442 lines
14 KiB
Python
442 lines
14 KiB
Python
from __future__ import annotations
|
|
|
|
from functools import reduce
|
|
from itertools import chain
|
|
from typing import Any, Callable, Generic, Iterator, Self, TypeVar, overload
|
|
import unicodedata
|
|
|
|
from advent.parser.parser_input import AllowedParserInput, ParserInput, create_parser_input
|
|
|
|
from .result import Result
|
|
|
|
T = TypeVar('T')
|
|
T1 = TypeVar('T1')
|
|
T2 = TypeVar('T2')
|
|
T3 = TypeVar('T3')
|
|
T4 = TypeVar('T4')
|
|
T5 = TypeVar('T5')
|
|
TR = TypeVar('TR')
|
|
|
|
|
|
ParserResult = Iterator[tuple[ParserInput, T]]
|
|
ParserFunc = Callable[[ParserInput], ParserResult[T]]
|
|
|
|
|
|
class P(Generic[T]):
|
|
def __init__(self, func: ParserFunc[T]):
|
|
self.func = func
|
|
|
|
def parse(self, input: AllowedParserInput) -> Result[T]:
|
|
parser_input = create_parser_input(input)
|
|
all_results = self.func(parser_input)
|
|
|
|
try:
|
|
_, result = next(all_results)
|
|
return Result.of(result)
|
|
except StopIteration:
|
|
return Result.fail("No result")
|
|
|
|
def parse_multi(self, input: AllowedParserInput) -> Iterator[T]:
|
|
parser_input = create_parser_input(input)
|
|
all_results = self.func(parser_input)
|
|
|
|
return (v for _, v in all_results)
|
|
|
|
@classmethod
|
|
def pure(cls, value: T) -> P[T]:
|
|
return P(lambda pp: iter([(pp, value)]))
|
|
|
|
@classmethod
|
|
def fail(cls) -> P[Any]:
|
|
return P(lambda _: iter([]))
|
|
|
|
@classmethod
|
|
def _fix(cls, p1: Callable[[P[Any]], P[T]]) -> P[T]:
|
|
""" Not really nice helper function, but it works"""
|
|
return [p._forward(q.func) for p in [P(None)] for q in [p1(p)]][0] # type: ignore
|
|
|
|
def _forward(self, func: ParserFunc[T]) -> Self:
|
|
self.func = func
|
|
return self
|
|
|
|
def bind(self, bind_func: Callable[[T], P[TR]]) -> P[TR]:
|
|
def inner(parserPos: ParserInput) -> ParserResult[TR]:
|
|
return (r for rs in (bind_func(v).func(pp)
|
|
for pp, v in self.func(parserPos)) for r in rs)
|
|
return P(inner)
|
|
|
|
def fmap(self, map_func: Callable[[T], TR]) -> P[TR]:
|
|
def inner(parserPos: ParserInput) -> ParserResult[TR]:
|
|
return ((pp, map_func(v)) for pp, v in self.func(parserPos))
|
|
return P(inner)
|
|
|
|
def safe_fmap(self, map_func: Callable[[T], TR]) -> P[TR]:
|
|
def inner(parserPos: ParserInput) -> ParserResult[TR]:
|
|
for pp, v in self.func(parserPos):
|
|
try:
|
|
yield pp, map_func(v)
|
|
except Exception:
|
|
pass
|
|
|
|
return P(inner)
|
|
|
|
def replace(self, value: TR) -> P[TR]:
|
|
return self.fmap(lambda _: value)
|
|
|
|
def ignore(self) -> P[tuple[()]]:
|
|
return self.fmap(lambda _: ())
|
|
|
|
def apply(self, p2: P[Callable[[T], TR]]) -> P[TR]:
|
|
return self.bind(lambda x: p2.bind(lambda y: P.pure(y(x))))
|
|
|
|
@classmethod
|
|
def first(cls, p1: P[T1], p2: P[Any]) -> P[T1]:
|
|
return p1.bind(lambda v1: p2.fmap(lambda _: v1))
|
|
|
|
@classmethod
|
|
def second(cls, p1: P[Any], p2: P[T2]) -> P[T2]:
|
|
return p1.bind(lambda _: p2)
|
|
|
|
def between(self, pre: P[Any], post: P[Any]) -> P[T]:
|
|
return P.map3(pre, self, post, lambda _1, v, _2: v)
|
|
|
|
def some(self) -> P[list[T]]:
|
|
return P._fix(lambda p: self.bind(
|
|
lambda x: P.either(p, P.pure([])).fmap(lambda ys: [x] + ys)))
|
|
|
|
def some_lazy(self) -> P[list[T]]:
|
|
return P._fix(lambda p: self.bind(
|
|
lambda x: P.either(P.pure([]), p).fmap(lambda ys: [x] + ys)))
|
|
|
|
def many(self) -> P[list[T]]:
|
|
return P.either(self.some(), P.pure([]))
|
|
|
|
def many_lazy(self) -> P[list[T]]:
|
|
return P.either(P.pure([]), self.some_lazy())
|
|
|
|
def satisfies(self, pred: Callable[[T], bool]) -> P[T]:
|
|
return self.bind(lambda v: P.pure(v) if pred(v) else P.fail())
|
|
|
|
def optional(self) -> P[T | None]:
|
|
return P.either(self, P.pure(None))
|
|
|
|
def optional_lazy(self) -> P[T | None]:
|
|
return P.either(P.pure(None), self)
|
|
|
|
@overload
|
|
def times(self, *, exact: int) -> P[list[T]]:
|
|
...
|
|
|
|
@overload
|
|
def times(self, *, min: int) -> P[list[T]]:
|
|
...
|
|
|
|
@overload
|
|
def times(self, *, max: int) -> P[list[T]]:
|
|
...
|
|
|
|
@overload
|
|
def times(self, *, min: int, max: int) -> P[list[T]]:
|
|
...
|
|
|
|
def times(self, *, max: int | None = None, min: int | None = None,
|
|
exact: int | None = None) -> P[list[T]]:
|
|
match (exact, min, max):
|
|
case (int(e), None, None):
|
|
return self.many().satisfies(lambda lst: len(lst) == e)
|
|
case (None, int(mn), None):
|
|
return self.many().satisfies(lambda lst: len(lst) >= mn)
|
|
case (None, None, int(mx)):
|
|
return self.many().satisfies(lambda lst: len(lst) <= mx)
|
|
case (None, int(mn), int(mx)):
|
|
return self.many().satisfies(lambda lst: mn <= len(lst) <= mx)
|
|
case _:
|
|
raise Exception("Illegal combination of parameters")
|
|
|
|
@overload
|
|
def times_lazy(self, *, exact: int) -> P[list[T]]:
|
|
...
|
|
|
|
@overload
|
|
def times_lazy(self, *, min: int) -> P[list[T]]:
|
|
...
|
|
|
|
@overload
|
|
def times_lazy(self, *, max: int) -> P[list[T]]:
|
|
...
|
|
|
|
@overload
|
|
def times_lazy(self, *, min: int, max: int) -> P[list[T]]:
|
|
...
|
|
|
|
def times_lazy(self, *, max: int | None = None, min: int | None = None,
|
|
exact: int | None = None) -> P[list[T]]:
|
|
match (exact, min, max):
|
|
case (int(e), None, None):
|
|
return self.many_lazy().satisfies(lambda lst: len(lst) == e)
|
|
case (None, int(mn), None):
|
|
return self.many_lazy().satisfies(lambda lst: len(lst) >= mn)
|
|
case (None, None, int(mx)):
|
|
return self.many_lazy().satisfies(lambda lst: len(lst) <= mx)
|
|
case (None, int(mn), int(mx)):
|
|
return self.many_lazy().satisfies(lambda lst: mn <= len(lst) <= mx)
|
|
case _:
|
|
raise Exception("Illegal combination of parameters")
|
|
|
|
def sep_by(self, sep: P[Any]) -> P[list[T]]:
|
|
return P.map2(self, P.second(sep, self).many(), lambda f, r: [f] + r)
|
|
|
|
def sep_by_lazy(self, sep: P[Any]) -> P[list[T]]:
|
|
return P.map2(self, P.second(sep, self).many_lazy(), lambda f, r: [f] + r)
|
|
|
|
def no_match(self) -> P[tuple[()]]:
|
|
def inner(parserPos: ParserInput) -> ParserResult[tuple[()]]:
|
|
result = self.func(parserPos)
|
|
try:
|
|
next(result)
|
|
# Silently yields nothing so is an empty Generator
|
|
except StopIteration:
|
|
yield (parserPos, ())
|
|
|
|
return P(inner)
|
|
|
|
@classmethod
|
|
def map2(cls, p1: P[T1], p2: P[T2], func: Callable[[T1, T2], TR]) -> P[TR]:
|
|
return p1.bind(lambda v1: p2.fmap(lambda v2: func(v1, v2)))
|
|
|
|
@classmethod
|
|
def map3(cls, p1: P[T1], p2: P[T2], p3: P[T3], func: Callable[[T1, T2, T3], TR]) -> P[TR]:
|
|
return p1.bind(
|
|
lambda v1: p2.bind(
|
|
lambda v2: p3.fmap(
|
|
lambda v3: func(v1, v2, v3))))
|
|
|
|
@classmethod
|
|
def map4(cls, p1: P[T1], p2: P[T2], p3: P[T3], p4: P[T4],
|
|
func: Callable[[T1, T2, T3, T4], TR]) -> P[TR]:
|
|
return p1.bind(
|
|
lambda v1: p2.bind(
|
|
lambda v2: p3.bind(
|
|
lambda v3: p4.fmap(
|
|
lambda v4: func(v1, v2, v3, v4)))))
|
|
|
|
@classmethod
|
|
def map5(cls, p1: P[T1], p2: P[T2], p3: P[T3], p4: P[T4], p5: P[T5],
|
|
func: Callable[[T1, T2, T3, T4, T5], TR]) -> P[TR]:
|
|
return p1.bind(
|
|
lambda v1: p2.bind(
|
|
lambda v2: p3.bind(
|
|
lambda v3: p4.bind(
|
|
lambda v4: p5.fmap(
|
|
lambda v5: func(v1, v2, v3, v4, v5))))))
|
|
|
|
@classmethod
|
|
@overload
|
|
def seq(cls, p1: P[T1], p2: P[T2], /) -> P[tuple[T1, T2]]:
|
|
...
|
|
|
|
@classmethod
|
|
@overload
|
|
def seq(cls, p1: P[T1], p2: P[T2], p3: P[T3], /) -> P[tuple[T1, T2, T3]]:
|
|
...
|
|
|
|
@classmethod
|
|
@overload
|
|
def seq(cls, p1: P[T1], p2: P[T2], p3: P[T3], p4: P[T4], /) -> P[tuple[T1, T2, T3, T4]]:
|
|
...
|
|
|
|
@classmethod
|
|
@overload
|
|
def seq(cls, p1: P[T1], p2: P[T2], p3: P[T3], p4: P[T4],
|
|
p5: P[T5], /) -> P[tuple[T1, T2, T3, T4, T5]]:
|
|
...
|
|
|
|
@classmethod
|
|
def seq(cls, *ps: P[Any]) -> P[tuple[Any, ...]]:
|
|
return reduce(lambda p, x: x.bind(
|
|
lambda a: p.fmap(lambda b: chain([a], b))),
|
|
list(ps)[::-1], P.pure(iter([]))).fmap(tuple)
|
|
|
|
@classmethod
|
|
@overload
|
|
def sep_seq(cls, p1: P[T1], p2: P[T2], /, *, sep: P[Any]) -> P[tuple[T1, T2]]:
|
|
...
|
|
|
|
@classmethod
|
|
@overload
|
|
def sep_seq(cls, p1: P[T1], p2: P[T2], p3: P[T3], /, *, sep: P[Any]) -> P[tuple[T1, T2, T3]]:
|
|
...
|
|
|
|
@classmethod
|
|
@overload
|
|
def sep_seq(cls, p1: P[T1], p2: P[T2], p3: P[T3], p4: P[T4], /,
|
|
*, sep: P[Any]) -> P[tuple[T1, T2, T3, T4]]:
|
|
...
|
|
|
|
@classmethod
|
|
@overload
|
|
def sep_seq(cls, p1: P[T1], p2: P[T2], p3: P[T3], p4: P[T4],
|
|
p5: P[T5], /, *, sep: P[Any]) -> P[tuple[T1, T2, T3, T4, T5]]:
|
|
...
|
|
|
|
@classmethod
|
|
def sep_seq(cls, *ps: P[Any], sep: P[Any]) -> P[tuple[Any, ...]]:
|
|
first, *rest = list(ps)
|
|
return P.map2(first,
|
|
reduce(lambda p, x: P.second(sep, x.bind(
|
|
lambda a: p.fmap(lambda b: chain([a], b)))),
|
|
rest[::-1], P.pure(iter([]))),
|
|
lambda f, r: (f,) + tuple(r))
|
|
|
|
@classmethod
|
|
def either(cls, p1: P[T1], p2: P[T2], /) -> P[T1 | T2]:
|
|
def inner(parserPos: ParserInput):
|
|
yield from p1.func(parserPos)
|
|
yield from p2.func(parserPos)
|
|
|
|
return P(inner)
|
|
|
|
@classmethod
|
|
@overload
|
|
def choice(cls, p1: P[T1], p2: P[T2], p3: P[T3], /) -> P[T1 | T2 | T3]:
|
|
...
|
|
|
|
@classmethod
|
|
@overload
|
|
def choice(cls, p1: P[T1], p2: P[T2], p3: P[T3], p4: P[T4], /) -> P[T1 | T2 | T3 | T4]:
|
|
...
|
|
|
|
@classmethod
|
|
@overload
|
|
def choice(cls, p1: P[T1], p2: P[T2], p3: P[T3], p4: P[T4],
|
|
p5: P[T5], /) -> P[T1 | T2 | T3 | T4 | T5]:
|
|
...
|
|
|
|
@classmethod
|
|
def choice(cls, *ps: P[Any]) -> P[Any]:
|
|
def inner(parserPos: ParserInput) -> Iterator[Any]:
|
|
for p in ps:
|
|
yield from p.func(parserPos)
|
|
return P(inner)
|
|
|
|
@classmethod
|
|
def choice2(cls, *ps: P[T]) -> P[T]:
|
|
return P.choice(*ps)
|
|
|
|
# Start of String functions
|
|
|
|
@classmethod
|
|
def any_char(cls) -> P[str]:
|
|
def inner(parserPos: ParserInput) -> ParserResult[str]:
|
|
if parserPos.has_data():
|
|
yield parserPos.step()
|
|
return P(inner)
|
|
|
|
@classmethod
|
|
def eof(cls) -> P[tuple[()]]:
|
|
return P.any_char().no_match()
|
|
|
|
@classmethod
|
|
def char_func(cls, cmp: Callable[[str], bool]) -> P[str]:
|
|
return P.any_char().satisfies(cmp)
|
|
|
|
@classmethod
|
|
def char(cls, cmp: str) -> P[str]:
|
|
return P.char_func(lambda c: c == cmp)
|
|
|
|
@classmethod
|
|
def string(cls, cmp: str) -> P[str]:
|
|
return P.seq(*map(P.char, cmp)).replace(cmp)
|
|
|
|
@classmethod
|
|
def tchar(cls, cmp: str) -> P[str]:
|
|
return P.char(cmp).trim()
|
|
|
|
@classmethod
|
|
def tstring(cls, cmp: str) -> P[str]:
|
|
return P.string(cmp).trim()
|
|
|
|
@classmethod
|
|
def one_of(cls, s: str) -> P[str]:
|
|
return P.char_func(lambda c: c in s)
|
|
|
|
@classmethod
|
|
def any_decimal(cls) -> P[str]:
|
|
return P.char_func(lambda c: c.isdecimal())
|
|
|
|
@classmethod
|
|
def is_decimal(cls, num: int) -> P[str]:
|
|
return P.any_decimal().satisfies(lambda c: unicodedata.decimal(c) == num)
|
|
|
|
@classmethod
|
|
def is_not_decimal(cls, num: int) -> P[str]:
|
|
return P.any_decimal().satisfies(lambda c: unicodedata.decimal(c) != num)
|
|
|
|
@classmethod
|
|
def lower(cls) -> P[str]:
|
|
return P.char_func(lambda c: c.islower())
|
|
|
|
@classmethod
|
|
def upper(cls) -> P[str]:
|
|
return P.char_func(lambda c: c.isupper())
|
|
|
|
@classmethod
|
|
def eol(cls) -> P[tuple[()]]:
|
|
return P.char('\n').ignore()
|
|
|
|
@classmethod
|
|
def space(cls) -> P[str]:
|
|
return P.char_func(lambda c: c.isspace())
|
|
|
|
def word(self) -> P[str]:
|
|
return P.first(self.many().fmap(lambda cs: ''.join(str(c) for c in cs)), self.no_match())
|
|
|
|
@classmethod
|
|
def unsigned(cls) -> P[int]:
|
|
return P.either(P.first(P.is_decimal(0), P.any_decimal().no_match()).replace(0),
|
|
P.map2(P.is_not_decimal(0), P.any_decimal().word(),
|
|
lambda f, s: int(f + s)))
|
|
|
|
@classmethod
|
|
def signed(cls) -> P[int]:
|
|
return P.map2(P.one_of('+-').optional(), P.unsigned(),
|
|
lambda sign, num: num if sign != '-' else -num)
|
|
|
|
@classmethod
|
|
def tunsigned(cls) -> P[int]:
|
|
return P.unsigned().trim()
|
|
|
|
@classmethod
|
|
def tsigned(cls) -> P[int]:
|
|
return P.signed().trim()
|
|
|
|
def line(self) -> P[T]:
|
|
return P.first(self, P.eol())
|
|
|
|
def tline(self) -> P[T]:
|
|
return P.first(self.trim(), P.eol())
|
|
|
|
def in_parens(self) -> P[T]:
|
|
return self.between(P.char('('), P.char(')'))
|
|
|
|
def in_angles(self) -> P[T]:
|
|
return self.between(P.char('<'), P.char('>'))
|
|
|
|
def in_brackets(self) -> P[T]:
|
|
return self.between(P.char('['), P.char(']'))
|
|
|
|
def in_curleys(self) -> P[T]:
|
|
return self.between(P.char('{'), P.char('}'))
|
|
|
|
def trim_left(self) -> P[T]:
|
|
return P.second(WHITE_SPACE, self)
|
|
|
|
def trim_right(self) -> P[T]:
|
|
return P.first(self, WHITE_SPACE)
|
|
|
|
def trim(self) -> P[T]:
|
|
return self.between(WHITE_SPACE, WHITE_SPACE)
|
|
|
|
|
|
WHITE_SPACE: P[tuple[()]] = P.space().many().ignore()
|
|
SEP_SPACE: P[tuple[()]] = P.space().some().ignore()
|