gout
gout copied to clipboard
guonaihong
为什么SetJSON会把入参里的&替换成\u0026 ？
在SetJSON前检查还是https://uat.xxx.com?a=1&b=1 打断点看传给SetJSON也是&，可是gout实际发出去的却是\u0026 ？
2022-04-20 09:51:01.018 INFO log/writer.go:21 > POST /cgi-bin/menu/create?access_token=56_zk9P0-rpBoNUhlneUAVJfABALGU HTTP/1.1
2022-04-20 09:51:01.018 INFO log/writer.go:21 > Content-Type: application/json
2022-04-20 09:51:01.018 INFO log/writer.go:21 >
2022-04-20 09:51:01.018 INFO log/writer.go:21
2022-04-20 09:51:01.018 INFO log/writer.go:21 {"button":[{"type":"view","name":"今日歌曲","url":"https://uat.xxx.com?a=1\u0026b=1"}]}
2022-04-20 09:51:01.018 INFO log/writer.go:21
2022-04-20 09:51:01.018 INFO log/writer.go:21 < HTTP/1.1 200 OK
2022-04-20 09:51:01.018 INFO log/writer.go:21 < Connection: keep-alive
2022-04-20 09:51:01.018 INFO log/writer.go:21 < Content-Type: application/json; encoding=utf-8
2022-04-20 09:51:01.018 INFO log/writer.go:21 < Date: Wed, 20 Apr 2022 01:51:01 GMT
2022-04-20 09:51:01.018 INFO log/writer.go:21 < Content-Length: 141
2022-04-20 09:51:01.018 INFO log/writer.go:21
2022-04-20 09:51:01.018 INFO log/writer.go:21
2022-04-20 09:51:01.018 INFO log/writer.go:21 {"errcode":40033,"errmsg":"invalid charset. please check your request, if include \\uxxxx will create fail! rid: 625f6705-437db838-0f7f9178"}
感谢, pr. 可以使用标准库的, https://pkg.go.dev/encoding/json#Encoder.SetEscapeHTML 这个方法实现这个功能.
SetEscapeHTML specifies whether problematic HTML characters should be escaped inside JSON quoted strings. The default behavior is to escape &, <, and > to \u0026, \u003c, and \u003e to avoid certain safety problems that can arise when embedding JSON in HTML.
我可能希望. SetJSON的默认语义不变, 新加参数或者接口不转义HTML相关的字符编码.
感谢回复。因为前面有一句不用encoder，如果用encoder那就可以传escapeHtml，按照目前代码的说法，这样就要处理\n，处理\n更容易误伤吧，我看json.marshal的源码有没有escapeHtml的操作也是直接替换的；
我可能觉得请求包就应该要数据前后一致、所见即所得，用替换的方式也不影响原数据本身就有的\u0026等符号，所以“新加参数或者接口不转义HTML相关的字符编码”感觉意义不大呢而且还要处理\n。
您看看怎么取舍吧，辛苦啦
我先聊聊, 最小化改造的想法
我看json.marshal的源码有没有escapeHtml的操作也是直接替换的；
指的是HTMLEscape这个函数吗? 我刚刚也喵了下标准库源代码. json.HTMLEscape主要修改了4处地方, <, >, & 最后就是U+2028到U+2029之间范围的字符
func HTMLEscape(dst *bytes.Buffer, src []byte) {
// The characters can only appear in string literals,
// so just scan the string one byte at a time.
start := 0
for i, c := range src {
if c == '<' || c == '>' || c == '&' {
if start < i {
dst.Write(src[start:i])
}
dst.WriteString(`\u00`)
dst.WriteByte(hex[c>>4])
dst.WriteByte(hex[c&0xF])
start = i + 1
}
// Convert U+2028 and U+2029 (E2 80 A8 and E2 80 A9).
if c == 0xE2 && i+2 < len(src) && src[i+1] == 0x80 && src[i+2]&^1 == 0xA8 {
if start < i {
dst.Write(src[start:i])
}
dst.WriteString(`\u202`)
dst.WriteByte(hex[src[i+2]&0xF])
start = i + 3
}
}
if start < len(src) {
dst.Write(src[start:])
}
}
我可能觉得请求包就应该要数据前后一致、所见即所得
这道理说得通. 那是否要把HTMLEscape函数逆向编码下(待定)?
逆向编码的方式还挺麻烦的. 我最后想了下, 还是Encode的方式简单点
你看Encode实现, 这个函数没有继续使用是因为标准库自我聪明加了e.WriteByte('\n')这行.
func (enc *Encoder) Encode(v any) error { ■ undeclared name: any
if enc.err != nil {
return enc.err
}
e := newEncodeState()
err := e.marshal(v, encOpts{escapeHTML: enc.escapeHTML})
if err != nil {
return err
}
// Terminate each value with a newline.
// This makes the output look a little nicer
// when debugging, and some kind of space
// is required if the encoded value was a number,
// so that the reader knows there aren't more
// digits coming.
e.WriteByte('\n')
b := e.Bytes()
if enc.indentPrefix != "" || enc.indentValue != "" {
if enc.indentBuf == nil {
enc.indentBuf = new(bytes.Buffer)
}
enc.indentBuf.Reset()
err = Indent(enc.indentBuf, b, enc.indentPrefix, enc.indentValue)
if err != nil {
return err
}
b = enc.indentBuf.Bytes()
}
if _, err = enc.w.Write(b); err != nil {
enc.err = err
}
encodeStatePool.Put(e)
return err
}
既然知道他肯定多一个'\n', 那就可以去掉'\n' 写了一个最小化的原型代码, 可以做这件事
package main
import (
"bytes"
"encoding/json"
"fmt"
)
var bytesLine = []byte("\n")
func main() {
var b bytes.Buffer
en := json.NewEncoder(&b)
en.SetEscapeHTML(false)
en.Encode(map[string]interface{}{
"aa": "aa",
"bb": "bb",
"cc": "<>",
})
fmt.Printf("old.(%s)\n", b.String())
jsonBytes := b.Bytes()
if bytes.HasSuffix(jsonBytes, bytesLine) {
jsonBytes = jsonBytes[:len(jsonBytes)-1]
}
fmt.Printf("new.(%s)\n", jsonBytes)
}