server/decoder.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
from typing import Dict, Iterator, List, Union
from .resource import Resource
from dataclasses import dataclass
from urllib import parse
import json,re
from base64 import b64decode,b64encode
from .utils import decode_base64_with_filter,truncate_encoded_string



class BaseDecoder:
    def __init__(self, encode_str: str, nameinfo: str = ""):
        self.encode_str = encode_str
        self.decode_str: str
        self.nameinfo = nameinfo

    @staticmethod
    def _bytesto_str(iobytes: bytes) -> str:
        return iobytes.decode('utf-8')


    def decode(self) -> None:
        self.decode_str = self._bytesto_str(b64decode(self.encode_str))


@dataclass
class EncodedCfg:
    encoded_config_str: str
    nameinfo: str = ""




class ListDecoder(BaseDecoder):
    def iter_encode_config(self) -> Iterator[EncodedCfg]:
        for config_str in self.decode_str.splitlines():
            _config_str = re.sub(r"(vmess|ss)://", "", config_str)
            nameinfo = ""
            if "#" in _config_str:
                _config_str, nameinfo = _config_str.split("#", 1)
                nameinfo = parse.unquote(nameinfo)

            _encoded_config_str = _config_str + (
                (4 - len(_config_str) % 4) * "="
            )

            if "trojan" in config_str and "0.0.0.0" not in config_str and '专线' not in nameinfo and '443' not in config_str:
                # 使用正则表达式去除 type=... 参数，包括#前的部分
                # 这里我们匹配 &type= 后面跟着任意字符直到下一个 & 或 # 或字符串结束 ,去除type=tcp== 或者type=tcp= 这类影响链接的多余部分
                _encoded_config_str_without_type = re.sub(r'([&?]type=[^&#]*)', '',_encoded_config_str)

                if "allowInsecure" in config_str:
                    yield _encoded_config_str_without_type.replace('allowInsecure=0', 'allowInsecure=1')+"#"+truncate_encoded_string(nameinfo) # 适当缩短name
                else:
                    yield _encoded_config_str_without_type+"&allowInsecure=1"+"#"+truncate_encoded_string(nameinfo) # 适当缩短name

            # vmess无法获取到nameinfo 因为在里面的ps这个key对于的value作为remark
            if ("vmess" in config_str):
                vmess_decoded_data = decode_base64_with_filter(_config_str)
                vmess_info_json = json.loads(vmess_decoded_data)
                vmess_info_str = json.dumps(vmess_info_json,ensure_ascii=False)#不转译中文
                if ('倍率提示'not in vmess_info_str)and('导航' not in vmess_info_str) and('443' not in vmess_info_str):
                    # 获取 ps 对应的remark值
                    ps_value = vmess_info_json.get('ps', '')  # 如果 'ps' 键不存在，默认值为空字符串
                    # 如果 ps_value 不为空，则调用 truncate_encoded_string 函数并更新值。如果 ps_value 为空，则保持原值（空字符串）。
                    vmess_info_json['ps'] = truncate_encoded_string(ps_value) if ps_value else ps_value
                    vmess_node_return = 'vmess://'+b64encode(json.dumps(vmess_info_json,ensure_ascii=False).encode("utf-8")).decode("utf-8")
                    yield vmess_node_return

            # 不能简单判断ss:// 因为vmess://也包含ss://
            if (config_str.startswith("ss://")) and ("套餐" not in nameinfo) and('到期' not in nameinfo) and('流量' not in nameinfo) and('剩余' not in nameinfo) and ('专线' not in nameinfo):
                #shadowsocks节点保留
                yield 'ss://'+_config_str+'#'+truncate_encoded_string(nameinfo)

class ConfigDecoder(BaseDecoder):
    def get_json(self) -> Dict:
        if re.match(r"^{.*}$", self.decode_str):
            return json.loads(self.decode_str)
        return self._parse_ss()

    def _parse_ss(self):
        crypt, pw_at_addr, port = self.decode_str.split(":")
        pw, addr = pw_at_addr.split("@")
        return {
            "ps": self.nameinfo,
            "add": addr,
            "port": int(port),
            "method": crypt,
            "password": pw,
            "ss": True,
            "level": 0,
        }


def _get_resource_from_url(url: str) -> str:
    resource = Resource(url)
    return resource.get_encoded_data()


def _get_listencoded_cfg_from_encoded_str(encoded_str: str) -> List[EncodedCfg]:
    decoder = ListDecoder(encoded_str)
    decoder.decode()
    return [item for item in decoder.iter_encode_config()]





def decode_url_to_configs(url: str)->list:
    encoded_str = _get_resource_from_url(url)
    lst_encoded_cfg = _get_listencoded_cfg_from_encoded_str(encoded_str)#lst_encoded_cfg解析为节点后的列表,包含所有节点链接
    return lst_encoded_cfg