ircutils: add SanitizeText

ircutils/unicode.go

@@ -4,6 +4,8 @@
 package ircutils
 
 import (
+	"strings"
+	"unicode"
 	"unicode/utf8"
 )
 
@@ -23,3 +25,38 @@ func TruncateUTF8Safe(message string, byteLimit int) (result string) {
 	}
 	return message
 }
+
+// Sanitizes human-readable text to make it safe for IRC;
+// assumes UTF-8 and uses the replacement character where
+// applicable.
+func SanitizeText(message string, byteLimit int) (result string) {
+	var buf strings.Builder
+
+	for _, r := range message {
+		if r == '\x00' || r == '\r' {
+			continue
+		} else if r == '\n' {
+			if buf.Len()+2 <= byteLimit {
+				buf.WriteString("  ")
+				continue
+			} else {
+				break
+			}
+		} else if unicode.IsSpace(r) {
+			if buf.Len()+1 <= byteLimit {
+				buf.WriteString(" ")
+			} else {
+				break
+			}
+		} else {
+			rLen := utf8.RuneLen(r)
+			if buf.Len()+rLen <= byteLimit {
+				buf.WriteRune(r)
+			} else {
+				break
+			}
+		}
+	}
+
+	return buf.String()
+}

ircutils/unicode_test.go

@@ -31,3 +31,16 @@ func TestTruncateUTF8(t *testing.T) {
 	// shouldn't truncate the whole string
 	assertEqual(TruncateUTF8Safe("\xff\xff\xff\xff\xff\xff", 5), "\xff\xff")
 }
+
+func TestSanitize(t *testing.T) {
+	assertEqual(SanitizeText("abc", 10), "abc")
+	assertEqual(SanitizeText("abcdef", 5), "abcde")
+
+	assertEqual(SanitizeText("shivaram\x00shivaram\x00shivarampassphrase", 400), "shivaramshivaramshivarampassphrase")
+
+	assertEqual(SanitizeText("the quick brown fox\xffjumps over the lazy dog", 400), "the quick brown fox\xef\xbf\xbdjumps over the lazy dog")
+
+	// \r ignored, \n is two spaces
+	assertEqual(SanitizeText("the quick brown fox\r\njumps over the lazy dog", 400), "the quick brown fox  jumps over the lazy dog")
+	assertEqual(SanitizeText("the quick brown fox\njumps over the lazy dog", 400), "the quick brown fox  jumps over the lazy dog")
+}