Input sanitisation in userpages and user settings

common/sanitisation.go

@@ -0,0 +1,16 @@
+package common
+
+import (
+	"unicode"
+)
+
+// SanitiseString removes all control codes from a string.
+func SanitiseString(s string) string {
+	n := make([]rune, 0, len(s))
+	for _, c := range s {
+		if !unicode.Is(unicode.Other, c) {
+			n = append(n, c)
+		}
+	}
+	return string(n)
+}

common/sanitisation_test.go

@@ -0,0 +1,41 @@
+package common
+
+import "testing"
+
+const pen = "I trattori di palmizio 나는 펜이있다. 私はリンゴを持っています。" +
+	"啊! 苹果笔。 у меня есть ручка, Tôi có dứa. අන්නාසි පෑන"
+
+func TestSanitiseString(t *testing.T) {
+	tests := []struct {
+		name string
+		arg  string
+		want string
+	}{
+		{
+			"Normal",
+			pen,
+			pen,
+		},
+		{
+			"Arabic (rtl)",
+			"أناناس",
+			"أناناس",
+		},
+		{
+			"Null",
+			"A\x00B",
+			"AB",
+		},
+	}
+	for _, tt := range tests {
+		if got := SanitiseString(tt.arg); got != tt.want {
+			t.Errorf("%q. SanitiseString() = %v, want %v", tt.name, got, tt.want)
+		}
+	}
+}
+
+func BenchmarkSanitiseString(b *testing.B) {
+	for i := 0; i < b.N; i++ {
+		SanitiseString(pen)
+	}
+}