38029-vm/core/parser.py

import re

class Lexer:
    def __init__(self, code):
        self.code = code
        self.tokens = []
        self.pos = 0
        LB, RB, DQ, SQ, BS = chr(91), chr(93), chr(34), chr(39), chr(92)
        OP_LIST = r'==|~=|<=|>=|\.\.\.|\.\.|>>|<<|\+|\-|\*|/|%|\^|#|=|\<|\>|\(|\)|\{|\}|' + BS + LB + '|' + BS + RB + '|;|:|,|\.'
        self.rules = [
            ('COMMENT', re.compile('--' + LB + LB + '.*?' + RB + RB + '|--.*', re.DOTALL)),
            ('STRING', re.compile(DQ + r'(?:' + BS + BS + r'.|[^' + DQ + BS + BS + r'])*' + DQ + '|' + SQ + r'(?:' + BS + BS + r'.|[^' + SQ + BS + BS + r'])*' + SQ + '|' + LB + LB + '.*?' + RB + RB, re.DOTALL)),
            ('NUMBER', re.compile(r'\b\d+\.?\d*(?:[eE][+-]?\d+)?\b|\b0x[0-9a-fA-F]+\b')),
            ('KEYWORD', re.compile(r'\b(and|break|do|else|elseif|end|false|for|function|if|in|local|nil|not|or|repeat|return|then|true|until|while)\b')),
            ('IDENT', re.compile(r'[a-zA-Z_][a-zA-Z0-9_]*')),
            ('OP', re.compile(OP_LIST)),
            ('SPACE', re.compile(r'\s+'))
        ]
    def tokenize(self):
        while self.pos < len(self.code):
            match = None
            for name, regex in self.rules:
                match = regex.match(self.code, self.pos)
                if match:
                    if name != 'SPACE' and name != 'COMMENT': self.tokens.append((name, match.group(0)))
                    self.pos = match.end(); break
            if not match: self.pos += 1
        return self.tokens

class Parser:
    def __init__(self, tokens): self.tokens = tokens; self.pos = 0
    def peek(self, offset=0):
        idx = self.pos + offset
        return self.tokens[idx] if idx < len(self.tokens) else (None, None)
    def consume(self, t=None, v=None):
        tk = self.peek()
        if not tk or not tk[0]: return None
        if t and tk[0] != t: return None
        if v and tk[1] != v: return None
        self.pos += 1; return tk
    def parse(self):
        nodes = []
        while self.pos < len(self.tokens):
            node = self.parse_statement()
            if node: nodes.append(node)
            else: self.pos += 1
        return nodes
    def parse_statement(self):
        tk = self.peek()
        if not tk or not tk[0]: return None
        if tk[1] == 'local':
            self.consume(); ident = self.consume('IDENT')
            if ident:
                if self.peek()[1] == '=':
                    self.consume(); return {'type': 'assign', 'name': ident[1], 'value': self.parse_expression(), 'local': True}
                return {'type': 'assign', 'name': ident[1], 'value': {'type': 'KEYWORD', 'value': 'nil'}, 'local': True}
            return None
        if tk[1] == 'return':
            self.consume()
            return {'type': 'return', 'value': self.parse_expression()}
        if tk[1] == ';':
            self.consume(); return None
        start = self.pos; expr = self.parse_prefix_expression()
        if expr:
            if self.peek()[1] == '=':
                self.consume(); return {'type': 'assign', 'target': expr, 'value': self.parse_expression()}
            elif expr['type'] in ['call', 'method_call']: return expr
        self.pos = start; return None
    def parse_prefix_expression(self):
        tk = self.peek()
        if not tk or tk[0] != 'IDENT': return None
        expr = {'type': 'variable', 'name': self.consume()[1]}
        while True:
            nt = self.peek()
            if nt[1] == '.':
                self.consume(); m = self.consume('IDENT')
                if m: expr = {'type': 'index', 'base': expr, 'key': {'type': 'STRING', 'value': '"'+m[1]+'"'}}
                else: break
            elif nt[1] == ':':
                self.consume(); m = self.consume('IDENT')
                if m:
                    if self.peek()[1] == '(':
                        self.consume(); args = []
                        if self.peek()[1] != ')':
                            args.append(self.parse_expression())
                            while self.peek()[1] == ',': self.consume(); args.append(self.parse_expression())
                        self.consume('OP', ')'); expr = {'type': 'method_call', 'base': expr, 'method': m[1], 'args': args}
                    else: expr = {'type': 'index', 'base': expr, 'key': {'type': 'STRING', 'value': '"'+m[1]+'"'}}
                else: break
            elif nt[1] == '[':
                self.consume(); key = self.parse_expression(); self.consume('OP', ']'); expr = {'type': 'index', 'base': expr, 'key': key}
            elif nt[1] == '(':
                self.consume(); args = []
                if self.peek()[1] != ')':
                    args.append(self.parse_expression())
                    while self.peek()[1] == ',': self.consume(); args.append(self.parse_expression())
                self.consume('OP', ')'); expr = {'type': 'call', 'func': expr, 'args': args}
            else: break
        return expr
    def parse_expression(self):
        left = self.parse_primary_expression()
        if not left: return None
        while self.peek()[1] in ['+', '-', '*', '/', '..', '==', '<', '>', '<=', '>=', '~=']:
            op = self.consume()[1]; right = self.parse_primary_expression()
            if not right: break
            left = {'type': 'binary', 'op': op, 'left': left, 'right': right}
        return left
    def parse_primary_expression(self):
        tk = self.peek()
        if not tk: return None
        if tk[1] in ['-', '#', 'not']:
            op = self.consume()[1]; sub = self.parse_primary_expression()
            return {'type': 'unary', 'op': op, 'operand': sub} if sub else None
        if tk[0] in ['STRING', 'NUMBER']: v = tk[1]; self.consume(); return {'type': tk[0], 'value': v}
        if tk[1] in ['true', 'false', 'nil']: return {'type': 'KEYWORD', 'value': self.consume()[1]}
        if tk[1] == '(': self.consume(); e = self.parse_expression(); self.consume('OP', ')'); return e
        return self.parse_prefix_expression()