From 58adf941723f4c2ba8779bdf7f8e0fea8de9f7fe Mon Sep 17 00:00:00 2001
From: John Pennycook <john.pennycook@intel.com>
Date: Wed, 18 Dec 2024 09:32:24 +0000
Subject: [PATCH] Refactor Token class using dataclass

Token is simple enough that @dataclass is sufficient to generate the desired
__init__ and __repr__ definitions.

Signed-off-by: John Pennycook <john.pennycook@intel.com>
---
 codebasin/preprocessor.py | 46 ++++++++++++---------------------------
 1 file changed, 14 insertions(+), 32 deletions(-)

diff --git a/codebasin/preprocessor.py b/codebasin/preprocessor.py
index 2323525..d97b19f 100644
--- a/codebasin/preprocessor.py
+++ b/codebasin/preprocessor.py
@@ -13,6 +13,7 @@
 import os
 from collections.abc import Callable, Iterable
 from copy import copy
+from dataclasses import dataclass, field
 from enum import Enum
 from typing import Self
 
@@ -68,19 +69,16 @@ class MacroExpandOverflow(ValueError):
     """
 
 
+@dataclass
 class Token:
     """
     Represents a token constructed by the parser.
     """
 
-    def __init__(self, line, col, prev_white, token):
-        self.line = line
-        self.col = col
-        self.prev_white = prev_white
-        self.token = token
-
-    def __repr__(self):
-        return _representation_string(self)
+    line: int
+    col: int
+    prev_white: bool
+    token: str
 
     def __str__(self):
         return "\n".join(self.spelling())
@@ -100,15 +98,14 @@ def sanitized_str(self):
         return str(self)
 
 
+@dataclass
 class CharacterConstant(Token):
     """
     Represents a character constant.
     """
 
-    def __repr__(self):
-        return _representation_string(self)
-
 
+@dataclass
 class NumericalConstant(Token):
     """
     Represents a 'preprocessing number'.
@@ -116,18 +113,13 @@ class NumericalConstant(Token):
     not be valid syntax).
     """
 
-    def __repr__(self):
-        return _representation_string(self)
-
 
+@dataclass
 class StringConstant(Token):
     """
     Represents a string constant.
     """
 
-    def __repr__(self):
-        return _representation_string(self)
-
     def spelling(self):
         """
         Return the string representation of this token in the input code.
@@ -155,45 +147,35 @@ def sanitized_str(self):
         return "".join(out)
 
 
+@dataclass
 class Identifier(Token):
     """
     Represents a C identifier.
     """
 
-    def __init__(self, line, col, prev_white, token):
-        super().__init__(line, col, prev_white, token)
-        self.expandable = True
-
-    def __repr__(self):
-        return _representation_string(self)
+    expandable: bool = field(default=True, init=True)
 
 
+@dataclass
 class Operator(Token):
     """
     Represents a C operator.
     """
 
-    def __repr__(self):
-        return _representation_string(self)
-
 
+@dataclass
 class Punctuator(Token):
     """
     Represents a punctuator (e.g. parentheses)
     """
 
-    def __repr__(self):
-        return _representation_string(self)
-
 
+@dataclass
 class Unknown(Token):
     """
     Represents an unknown token.
     """
 
-    def __repr__(self):
-        return _representation_string(self)
-
 
 class Lexer:
     """