public void UrlEncode_BadSurrogates_ReturnsUnicodeReplacementChar()
{
// Arrange
UrlEncoder encoder = UrlEncoder.Create(UnicodeRanges.All); // allow all codepoints
// "a<unpaired leading>b<unpaired trailing>c<trailing before leading>d<unpaired trailing><valid>e<high at end of string>"
const string input = "a\uD800b\uDFFFc\uDFFF\uD800d\uDFFF\uD800\uDFFFe\uD800";
const string expected = "a%EF%BF%BDb%EF%BF%BDc%EF%BF%BD%EF%BF%BDd%EF%BF%BD%F0%90%8F%BFe%EF%BF%BD"; // 'D800' 'DFFF' was preserved since it's valid
// Act
string retVal = encoder.Encode(input);
// Assert
Assert.Equal(expected, retVal);
}
public void Default_EquivalentToBasicLatin()
{
// Arrange
UrlEncoder controlEncoder = UrlEncoder.Create(UnicodeRanges.BasicLatin);
UrlEncoder testEncoder = UrlEncoder.Default;
// Act & assert
for (int i = 0; i <= char.MaxValue; i++)
{
if (!IsSurrogateCodePoint(i))
{
string input = new string((char)i, 1);
Assert.Equal(controlEncoder.Encode(input), testEncoder.Encode(input));
}
}
}
public void Ctor_WithTextEncoderSettings()
{
// Arrange
var filter = new TextEncoderSettings();
filter.AllowCharacters('a', 'b');
filter.AllowCharacters('\0', '&', '\uFFFF', 'd');
UrlEncoder encoder = UrlEncoder.Create(filter);
// Act & assert
Assert.Equal("a", encoder.Encode("a"));
Assert.Equal("b", encoder.Encode("b"));
Assert.Equal("%63", encoder.Encode("c"));
Assert.Equal("d", encoder.Encode("d"));
Assert.Equal("%00", encoder.Encode("\0")); // we still always encode control chars
Assert.Equal("%26", encoder.Encode("&")); // we still always encode HTML-special chars
Assert.Equal("%EF%BF%BF", encoder.Encode("\uFFFF")); // we still always encode non-chars and other forbidden chars
}
public void UrlEncode_DoesNotOutputHtmlSensitiveCharacters()
{
// Per the design document, we provide additional defense-in-depth
// by never emitting HTML-sensitive characters unescaped.
// Arrange
UrlEncoder urlEncoder = UrlEncoder.Create(UnicodeRanges.All);
HtmlEncoder htmlEncoder = HtmlEncoder.Create(UnicodeRanges.All);
// Act & assert
for (int i = 0; i <= 0x10FFFF; i++)
{
if (IsSurrogateCodePoint(i))
{
continue; // surrogates don't matter here
}
string urlEncoded = urlEncoder.Encode(char.ConvertFromUtf32(i));
string thenHtmlEncoded = htmlEncoder.Encode(urlEncoded);
Assert.Equal(urlEncoded, thenHtmlEncoded); // should have contained no HTML-sensitive characters
}
}
public void UrlEncode_AllRangesAllowed_StillEncodesForbiddenChars()
{
// Arrange
UrlEncoder encoder = UrlEncoder.Create(UnicodeRanges.All);
// Act & assert - BMP chars
for (int i = 0; i <= 0xFFFF; i++)
{
string input = new string((char)i, 1);
string expected;
if (IsSurrogateCodePoint(i))
{
expected = "%EF%BF%BD"; // unpaired surrogate -> Unicode replacement char
}
else
{
bool mustEncode = true;
// RFC 3987, Sec. 2.2 gives the list of allowed chars
// (We allow 'ipchar' except for "'", "&", "+", "%", and "="
if (('a' <= i && i <= 'z') || ('A' <= i && i <= 'Z') || ('0' <= i && i <= '9'))
{
mustEncode = false; // ALPHA / DIGIT
}
else if ((0x00A0 <= i && i <= 0xD7FF) | (0xF900 <= i && i <= 0xFDCF) | (0xFDF0 <= i && i <= 0xFFEF))
{
mustEncode = !UnicodeTestHelpers.IsCharacterDefined((char)i); // 'ucschar'
}
else
{
switch (i)
{
// iunreserved
case '-':
case '.':
case '_':
case '~':
// isegment-nz-nc
case '@':
// sub-delims
case '!':
case '$':
case '(':
case ')':
case '*':
case ',':
case ';':
mustEncode = false;
break;
}
}
if (mustEncode)
{
expected = GetKnownGoodPercentEncodedValue(i);
}
else
{
expected = input; // no encoding
}
}
string retVal = encoder.Encode(input);
Assert.Equal(expected, retVal);
}
// Act & assert - astral chars
for (int i = 0x10000; i <= 0x10FFFF; i++)
{
string input = char.ConvertFromUtf32(i);
string expected = GetKnownGoodPercentEncodedValue(i);
string retVal = encoder.Encode(input);
Assert.Equal(expected, retVal);
}
}
