json字符串类型的value换行方案

json字符串类型的value换行方案

按照标准json语法,字符串类型的value是不能换行写的.
例如,以下是错误的写法

{
    "key":"hello
 world"
}

但是遇到了需要在json中写代码与服务器交互的情况,无奈只能这样写:

{
    "key":"var a = 1\nvar b = 2\n var c=a+b"
}

代码行数少了还好,多了极其**,于是想办法可以在json中换行写代码.

尝试1: json5

JSON5官网

JSON5号称"JSON for Humans",相比标准json主要特点如下:

• 字符串value支持换行
• key可以不加""
• key和value可以使用''代替""
• 支持注释
• 二进制数值
• ...

栗子:
以下是一个合法的json5

{
  // comments
  unquoted: 'and you can quote me on that',
  singleQuotes: 'I can use "double quotes" here',
  lineBreaks: "Look, Mom! \
No \\n's!",
  hexadecimal: 0xdecaf,
  leadingDecimalPoint: .8675309, andTrailing: 8675309.,
  positiveSign: +1,
  trailingComma: 'in objects', andIn: ['arrays',],
  "backwardsCompatible": "with JSON",
}

• 优点:现成
• 缺点:

  • 换行需要用 \ \标识;

  • 解析之后因为没有了换行,代码无法执行,所以是达不到目的的;

    {
        "key":"var a = 1\
    var b = 2\
    var c=a+b"
    }
    
    解析之后的结果为
    {
        "key":"var a = 1var b = 2var c=a+b"
    }
    

尝试2: 编写自己的json解释器

• 优点: 可根据自己定义的规则解析数据
• 缺点: 成本高.

以下为用python实现的json解释器,支持换行,目前没有做完整的错误处理,请不要用于生产:

"""
Topic: 下降解析器
Desc :
"""
import re
import collections

# Token specification
NUM = r'(?P<NUM>\d+)'
STR_D = r'"(?P<STR_D>[^"]*?)"'
STR_S = r'(?P<STR_S>\'.*?\')'
COLON = r'(?P<COLON>:)'
COMMA = r'(?P<COMMA>,)'
LLB = r'(?P<LLB>\[)'
RLB = r'(?P<RLB>\])'
LDB = r'(?P<LDB>\{)'
RDB = r'(?P<RDB>\})'
WS = r'(?P<WS>\s+)'
NULL = r'(?P<NULL>null)'
FALSE = r'(?P<FALSE>false)'
TRUE = r'(?P<TRUE>true)'

master_pat = re.compile('|'.join([NUM, STR_D, STR_S, LLB, RLB,
                                  LDB, RDB, COLON, COMMA, FALSE, TRUE, NULL, WS]), re.M | re.S)
# Tokenizer
Token = collections.namedtuple('Token', ['type', 'value'])


def generate_tokens(text):
    scanner = master_pat.scanner(text)
    for m in iter(scanner.match, None):
        tok = Token(m.lastgroup, m.group())
        if tok.type != 'WS':
            yield tok


def simple_join(text):
    li = []
    for t in generate_tokens(text):
        res = t[1]
        if t[0] == "STR_S":
            res = t[1].replace("'", "\"")
        elif t[0].startswith("STR_"):
            res = t[1].replace("\n", "\\n").replace("\r\n", "\\n")

        li.append(str(res))

    return "".join(li)


class Evaluator:
    def __init__(self, text):
        self.text = text

    def parse(self):
        self.tokens = generate_tokens(self.text)
        self.nexttok = next(self.tokens, None)
        self._advance()
        return self._parse()

    def _advance(self):
        '''Advance one token ahead'''
        self.tok, self.nexttok = self.nexttok, next(self.tokens, None)

    def _parse(self):
        if self.tok[0] == "LLB":
            return self.get_list()
        if self.tok[0] == "LDB":
            return self.get_dict()
        if self.tok[0].startswith("STR_"):
            return self.tok[1].strip('"')
        if self.tok[0] == "NUM":
            return int(self.tok[1])
        if self.tok[0] == "NULL":
            return None
        if self.tok[0] == "FALSE":
            return False
        if self.tok[0] == "TRUE":
            return True
        raise Exception("未知token:{}".format(self.tok[1]))

    def get_dict(self):
        """
        { 开头
        中间内容必须为 k:v,
        结尾必须为 }
        """
        res = {}
        self._advance()
        # }
        if self.tok[0] == "RDB":
            return {}

        def parse_value():
            if not self.tok[0].startswith("STR_"):
                raise Exception(f"KEY需是字符串形式,{self.tok[1]}")
            new_key = self._parse()
            self._advance()
            if self.tok[0] != "COLON":
                raise Exception("KEY和VALUE需用:分割")
            self._advance()
            res[new_key] = self._parse()
            self._advance()

        parse_value()
        while self.tok[0] == "COMMA":
            self._advance()
            parse_value()

        return res

    def get_list(self):
        res = []
        self._advance()
        # }
        if self.tok[0] == "RLB":
            return []

        def parse_value():
            if self.tok[0] in ["NUM", "STR_D", "STR_S", "NULL", "FALSE", "TRUE"]:
                res.append(self._parse())
            elif self.tok[0] == "LLB":
                res.append(self.get_list())
            elif self.tok[0] == "LDB":
                res.append(self.get_dict())

        parse_value()
        self._advance()
        while self.tok[0] == "COMMA":
            self._advance()
            parse_value()
            self._advance()
        return res


if __name__ == '__main__':
    text = """
        {
        "k1":1,
        "k2":"v2",
        "is_true":true,
        "is_none":null,
        "k2":"v2",
        "k3":{
        "a1":"a1",
        "a2":"a2",
        "code":"
        var code = 0
        code = 1
        "
        },
        "list":[1,2,3,4]
    }
    """
    # test(text)
    e = Evaluator(text)
    res = e.parse()
    print(res) #{'k1': 1, 'k2': 'v2', 'is_true': True, 'is_none': None, 'k3': {'a1': 'a1', 'a2': 'a2', 'code': '\n        var code = 0\n        code = 1\n        '}, 'list': [1, 2, 3, 4]}

尝试3: 全局替换 \n 为 \n

• 优点:简单粗暴

  {
      "key":"
  var a = 1
  var b = 2
  var c=a+b"
  }
  
  替换之后的结果为
  {
      "key":"\nvar a = 1\nvar b = 2\n var c=a+b"
  }
  

• 缺点:虽然写的时候有换行,但是上传到服务器再查询的时候只能看到\n,依然缺乏可读性;

尝试4(最终采用方案): 结合UI将json生成树状节点,将代码类型的value单独显示.

• 优点
  • 难度一般
  • 按常规方式解析json
  • 上传到服务器再查询的时候也能看到换行.