from __future__ import annotations from dataclasses import dataclass from typing import Union class LexError(Exception): pass @dataclass class Token: kind: str value: Union[int, float, str, None] def tokenize(src: str) -> list[Token]: tokens: list[Token] = [] i = 0 n = len(src) while i < n: ch = src[i] if ch in ' \t\r\n': i += 1 continue if ch.isdigit() or ch == '.': j = i while j < n and src[j].isdigit(): j += 1 if j < n and src[j] == '.': j += 1 while j < n and src[j].isdigit(): j += 1 tokens.append(Token('NUMBER', float(src[i:j]))) else: tokens.append(Token('NUMBER', int(src[i:j]))) i = j continue if ch == '+': tokens.append(Token('PLUS', '+')) elif ch == '-': tokens.append(Token('MINUS', '-')) elif ch == '*': tokens.append(Token('STAR', '*')) elif ch == '/': tokens.append(Token('SLASH', '/')) elif ch == '(': tokens.append(Token('LPAREN', '(')) elif ch == ')': tokens.append(Token('RPAREN', ')')) else: raise LexError(f"unexpected character {ch!r} at position {i}") i += 1 tokens.append(Token('EOF', None)) return tokens