Go语言解析邮件出现乱码的解决方法与避坑指南

const raw = "From: =?GB2312?B?zajWqrf+zvE=?= <noreply@example.com>\r\n" +
	"To: someone@example.com\r\n" +
	"Subject: =?GBK?B?xPq1xNHp1qTC686q?= =?UTF-8?B?77yaODQ3MjAx?=\r\n" +
	"MIME-Version: 1.0\r\n" +
	"Content-Type: multipart/alternative; boundary=\"b1\"\r\n" +
	"\r\n" +
	"--b1\r\n" +
	"Content-Type: text/plain; charset=GBK\r\n" +
	"Content-Transfer-Encoding: quoted-printable\r\n" +
	"\r\n" +
	"=C4=FA=B5=C4=D1=E9=D6=A4=C2=EB=CA=C7 847201=A3=AC10 =B7=D6=D6=D3=C4=DA=D3=D0=D0=A7=A1=A3\r\n" +
	"--b1\r\n" +
	"Content-Type: text/html; charset=UTF-8\r\n" +
	"Content-Transfer-Encoding: base64\r\n" +
	"\r\n" +
	"PHA+5oKo55qE6aqM6K+B56CB5pivIDxiPjg0NzIwMTwvYj48L3A+\r\n" +
	"--b1--\r\n"

func charsetReader(charset string, input io.Reader) (io.Reader, error) {
	switch strings.ToLower(charset) {
	case "utf-8", "us-ascii", "":
		return input, nil
	case "gbk", "gb2312", "gb18030":
		// GB2312 和 GBK 都是 GB18030 的子集，统一按 GB18030 解最稳
		return transform.NewReader(input, simplifiedchinese.GB18030.NewDecoder()), nil
	}
	return nil, fmt.Errorf("unsupported charset: %s", charset)
}
var dec = &mime.WordDecoder{CharsetReader: charsetReader}
func decodeHeader(s string) string {
	out, err := dec.DecodeHeader(s)
	if err != nil {
		return s // 解不了就原样返回，别让一个头拖垮整封信
	}
	return out
}

type Part struct {
	ContentType string
	Body        string
}
func walk(r io.Reader, contentType, cte string, out *[]Part) error {
	mediaType, params, err := mime.ParseMediaType(contentType)
	if err != nil {
		// Content-Type 写错的邮件并不少见，兜底当纯文本
		mediaType, params = "text/plain", map[string]string{}
	}
	if strings.HasPrefix(mediaType, "multipart/") {
		mr := multipart.NewReader(r, params["boundary"])
		for {
			p, err := mr.NextRawPart()
			if err == io.EOF {
				return nil
			}
			if err != nil {
				return err
			}
			if err := walk(p, p.Header.Get("Content-Type"), p.Header.Get("Content-Transfer-Encoding"), out); err != nil {
				return err
			}
		}
	}
	// 叶子节点：先解传输编码，再解字符集，顺序不能反
	var body io.Reader = r
	switch strings.ToLower(strings.TrimSpace(cte)) {
	case "quoted-printable":
		body = quotedprintable.NewReader(body)
	case "base64":
		body = base64Reader(body)
	}
	cr, err := charsetReader(params["charset"], body)
	if err != nil {
		return err
	}
	b, err := io.ReadAll(cr)
	if err != nil {
		return err
	}
	*out = append(*out, Part{ContentType: mediaType, Body: string(b)})
	return nil
}

func base64Reader(r io.Reader) io.Reader {
	b, _ := io.ReadAll(r)
	clean := strings.Map(func(c rune) rune {
		if c == '\r' || c == '\n' || c == ' ' || c == '\t' {
			return -1
		}
		return c
	}, string(b))
	return base64.NewDecoder(base64.StdEncoding, strings.NewReader(clean))
}

func main() {
	msg, err := mail.ReadMessage(strings.NewReader(raw))
	if err != nil {
		panic(err)
	}
	fmt.Println("原始 Subject:", msg.Header.Get("Subject"))
	fmt.Println("解码 Subject:", decodeHeader(msg.Header.Get("Subject")))
	fmt.Println("解码 From   :", decodeHeader(msg.Header.Get("From")))
	var parts []Part
	if err := walk(msg.Body, msg.Header.Get("Content-Type"), msg.Header.Get("Content-Transfer-Encoding"), &parts); err != nil {
		panic(err)
	}
	for _, p := range parts {
		fmt.Printf("[%s] %s\n", p.ContentType, strings.TrimSpace(p.Body))
	}
}

import (
	"encoding/base64"
	"fmt"
	"io"
	"mime"
	"mime/multipart"
	"mime/quotedprintable"
	"net/mail"
	"strings"
	"golang.org/x/text/encoding/simplifiedchinese"
	"golang.org/x/text/transform"
)