push_data(data):
size = len(data)
if size < OP_PUSHDATA1:
yield BCDataStream.uint8.pack(size)
elif size <= 0xFF:
yield BCDataStream.uint8.pack(OP_PUSHDATA1)
yield BCDataStream.uint8.pack(size)
elif size <= 0xFFFF:
yield BCDataStream.uint8.pack(OP_PUSHDATA2)
yield BCDataStream.uint16.pack(size)
else:
yield BCDataStream.uint8.pack(OP_PUSHDATA4)
yield BCDataStream.uint32.pack(size)
yield bytes(data)
def read_data(token, stream):
if token < OP_PUSHDATA1:
return stream.read(token)
if token == OP_PUSHDATA1:
return stream.read(stream.read_uint8())
if token == OP_PUSHDATA2:
return stream.read(stream.read_uint16())
return stream.read(stream.read_uint32())
# opcode for OP_1 - OP_16
# pylint: disable=invalid-name
SMALL_INTEGER = namedtuple('SMALL_INTEGER', 'name')
def is_small_integer(token):
return OP_1 <= token <= OP_16
def push_small_integer(num):
assert 1 <= num <= 16
yield BCDataStream.uint8.pack(OP_1 + (num - 1))
def read_small_integer(token):
return (token - OP_1) + 1
class Token(namedtuple('Token', 'value')):
__slots__ = ()
def __repr__(self):
name = None
for var_name, var_value in globals().items():
if var_name.startswith('OP_') and var_value == self.value:
name = var_name
break
return name or self.value
class DataToken(Token):
__slots__ = ()
def __repr__(self):
return f'"{hexlify(self.value)}"'
class SmallIntegerToken(Token):
__slots__ = ()
def __repr__(self):
return f'SmallIntegerToken({self.value})'
def token_producer(source):
token = source.read_uint8()
while token is not None:
if is_push_data_token(token):
yield DataToken(read_data(token, source))
elif is_small_integer(token):
yield SmallIntegerToken(read_small_integer(token))
else:
yield Token(token)
token = source.read_uint8()
def tokenize(source):
return list(token_producer(source))
class ScriptError(Exception):
""" General script handling error. """
class ParseError(ScriptError):
""" Script parsing error. """
class Parser:
def __init__(self, opcodes, tokens):
self.opcodes = opcodes
self.tokens = tokens
self.values = {}
self.token_index = 0
self.opcode_index = 0
def parse(self):
while self.token_index < len(self.tokens) and self.opcode_index < len(self.opcodes):
token = self.tokens[self.token_index]
opcode = self.opcodes[self.opcode_index]
if token.value == 0 and isinstance(opcode, PUSH_SINGLE):
token = DataToken(b'')
if isinstance(token, DataToken):
if isinstance(opcode, (PUSH_SINGLE, PUSH_INTEGER, PUSH_SUBSCRIPT)):
self.push_single(opcode, token.value)
elif isinstance(opcode, PUSH_MANY):
self.consume_many_non_greedy()
else:
raise ParseError(f"DataToken found but opcode was '{opcode}'.")
elif isinstance(token, SmallIntegerToken):
if isinstance(opcode, SMALL_INTEGER):
self.values[opcode.name] = token.value
else:
raise ParseError(f"SmallIntegerToken found but opcode was '{opcode}'.")
elif token.value == opcode:
pass
else:
raise ParseError(f"Token is '{token.value}' and opcode is '{opcode}'.")
self.token_index += 1
self.opcode_index += 1
if self.token_index < len(self.tokens):
raise ParseError("Parse completed without all tokens being consumed.")
if self.opcode_index < len(self.opcodes):
raise ParseError("Parse completed without all opcodes being consumed.")
return self
def consume_many_non_greedy(self):
""" Allows PUSH_MANY to consume data without being greedy
in cases when one or more PUSH_SINGLEs follow a PUSH_MANY. This will
prioritize giving all PUSH_SINGLEs some data and only after that
subsume the rest into PUSH_MANY.
"""
token_values = []
while self.token_index < len(self.tokens):
token = self.tokens[self.token_index]
if not isinstance(token, DataToken):
self.token_index -= 1
break
token_values.append(token.value)
self.token_index += 1
push_opcodes = []
push_many_count = 0
while self.opcode_index < len(self.opcodes):
opcode = self.opcodes[self.opcode_index]
if not is_push_data_opcode(opcode):
self.opcode_index -= 1
break
if isinstance(opcode, PUSH_MANY):
push_many_count += 1
push_opcodes.append(opcode)
self.opcode_index += 1
if push_many_count > 1:
raise ParseError(
"Cannot have more than one consecutive PUSH_MANY, as there is no way to tell which"
" token value should go into which PUSH_MANY."
)
if len(push_opcodes) > len(token_values):
raise ParseError(
"Not enough token values to match all of the PUSH_MANY and PUSH_SINGLE opcodes."
)
many_opcode = push_opcodes.pop(0)
# consume data into PUSH_SINGLE opcodes, working backwards
for opcode in reversed(push_opcodes):
self.push_single(opcode, token_values.pop())
# finally PUSH_MANY gets everything that's left
self.values[many_opcode.name] = token_values
def push_single(self, opcode, value):
if isinstance(opcode, PUSH_SINGLE):
self.values[opcode.name] = value
elif isinstance(opcode, PUSH_INTEGER):
self.values[opcode.name] = int.from_bytes(value, 'little')
elif isinstance(opcode, PUSH_SUBSCRIPT):
self.values[opcode.name] = Script.from_source_with_template(value, opcode.template)
else:
raise ParseError(f"Not a push single or subscript: {opcode}")
class Template:
__slots__ = 'name', 'opcodes'
def __init__(self, name, opcodes):
self.name = name
self.opcodes = opcodes
def parse(self, tokens):
return Parser(self.opcodes, tokens).parse().values if self.opcodes else {}
def generate(self, values):
source = BCDataStream()
for opcode in self.opcodes:
if isinstance(opcode, PUSH_SINGLE):
data = values[opcode.name]
source.write_many(push_data(data))
elif isinstance(opcode, PUSH_INTEGER):
data = values[opcode.name]
source.write_many(push_data(
data.to_bytes((data.bit_length() + 8) // 8, byteorder='little', signed=True)
))
elif isinstance(opcode, PUSH_SUBSCRIPT):
data = values[opcode.name]
source.write_many(push_data(data.source))
elif isinstance(opcode, PUSH_MANY):
for data in values[opcode.name]:
source.write_many(push_data(data))
elif isinstance(opcode, SMALL_INTEGER):
data = values[opcode.name]
source.write_many(push_small_integer(data))
else:
source.write_uint8(opcode)
return source.get_bytes()
class Script:
__slots__ = 'source', '_template', '_values', '_template_hint'
templates: List[Template] = []
NO_SCRIPT = Template('no_script', None) # special case
def __init__(self, source=None, template=None, values=None, template_hint=None):
self.source = source
self._template = template
self._values = values
self._template_hint = template_hint
if source is None and template and values:
self.generate()
@property
def template(self):
if self._template is None:
self.parse(self._template_hint)
return self._template
@property
def values(self):
if self._values is None:
self.parse(self._template_hint)
return self._values
@property
def tokens(self):
return tokenize(BCDataStream(self.source))
@classmethod
def from_source_with_template(cls, source, template):
return cls(source, template_hint=template)
def parse(self, template_hint=None):
tokens = self.tokens
if not tokens and not template_hint:
template_hint = self.NO_SCRIPT
for template in chain((template_hint,), self.templates):
if not template:
continue
try:
self._values = template.parse(tokens)
self._template = template
return
except ParseError:
continue
raise ValueError(f'No matching templates for source: {hexlify(self.source)}')
def generate(self):
self.source = self.template.generate(self._values)
class InputScript(Script):
__slots__ = ()
REDEEM_PUBKEY = Template('pubkey', (
PUSH_SINGLE('signature'),
))
REDEEM_PUBKEY_HASH = Template('pubkey_hash', (
PUSH_SINGLE('signature'), PUSH_SINGLE('pubkey')
))
MULTI_SIG_SCRIPT = Template('multi_sig', (
SMALL_INTEGER('signatures_count'), PUSH_MANY('pubkeys'), SMALL_INTEGER('pubkeys_count'),
OP_CHECKMULTISIG
))
REDEEM_SCRIPT_HASH_MULTI_SIG = Template('script_hash+mult_sig', (
PUSH_SINGLE('signature'), PUSH_SINGLE('pubkey'), PUSH_SUBSCRIPT('script', MULTI_SIG_SCRIPT)
))
TIME_LOCK_SCRIPT = Template('timelock', (
PUSH_INTEGER('height'), OP_CHECKLOCKTIMEVERIFY, OP_DROP,
# rest is identical to OutputScript.PAY_PUBKEY_HASH:
OP_DUP, OP_HASH160, PUSH_SINGLE('pubkey_hash'), OP_EQUALVERIFY, OP_CHECKSIG
))
REDEEM_SCRIPT_HASH_TIME_LOCK = Template('script_hash+timelock', (
PUSH_SINGLE('signature'), PUSH_SINGLE('pubkey'), PUSH_SUBSCRIPT('script', TIME_LOCK_SCRIPT)
))
templates = [
REDEEM_PUBKEY,
REDEEM_PUBKEY_HASH,
REDEEM_SCRIPT_HASH_TIME_LOCK,
REDEEM_SCRIPT_HASH_MULTI_SIG,
]
@classmethod
def redeem_pubkey_hash(cls, signature, pubkey):
return cls(template=cls.REDEEM_PUBKEY_HASH, values={
'signature': signature,
'pubkey': pubkey
})
@classmethod
def redeem_mult_sig_script_hash(cls, signatures, pubkeys):
return cls(template=cls.REDEEM_SCRIPT_HASH_MULTI_SIG, values={
'signatures': signatures,
'script': cls(template=cls.MULTI_SIG_SCRIPT, values={
'signatures_count': len(signatures),
'pubkeys': pubkeys,
'pubkeys_count': len(pubkeys)
})
})
@classmethod
def redeem_time_lock_script_hash(cls, signature, pubkey, height=None, pubkey_hash=None, script_source=None):
if height and pubkey_hash:
script = cls(template=cls.TIME_LOCK_SCRIPT, values={
'height': height,
'pubkey_hash': pubkey_hash
})
elif script_source:
script = cls(source=script_source, template=cls.TIME_LOCK_SCRIPT)
script.parse(script.template)
else:
raise ValueError("script_source or both height and pubkey_hash are required.")
return cls(template=cls.REDEEM_SCRIPT_HASH_TIME_LOCK, values={
'signature': signature,
'pubkey': pubkey,
'script': script
})
@property
def is_script_hash(self):
return self.template.name.startswith('script_hash+')
class OutputScript(Script):
__slots__ = ()
# output / payment script templates (aka scriptPubKey)
PAY_PUBKEY_FULL = Template('pay_pubkey_full', (
PUSH_SINGLE('pubkey'), OP_CHECKSIG
))
PAY_PUBKEY_HASH = Template('pay_pubkey_hash', (
OP_DUP, OP_HASH160, PUSH_SINGLE('pubkey_hash'), OP_EQUALVERIFY, OP_CHECKSIG
))
PAY_SCRIPT_HASH = Template('pay_script_hash', (
OP_HASH160, PUSH_SINGLE('script_hash'), OP_EQUAL
))
PAY_SEGWIT = Template('pay_script_hash+segwit', (
OP_0, PUSH_SINGLE('script_hash')
))
RETURN_DATA = Template('return_data', (
OP_RETURN, PUSH_SINGLE('data')
))
CLAIM_NAME_OPCODES = (
OP_CLAIM_NAME, PUSH_SINGLE('claim_name'), PUSH_SINGLE('claim'),
OP_2DROP, OP_DROP
)
CLAIM_NAME_PUBKEY = Template('claim_name+pay_pubkey_hash', (
CLAIM_NAME_OPCODES + PAY_PUBKEY_HASH.opcodes
))
CLAIM_NAME_SCRIPT = Template('claim_name+pay_script_hash', (
CLAIM_NAME_OPCODES + PAY_SCRIPT_HASH.opcodes
))
SUPPORT_CLAIM_OPCODES = (
OP_SUPPORT_CLAIM, PUSH_SINGLE('claim_name'), PUSH_SINGLE('claim_id'),
OP_2DROP, OP_DROP
)
SUPPORT_CLAIM_PUBKEY = Template('support_claim+pay_pubkey_hash', (
SUPPORT_CLAIM_OPCODES + PAY_PUBKEY_HASH.opcodes
))
SUPPORT_CLAIM_SCRIPT = Template('support_claim+pay_script_hash', (
SUPPORT_CLAIM_OPCODES + PAY_SCRIPT_HASH.opcodes
))
SUPPORT_CLAIM_DATA_OPCODES = (
OP_SUPPORT_CLAIM, PUSH_SINGLE('claim_name'), PUSH_SINGLE('claim_id'), PUSH_SINGLE('support'),
OP_2DROP, OP_2DROP
)
SUPPORT_CLAIM_DATA_PUBKEY = Template('support_claim+data+pay_pubkey_hash', (
SUPPORT_CLAIM_DATA_OPCODES + PAY_PUBKEY_HASH.opcodes
))
SUPPORT_CLAIM_DATA_SCRIPT = Template('support_claim+data+pay_script_hash', (
SUPPORT_CLAIM_DATA_OPCODES + PAY_SCRIPT_HASH.opcodes
))
UPDATE_CLAIM_OPCODES = (
OP_UPDATE_CLAIM, PUSH_SINGLE('claim_name'), PUSH_SINGLE('claim_id'), PUSH_SINGLE('claim'),
OP_2DROP, OP_2DROP
)
UPDATE_CLAIM_PUBKEY = Template('update_claim+pay_pubkey_hash', (
UPDATE_CLAIM_OPCODES + PAY_PUBKEY_HASH.opcodes
))
UPDATE_CLAIM_SCRIPT = Template('update_claim+pay_script_hash', (
UPDATE_CLAIM_OPCODES + PAY_SCRIPT_HASH.opcodes
))
templates = [
PAY_PUBKEY_FULL,
PAY_PUBKEY_HASH,
PAY_SCRIPT_HASH,
PAY_SEGWIT,
RETURN_DATA,
CLAIM_NAME_PUBKEY,
CLAIM_NAME_SCRIPT,
SUPPORT_CLAIM_PUBKEY,
SUPPORT_CLAIM_SCRIPT,
SUPPORT_CLAIM_DATA_PUBKEY,
SUPPORT_CLAIM_DATA_SCRIPT,
UPDATE_CLAIM_PUBKEY,
UPDATE_CLAIM_SCRIPT,
]
@classmethod
def pay_pubkey_hash(cls, pubkey_hash):
return cls(template=cls.PAY_PUBKEY_HASH, values={
'pubkey_hash': pubkey_hash
})
@classmethod
def pay_script_hash(cls, script_hash):
return cls(template=cls.PAY_SCRIPT_HASH, values={
'script_hash': script_hash
})
@classmethod
def return_data(cls, data):
return cls(template=cls.RETURN_DATA, values={
'data': data
})
@property
def is_pay_pubkey(self):
return self.template.name.endswith('pay_pubkey_full')
@classmethod
def pay_claim_name_pubkey_hash(cls, claim_name, claim, pubkey_hash):
return cls(template=cls.CLAIM_NAME_PUBKEY, values={
'claim_name': claim_name,
'claim': claim,
'pubkey_hash': pubkey_hash
})
@classmethod
def pay_update_claim_pubkey_hash(cls, claim_name, claim_id, claim, pubkey_hash):
return cls(template=cls.UPDATE_CLAIM_PUBKEY, values={
'claim_name': claim_name,
'claim_id': claim_id,
'claim': claim,
'pubkey_hash': pubkey_hash
})
@classmethod
def pay_support_pubkey_hash(cls, claim_name: bytes, claim_id: bytes, pubkey_hash: bytes):
return cls(template=cls.SUPPORT_CLAIM_PUBKEY, values={
'claim_name': claim_name,
'claim_id': claim_id,
'pubkey_hash': pubkey_hash
})
@classmethod
def pay_support_data_pubkey_hash(
cls, claim_name: bytes, claim_id: bytes, support, pubkey_hash: bytes):
return cls(template=cls.SUPPORT_CLAIM_DATA_PUBKEY, values={
'claim_name': claim_name,
'claim_id': claim_id,
'support': support,
'pubkey_hash': pubkey_hash
})
@property
def is_pay_pubkey_hash(self):
return self.template.name.endswith('pay_pubkey_hash')
@property
def is_pay_script_hash(self):
return self.template.name.endswith('pay_script_hash')
@property
def is_return_data(self):
return self.template.name.endswith('return_data')
@property
def is_claim_name(self):
return self.template.name.startswith('claim_name+')
@property
def is_update_claim(self):
return self.template.name.startswith('update_claim+')
@property
def is_support_claim(self):
return self.template.name.startswith('support_claim+')
@property
def is_support_claim_data(self):
return self.template.name.startswith('support_claim+data+')
@property
def is_claim_involved(self):
return any((self.is_claim_name, self.is_support_claim, self.is_update_claim))