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