Merge pull request #2159 from ergochat/casefolding.2

fix #2099

default.yaml

@@ -134,9 +134,10 @@ server:
     # the recommended default is 'ascii' (traditional ASCII-only identifiers).
     # the other options are 'precis', which allows UTF8 identifiers that are "sane"
     # (according to UFC 8265), with additional mitigations for homoglyph attacks,
-    # and 'permissive', which allows identifiers containing unusual characters like
+    # 'permissive', which allows identifiers containing unusual characters like
     # emoji, at the cost of increased vulnerability to homoglyph attacks and potential
-    # client compatibility problems. we recommend leaving this value at its default;
+    # client compatibility problems, and the legacy mappings 'rfc1459' and
+    # 'rfc1459-strict'. we recommend leaving this value at its default;
     # however, note that changing it once the network is already up and running is
     # problematic.
     casemapping: "ascii"

irc/config.go

@@ -453,6 +453,10 @@ func (cm *Casemapping) UnmarshalYAML(unmarshal func(interface{}) error) (err err
 		result = CasemappingPRECIS
 	case "permissive", "fun":
 		result = CasemappingPermissive
+	case "rfc1459":
+		result = CasemappingRFC1459
+	case "rfc1459-strict":
+		result = CasemappingRFC1459Strict
 	default:
 		return fmt.Errorf("invalid casemapping value: %s", orig)
 	}
@@ -1612,7 +1616,16 @@ func (config *Config) generateISupport() (err error) {
 	isupport.Initialize()
 	isupport.Add("AWAYLEN", strconv.Itoa(config.Limits.AwayLen))
 	isupport.Add("BOT", "B")
-	isupport.Add("CASEMAPPING", "ascii")
+	var casemappingToken string
+	switch config.Server.Casemapping {
+	default:
+		casemappingToken = "ascii" // this is published for ascii, precis, or permissive
+	case CasemappingRFC1459:
+		casemappingToken = "rfc1459"
+	case CasemappingRFC1459Strict:
+		casemappingToken = "rfc1459-strict"
+	}
+	isupport.Add("CASEMAPPING", casemappingToken)
 	isupport.Add("CHANLIMIT", fmt.Sprintf("%s:%d", chanTypes, config.Channels.MaxChannelsPerClient))
 	isupport.Add("CHANMODES", chanmodesToken)
 	if config.History.Enabled && config.History.ChathistoryMax > 0 {

irc/strings.go

@@ -60,6 +60,10 @@ const (
 	// confusables detection: standard skeleton algorithm (which may be ineffective
 	// over the larger set of permitted identifiers)
 	CasemappingPermissive
+	// rfc1459 is a legacy mapping as defined here: https://modern.ircdocs.horse/#casemapping-parameter
+	CasemappingRFC1459
+	// rfc1459-strict is a legacy mapping as defined here: https://modern.ircdocs.horse/#casemapping-parameter
+	CasemappingRFC1459Strict
 )
 
 // XXX this is a global variable without explicit synchronization.
@@ -110,6 +114,10 @@ func casefoldWithSetting(str string, setting Casemapping) (string, error) {
 		return foldASCII(str)
 	case CasemappingPermissive:
 		return foldPermissive(str)
+	case CasemappingRFC1459:
+		return foldRFC1459(str, false)
+	case CasemappingRFC1459Strict:
+		return foldRFC1459(str, true)
 	}
 }
 
@@ -214,7 +222,7 @@ func Skeleton(name string) (string, error) {
 	switch globalCasemappingSetting {
 	default:
 		return realSkeleton(name)
-	case CasemappingASCII:
+	case CasemappingASCII, CasemappingRFC1459, CasemappingRFC1459Strict:
 		// identity function is fine because we independently case-normalize in Casefold
 		return name, nil
 	}
@@ -302,6 +310,23 @@ func foldASCII(str string) (result string, err error) {
 	return strings.ToLower(str), nil
 }
 
+var (
+	rfc1459Replacer       = strings.NewReplacer("[", "{", "]", "}", "\\", "|", "~", "^")
+	rfc1459StrictReplacer = strings.NewReplacer("[", "{", "]", "}", "\\", "|")
+)
+
+func foldRFC1459(str string, strict bool) (result string, err error) {
+	asciiFold, err := foldASCII(str)
+	if err != nil {
+		return "", err
+	}
+	replacer := rfc1459Replacer
+	if strict {
+		replacer = rfc1459StrictReplacer
+	}
+	return replacer.Replace(asciiFold), nil
+}
+
 func IsPrintableASCII(str string) bool {
 	for i := 0; i < len(str); i++ {
 		// allow space here because it's technically printable;

irc/strings_test.go

@@ -279,3 +279,31 @@ func TestFoldASCIIInvalid(t *testing.T) {
 		t.Errorf("control characters should be invalid in identifiers")
 	}
 }
+
+func TestFoldRFC1459(t *testing.T) {
+	folder := func(str string) (string, error) {
+		return foldRFC1459(str, false)
+	}
+	tester := func(first, second string, equal bool) {
+		validFoldTester(first, second, equal, folder, t)
+	}
+	tester("shivaram", "SHIVARAM", true)
+	tester("shivaram[a]", "shivaram{a}", true)
+	tester("shivaram\\a]", "shivaram{a}", false)
+	tester("shivaram\\a]", "shivaram|a}", true)
+	tester("shivaram~a]", "shivaram^a}", true)
+}
+
+func TestFoldRFC1459Strict(t *testing.T) {
+	folder := func(str string) (string, error) {
+		return foldRFC1459(str, true)
+	}
+	tester := func(first, second string, equal bool) {
+		validFoldTester(first, second, equal, folder, t)
+	}
+	tester("shivaram", "SHIVARAM", true)
+	tester("shivaram[a]", "shivaram{a}", true)
+	tester("shivaram\\a]", "shivaram{a}", false)
+	tester("shivaram\\a]", "shivaram|a}", true)
+	tester("shivaram~a]", "shivaram^a}", false)
+}

traditional.yaml

@@ -108,9 +108,10 @@ server:
     # the recommended default is 'ascii' (traditional ASCII-only identifiers).
     # the other options are 'precis', which allows UTF8 identifiers that are "sane"
     # (according to UFC 8265), with additional mitigations for homoglyph attacks,
-    # and 'permissive', which allows identifiers containing unusual characters like
+    # 'permissive', which allows identifiers containing unusual characters like
     # emoji, at the cost of increased vulnerability to homoglyph attacks and potential
-    # client compatibility problems. we recommend leaving this value at its default;
+    # client compatibility problems, and the legacy mappings 'rfc1459' and
+    # 'rfc1459-strict'. we recommend leaving this value at its default;
     # however, note that changing it once the network is already up and running is
     # problematic.
     casemapping: "ascii"