json中非法字符

小灰灰 2022-07-16 04:17 266阅读 0赞

原文：

Hi there,   
  
I just discovered Go and decided to port a little program to Go.   
  
The program reads JSON-Data from an URL and process the Data. The Go   
port works well till now.   
  
I dont have any influence on the JSON data and so sometimes there are   
control character in it and my program crashes with "invalid character   
'\\x12' in string literal"   
  
here the code sample of my program:

![复制代码][copycode.gif]

http_return, err := http.Get(newurl) 
    var http_body []byte; 
    if err == nil { 
      http_body, err = ioutil.ReadAll(http_return.Body) 
      http_return.Body.Close() 
    } 
    param_info := make(map[string]interface{}) 
    param_err := json.Unmarshal(http_body, &param_info)

![复制代码][copycode.gif]

**This Unmarshal call crashes with "invalid character '\\x12' in string **  
**literal" **  
  
How do I remove this or any similar control character from the JSON   
data. I dont need those characters so they can simply be removed.   
  
Thanks   
Nico

==============================================================

意思大概就是说，json是通过网络传输或者其它方法获取的，然后里面可能包含特殊字符,这样呢，在go里面用json 解析就会报错。

于是有个哥们就出了一招。

========================

If you don't mind them being replaced with spaces, you could do something like:

![复制代码][copycode.gif]

for i, ch := range http_body {
    
      switch {
      case ch > '~':   http_body[i] = ' '
      case ch == '\r':
      case ch == '\n':
      case ch == '\t':
      case ch < ' ':   http_body[i] = ' '
      }
    }

![复制代码][copycode.gif]

========================================

具体什么原理呢？还不清楚。

大概是根据字符集来判断哪些是json里面的非法字符，然后替换成空格。

计算机，字符集。。。。。后面需要研究一下了。

===================================

看一下ASCII可显示字符表，就明白了。 空格开始，~结束。加上\\r\\n\\t这3个字符。

[copycode.gif]: /images/20220716/20bbb1a5250846b8a482b3266b788595.png