diff options
author | Wim <wim@42.be> | 2017-07-07 23:39:38 +0200 |
---|---|---|
committer | Wim <wim@42.be> | 2017-07-07 23:39:38 +0200 |
commit | bfc7130ed85f66da94a33ab4326a15a149c4a291 (patch) | |
tree | 0dae0e5c5d890a97e4e57c1f94febb9fe0430969 /bridge/irc | |
parent | a0938d93869904ebf6d9938485c248b976150fac (diff) | |
download | matterbridge-msglm-bfc7130ed85f66da94a33ab4326a15a149c4a291.tar.gz matterbridge-msglm-bfc7130ed85f66da94a33ab4326a15a149c4a291.tar.bz2 matterbridge-msglm-bfc7130ed85f66da94a33ab4326a15a149c4a291.zip |
Try to detect the charset and convert it to utf-8. (irc). Closes #209 #210
Diffstat (limited to 'bridge/irc')
-rw-r--r-- | bridge/irc/irc.go | 27 |
1 files changed, 27 insertions, 0 deletions
diff --git a/bridge/irc/irc.go b/bridge/irc/irc.go index 6d85e7ba..b86ff508 100644 --- a/bridge/irc/irc.go +++ b/bridge/irc/irc.go @@ -5,8 +5,13 @@ import ( "fmt" "github.com/42wim/matterbridge/bridge/config" log "github.com/Sirupsen/logrus" + "github.com/paulrosania/go-charset/charset" + _ "github.com/paulrosania/go-charset/data" + "github.com/saintfish/chardet" ircm "github.com/sorcix/irc" "github.com/thoj/go-ircevent" + "io" + "io/ioutil" "regexp" "sort" "strconv" @@ -251,6 +256,28 @@ func (b *Birc) handlePrivMsg(event *irc.Event) { // strip IRC colors re := regexp.MustCompile(`[[:cntrl:]](\d+,|)\d+`) msg = re.ReplaceAllString(msg, "") + + // detect what were sending so that we convert it to utf-8 + detector := chardet.NewTextDetector() + result, err := detector.DetectBest([]byte(msg)) + if err != nil { + flog.Infof("detection failed for msg: %#v", msg) + return + } + flog.Debugf("detected %s confidence %#v", result.Charset, result.Confidence) + var r io.Reader + r, err = charset.NewReader(result.Charset, strings.NewReader(msg)) + // if we're not sure, just pick ISO-8859-1 + if result.Confidence < 80 { + r, err = charset.NewReader("ISO-8859-1", strings.NewReader(msg)) + } + if err != nil { + flog.Errorf("charset to utf-8 conversion failed: %s", err) + return + } + output, _ := ioutil.ReadAll(r) + msg = string(output) + flog.Debugf("Sending message from %s on %s to gateway", event.Arguments[0], b.Account) b.Remote <- config.Message{Username: event.Nick, Text: msg, Channel: event.Arguments[0], Account: b.Account, UserID: event.User + "@" + event.Host} } |