4
0
mirror of https://github.com/cwinfo/matterbridge.git synced 2025-06-26 07:29:24 +00:00

Refactor utf-8 conversion (irc) (#1767)

This commit is contained in:
Wim
2022-03-19 23:14:56 +01:00
committed by GitHub
parent 663850a2b8
commit f0f801402d
11 changed files with 719 additions and 618 deletions

32
bridge/irc/charset.go Normal file
View File

@ -0,0 +1,32 @@
package birc
import (
"golang.org/x/text/encoding"
"golang.org/x/text/encoding/japanese"
"golang.org/x/text/encoding/korean"
"golang.org/x/text/encoding/simplifiedchinese"
"golang.org/x/text/encoding/traditionalchinese"
"golang.org/x/text/encoding/unicode"
)
var encoders = map[string]encoding.Encoding{
"utf-8": unicode.UTF8,
"iso-2022-jp": japanese.ISO2022JP,
"big5": traditionalchinese.Big5,
"gbk": simplifiedchinese.GBK,
"euc-kr": korean.EUCKR,
"gb2312": simplifiedchinese.HZGB2312,
"shift-jis": japanese.ShiftJIS,
"euc-jp": japanese.EUCJP,
"gb18030": simplifiedchinese.GB18030,
}
func toUTF8(from string, input string) string {
enc, ok := encoders[from]
if !ok {
return input
}
res, _ := enc.NewDecoder().String(input)
return res
}

View File

@ -11,7 +11,6 @@ import (
"github.com/42wim/matterbridge/bridge/config"
"github.com/42wim/matterbridge/bridge/helper"
"github.com/lrstanley/girc"
"github.com/missdeer/golib/ic"
"github.com/paulrosania/go-charset/charset"
"github.com/saintfish/chardet"
@ -24,12 +23,12 @@ func (b *Birc) handleCharset(msg *config.Message) error {
if b.GetString("Charset") != "" {
switch b.GetString("Charset") {
case "gbk", "gb18030", "gb2312", "big5", "euc-kr", "euc-jp", "shift-jis", "iso-2022-jp":
msg.Text = ic.ConvertString("utf-8", b.GetString("Charset"), msg.Text)
msg.Text = toUTF8(b.GetString("Charset"), msg.Text)
default:
buf := new(bytes.Buffer)
w, err := charset.NewWriter(b.GetString("Charset"), buf)
if err != nil {
b.Log.Errorf("charset from utf-8 conversion failed: %s", err)
b.Log.Errorf("charset to utf-8 conversion failed: %s", err)
return err
}
fmt.Fprint(w, msg.Text)
@ -227,7 +226,7 @@ func (b *Birc) handlePrivMsg(client *girc.Client, event girc.Event) {
}
switch mycharset {
case "gbk", "gb18030", "gb2312", "big5", "euc-kr", "euc-jp", "shift-jis", "iso-2022-jp":
rmsg.Text = ic.ConvertString("utf-8", b.GetString("Charset"), rmsg.Text)
rmsg.Text = toUTF8(b.GetString("Charset"), rmsg.Text)
default:
r, err := charset.NewReader(mycharset, strings.NewReader(rmsg.Text))
if err != nil {