144 lines
3.9 KiB
Python
144 lines
3.9 KiB
Python
"""Recursive-descent parser for arithmetic expressions.
|
|
|
|
AST node types:
|
|
Num(value) -- a numeric literal (int or float)
|
|
BinOp(op, left, right) -- binary operation; op in ('+','-','*','/')
|
|
Unary(op, operand) -- unary operation; op == '-'
|
|
|
|
Grammar (precedence low → high):
|
|
expr := term (('+' | '-') term)*
|
|
term := unary (('*' | '/') unary)*
|
|
unary := '-' unary | primary
|
|
primary := NUMBER | '(' expr ')'
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
from dataclasses import dataclass
|
|
from typing import Union
|
|
|
|
from calc.lexer import Token, tokenize
|
|
|
|
|
|
class ParseError(Exception):
|
|
"""Raised on syntactically invalid input."""
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# AST nodes
|
|
# ---------------------------------------------------------------------------
|
|
|
|
@dataclass
|
|
class Num:
|
|
value: Union[int, float]
|
|
|
|
def __repr__(self) -> str:
|
|
return f"Num({self.value!r})"
|
|
|
|
|
|
@dataclass
|
|
class BinOp:
|
|
op: str
|
|
left: "Node"
|
|
right: "Node"
|
|
|
|
def __repr__(self) -> str:
|
|
return f"BinOp({self.op!r}, {self.left!r}, {self.right!r})"
|
|
|
|
|
|
@dataclass
|
|
class Unary:
|
|
op: str
|
|
operand: "Node"
|
|
|
|
def __repr__(self) -> str:
|
|
return f"Unary({self.op!r}, {self.operand!r})"
|
|
|
|
|
|
Node = Union[Num, BinOp, Unary]
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Parser
|
|
# ---------------------------------------------------------------------------
|
|
|
|
class _Parser:
|
|
def __init__(self, tokens: list[Token]) -> None:
|
|
self._tokens = tokens
|
|
self._pos = 0
|
|
|
|
def _peek(self) -> Token:
|
|
return self._tokens[self._pos]
|
|
|
|
def _consume(self, kind: str | None = None) -> Token:
|
|
tok = self._tokens[self._pos]
|
|
if kind is not None and tok.kind != kind:
|
|
raise ParseError(
|
|
f"Expected {kind}, got {tok.kind!r} ({tok.value!r})"
|
|
)
|
|
self._pos += 1
|
|
return tok
|
|
|
|
def parse(self) -> Node:
|
|
if self._peek().kind == 'EOF':
|
|
raise ParseError("Empty input")
|
|
node = self._expr()
|
|
if self._peek().kind != 'EOF':
|
|
tok = self._peek()
|
|
raise ParseError(
|
|
f"Unexpected token {tok.kind!r} ({tok.value!r}) after expression"
|
|
)
|
|
return node
|
|
|
|
def _expr(self) -> Node:
|
|
node = self._term()
|
|
while self._peek().kind in ('PLUS', 'MINUS'):
|
|
op_tok = self._consume()
|
|
op = op_tok.value # '+' or '-'
|
|
right = self._term()
|
|
node = BinOp(op, node, right)
|
|
return node
|
|
|
|
def _term(self) -> Node:
|
|
node = self._unary()
|
|
while self._peek().kind in ('STAR', 'SLASH'):
|
|
op_tok = self._consume()
|
|
op = op_tok.value # '*' or '/'
|
|
right = self._unary()
|
|
node = BinOp(op, node, right)
|
|
return node
|
|
|
|
def _unary(self) -> Node:
|
|
if self._peek().kind == 'MINUS':
|
|
self._consume('MINUS')
|
|
operand = self._unary()
|
|
return Unary('-', operand)
|
|
return self._primary()
|
|
|
|
def _primary(self) -> Node:
|
|
tok = self._peek()
|
|
if tok.kind == 'NUMBER':
|
|
self._consume('NUMBER')
|
|
return Num(tok.value)
|
|
if tok.kind == 'LPAREN':
|
|
self._consume('LPAREN')
|
|
node = self._expr()
|
|
if self._peek().kind != 'RPAREN':
|
|
raise ParseError(
|
|
f"Expected ')' but got {self._peek().kind!r}"
|
|
)
|
|
self._consume('RPAREN')
|
|
return node
|
|
if tok.kind == 'EOF':
|
|
raise ParseError("Unexpected end of input")
|
|
raise ParseError(
|
|
f"Unexpected token {tok.kind!r} ({tok.value!r})"
|
|
)
|
|
|
|
|
|
def parse(tokens: list[Token]) -> Node:
|
|
"""Parse *tokens* (from lexer.tokenize) into an AST Node.
|
|
|
|
Raises ParseError on malformed input.
|
|
"""
|
|
return _Parser(tokens).parse()
|