URI＃getQuery（）无法解码非法字符

Question

为什么

new URI(
    null,
    null,
    "/monitor/123456/data",
    "filterParams=[{\"id\":12,\"type\":\"dimension\"}]",
    null).getQuery()

返回

  

filterParams = [％7B％22id％22：12，％22type％22：％22dimension％22％7D]

而不是

  

“filterParams = [{\” ID \ “：12，\” 类型\ “：\” 维度\ “}]”

根据URI#getQuery()文档

  

此方法返回的字符串等于#getRawQuery()方法返回的字符串，但所有转义的八位字节序列都被解码。

这是JDK错误还是我在这里缺少一些理解？

Answer 1

根据私有方法URI.decode上面的评论，它不是一个bug而是一个特征：）。

/*
Evaluates all escapes in s, applying UTF-8 decoding if needed.  Assumes
that escapes are well-formed syntactically, i.e., of the form %XX.  If a
sequence of escaped octets is not valid UTF-8 then the erroneous octets
are replaced with '\uFFFD'.
Exception: any "%" found between "[]" is left alone. It is an IPv6  literal with a scope_id
*/
private static String decode(String s) {...}