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: tokens = [] i = 0 while i < len(src): ch = src[i] if ch in ' \t': i += 1 continue if ch.isdigit() or (ch == '.' and i + 1 < len(src) and src[i + 1].isdigit()): j = i while j < len(src) and src[j].isdigit(): j += 1 if j < len(src) and src[j] == '.': j += 1 while j < len(src) and src[j].isdigit(): j += 1 value = float(src[i:j]) else: value = int(src[i:j]) tokens.append(Token('NUMBER', value)) i = j continue if ch == '+': tokens.append(Token('PLUS', '+')) i += 1 elif ch == '-': tokens.append(Token('MINUS', '-')) i += 1 elif ch == '*': tokens.append(Token('STAR', '*')) i += 1 elif ch == '/': tokens.append(Token('SLASH', '/')) i += 1 elif ch == '(': tokens.append(Token('LPAREN', '(')) i += 1 elif ch == ')': tokens.append(Token('RPAREN', ')')) i += 1 else: raise LexError(f"unexpected character {ch!r} at position {i}") tokens.append(Token('EOF', None)) return tokens