Merge pull request #1598 from slingamn/issue1596_truncation

fix #1596

go.mod

@@ -10,7 +10,7 @@ require (
 	github.com/go-sql-driver/mysql v1.5.0
 	github.com/go-test/deep v1.0.6 // indirect
 	github.com/gorilla/websocket v1.4.2
-	github.com/goshuirc/irc-go v0.0.0-20210311004346-ea7a188a73fe
+	github.com/goshuirc/irc-go v0.0.0-20210318074529-bdc2c2cd2fef
 	github.com/onsi/ginkgo v1.12.0 // indirect
 	github.com/onsi/gomega v1.9.0 // indirect
 	github.com/oragono/confusables v0.0.0-20201108231250-4ab98ab61fb1

go.sum

@@ -44,6 +44,8 @@ github.com/goshuirc/irc-go v0.0.0-20210304031553-cf78e9176f96 h1:sihI3HsrJWyS4Mt
 github.com/goshuirc/irc-go v0.0.0-20210304031553-cf78e9176f96/go.mod h1:q/JhvvKLmif3y9q8MDQM+gRCnjEKnu5ClF298TTXJug=
 github.com/goshuirc/irc-go v0.0.0-20210311004346-ea7a188a73fe h1:5UsPgeXJBkFgJK3Ml0nj6ljasjd26xiUxALnDJHmipE=
 github.com/goshuirc/irc-go v0.0.0-20210311004346-ea7a188a73fe/go.mod h1:q/JhvvKLmif3y9q8MDQM+gRCnjEKnu5ClF298TTXJug=
+github.com/goshuirc/irc-go v0.0.0-20210318074529-bdc2c2cd2fef h1:07e6GcSuNh1BoZJigrvaJSpe2PsYJgkYETOuGKpM2co=
+github.com/goshuirc/irc-go v0.0.0-20210318074529-bdc2c2cd2fef/go.mod h1:q/JhvvKLmif3y9q8MDQM+gRCnjEKnu5ClF298TTXJug=
 github.com/hpcloud/tail v1.0.0 h1:nfCOvKYfkgYP8hkirhJocXT2+zOD8yUNjXaWfTlyFKI=
 github.com/hpcloud/tail v1.0.0/go.mod h1:ab1qPbhIpdTxEkNHXyeSf5vhxWSCs/tWer42PpOxQnU=
 github.com/onsi/ginkgo v1.6.0/go.mod h1:lLunBs/Ym6LB5Z9jYTR76FiuTmxDTDusOGeTQH+WWjE=

irc/channel.go

@@ -13,6 +13,8 @@ import (
 
 	"sync"
 
+	"github.com/goshuirc/irc-go/ircutils"
+
 	"github.com/oragono/oragono/irc/caps"
 	"github.com/oragono/oragono/irc/history"
 	"github.com/oragono/oragono/irc/modes"
@@ -1259,10 +1261,7 @@ func (channel *Channel) SetTopic(client *Client, topic string, rb *ResponseBuffe
 		return
 	}
 
-	topicLimit := client.server.Config().Limits.TopicLen
-	if len(topic) > topicLimit {
-		topic = topic[:topicLimit]
-	}
+	topic = ircutils.TruncateUTF8Safe(topic, client.server.Config().Limits.TopicLen)
 
 	channel.stateMutex.Lock()
 	chname := channel.name
@@ -1520,10 +1519,7 @@ func (channel *Channel) Kick(client *Client, target *Client, comment string, rb
 		return
 	}
 
-	kicklimit := channel.server.Config().Limits.KickLen
-	if len(comment) > kicklimit {
-		comment = comment[:kicklimit]
-	}
+	comment = ircutils.TruncateUTF8Safe(comment, channel.server.Config().Limits.KickLen)
 
 	message := utils.MakeMessage(comment)
 	details := client.Details()

irc/handlers.go

@@ -22,6 +22,9 @@ import (
 
 	"github.com/goshuirc/irc-go/ircfmt"
 	"github.com/goshuirc/irc-go/ircmsg"
+	"github.com/goshuirc/irc-go/ircutils"
+	"golang.org/x/crypto/bcrypt"
+
 	"github.com/oragono/oragono/irc/caps"
 	"github.com/oragono/oragono/irc/custime"
 	"github.com/oragono/oragono/irc/flatip"
@@ -30,7 +33,6 @@ import (
 	"github.com/oragono/oragono/irc/modes"
 	"github.com/oragono/oragono/irc/sno"
 	"github.com/oragono/oragono/irc/utils"
-	"golang.org/x/crypto/bcrypt"
 )
 
 // helper function to parse ACC callbacks, e.g., mailto:person@example.com, tel:16505551234
@@ -346,10 +348,7 @@ func awayHandler(server *Server, client *Client, msg ircmsg.Message, rb *Respons
 	if len(msg.Params) > 0 {
 		isAway = true
 		awayMessage = msg.Params[0]
-		awayLen := server.Config().Limits.AwayLen
-		if len(awayMessage) > awayLen {
-			awayMessage = awayMessage[:awayLen]
-		}
+		awayMessage = ircutils.TruncateUTF8Safe(awayMessage, server.Config().Limits.AwayLen)
 	}
 
 	rb.session.SetAway(awayMessage)

vendor/github.com/goshuirc/irc-go/ircutils/doc.go

@@ -0,0 +1,9 @@
+// written by Daniel Oaks <daniel@danieloaks.net>
+// released under the ISC license
+
+/*
+Package ircutils provides small, useful utility functions and classes.
+
+This package is in an alpha stage.
+*/
+package ircutils

vendor/github.com/goshuirc/irc-go/ircutils/hostnames.go

@@ -0,0 +1,41 @@
+// written by Daniel Oaks <daniel@danieloaks.net>
+// released under the ISC license
+
+package ircutils
+
+import "strings"
+
+var allowedHostnameChars = "abcdefghijklmnopqrstuvwxyz1234567890-."
+
+// HostnameIsValid provides a way for servers to check whether a looked-up client
+// hostname is valid (see InspIRCd #1033 for why this is required).
+//
+// This function shouldn't be called by clients since they don't need to validate
+// hostnames for IRC use, just by servers that need to confirm hostnames of incoming
+// clients.
+//
+// In addition to this function, servers should impose their own limits on max
+// hostname length -- this function limits it to 200 but most servers will probably
+// want to make it smaller than that.
+func HostnameIsValid(hostname string) bool {
+	// IRC hostnames specifically require a period, rough limit of 200 chars
+	if !strings.Contains(hostname, ".") || len(hostname) < 1 || len(hostname) > 200 {
+		return false
+	}
+
+	// ensure each part of hostname is valid
+	for _, part := range strings.Split(hostname, ".") {
+		if len(part) < 1 || len(part) > 63 || strings.HasPrefix(part, "-") || strings.HasSuffix(part, "-") {
+			return false
+		}
+	}
+
+	// ensure all chars of hostname are valid
+	for _, char := range strings.Split(strings.ToLower(hostname), "") {
+		if !strings.Contains(allowedHostnameChars, char) {
+			return false
+		}
+	}
+
+	return true
+}

vendor/github.com/goshuirc/irc-go/ircutils/unicode.go

@@ -0,0 +1,25 @@
+// Copyright (c) 2021 Shivaram Lingamneni
+// Released under the MIT License
+
+package ircutils
+
+import (
+	"unicode/utf8"
+)
+
+// truncate a message, taking care not to make valid UTF8 into invalid UTF8
+func TruncateUTF8Safe(message string, byteLimit int) (result string) {
+	if len(message) <= byteLimit {
+		return message
+	}
+	message = message[:byteLimit]
+	for i := 0; i < (utf8.UTFMax - 1); i++ {
+		r, n := utf8.DecodeLastRuneInString(message)
+		if r == utf8.RuneError && n <= 1 {
+			message = message[:len(message)-1]
+		} else {
+			break
+		}
+	}
+	return message
+}

vendor/github.com/goshuirc/irc-go/ircutils/userhost.go

@@ -0,0 +1,56 @@
+// written by Daniel Oaks <daniel@danieloaks.net>
+// released under the ISC license
+
+package ircutils
+
+import "strings"
+
+// UserHost holds a username+host combination
+type UserHost struct {
+	Nick string
+	User string
+	Host string
+}
+
+// ParseUserhost takes a userhost string and returns a UserHost instance.
+func ParseUserhost(userhost string) UserHost {
+	var uh UserHost
+
+	if len(userhost) == 0 {
+		return uh
+	}
+
+	if strings.Contains(userhost, "!") {
+		usersplit := strings.SplitN(userhost, "!", 2)
+		var rest string
+		if len(usersplit) == 2 {
+			uh.Nick = usersplit[0]
+			rest = usersplit[1]
+		} else {
+			rest = usersplit[0]
+		}
+
+		hostsplit := strings.SplitN(rest, "@", 2)
+		if len(hostsplit) == 2 {
+			uh.User = hostsplit[0]
+			uh.Host = hostsplit[1]
+		} else {
+			uh.User = hostsplit[0]
+		}
+	} else {
+		hostsplit := strings.SplitN(userhost, "@", 2)
+		if len(hostsplit) == 2 {
+			uh.Nick = hostsplit[0]
+			uh.Host = hostsplit[1]
+		} else {
+			uh.User = hostsplit[0]
+		}
+	}
+
+	return uh
+}
+
+// // Canonical returns the canonical string representation of the userhost.
+// func (uh *UserHost) Canonical() string {
+// 	return ""
+// }

vendor/modules.txt

@@ -21,11 +21,12 @@ github.com/go-sql-driver/mysql
 # github.com/gorilla/websocket v1.4.2 => github.com/oragono/websocket v1.4.2-oragono1
 ## explicit
 github.com/gorilla/websocket
-# github.com/goshuirc/irc-go v0.0.0-20210311004346-ea7a188a73fe
+# github.com/goshuirc/irc-go v0.0.0-20210318074529-bdc2c2cd2fef
 ## explicit
 github.com/goshuirc/irc-go/ircfmt
 github.com/goshuirc/irc-go/ircmsg
 github.com/goshuirc/irc-go/ircreader
+github.com/goshuirc/irc-go/ircutils
 # github.com/onsi/ginkgo v1.12.0
 ## explicit
 # github.com/onsi/gomega v1.9.0