public void Ctor_UnicodeRanges()
{
// Act
var filter = new CodePointFilter(UnicodeRanges.LatinExtendedA, UnicodeRanges.LatinExtendedC);
// Assert
for (int i = 0; i < 0x0100; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x0100; i <= 0x017F; i++)
{
Assert.True(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x0180; i < 0x2C60; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x2C60; i <= 0x2C7F; i++)
{
Assert.True(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x2C80; i <= Char.MaxValue; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
}
public void Ctor_Parameterless_CreatesEmptyFilter()
{
// Act
var filter = new CodePointFilter();
// Assert
for (int i = 0; i <= Char.MaxValue; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
}
public void Ctor_OtherCodePointFilterAsInterface()
{
// Arrange
var originalFilter = new OddCodePointFilter();
// Act
var newFilter = new CodePointFilter(originalFilter);
// Assert
for (int i = 0; i <= Char.MaxValue; i++)
{
Assert.Equal((i % 2) == 1, newFilter.IsCharacterAllowed((char)i));
}
}
public void Ctor_OtherCodePointFilterAsConcreteType_Clones()
{
// Arrange
var originalFilter = new CodePointFilter().AllowChar('x');
// Act
var newFilter = new CodePointFilter(originalFilter).AllowChar('y');
// Assert
Assert.True(originalFilter.IsCharacterAllowed('x'));
Assert.False(originalFilter.IsCharacterAllowed('y'));
Assert.True(newFilter.IsCharacterAllowed('x'));
Assert.True(newFilter.IsCharacterAllowed('y'));
}
public void Ctor_WithCustomFilters()
{
// Arrange
var filter = new CodePointFilter().AllowChars("ab").AllowChars('\0', '&', '\uFFFF', 'd');
UnicodeEncoderBase encoder = new CustomUnicodeEncoderBase(filter);
// Act & assert
Assert.Equal("a", encoder.Encode("a"));
Assert.Equal("b", encoder.Encode("b"));
Assert.Equal("[U+0063]", encoder.Encode("c"));
Assert.Equal("d", encoder.Encode("d"));
Assert.Equal("[U+0000]", encoder.Encode("\0")); // we still always encode control chars
Assert.Equal("[U+0026]", encoder.Encode("&")); // we still always encode HTML-special chars
Assert.Equal("[U+FFFF]", encoder.Encode("\uFFFF")); // we still always encode non-chars and other forbidden chars
}
public void Ctor_WithCustomFilters()
{
// Arrange
var filter = new CodePointFilter().AllowChars("ab").AllowChars('\0', '&', '\uFFFF', 'd');
UnicodeEncoderBase encoder = new CustomUnicodeEncoderBase(filter);
// Act & assert
Assert.Equal("a", encoder.Encode("a"));
Assert.Equal("b", encoder.Encode("b"));
Assert.Equal("[U+0063]", encoder.Encode("c"));
Assert.Equal("d", encoder.Encode("d"));
Assert.Equal("[U+0000]", encoder.Encode("\0")); // we still always encode control chars
Assert.Equal("[U+0026]", encoder.Encode("&")); // we still always encode HTML-special chars
Assert.Equal("[U+FFFF]", encoder.Encode("\uFFFF")); // we still always encode non-chars and other forbidden chars
}
public void Ctor_WithCodePointFilter()
{
// Arrange
var filter = new CodePointFilter().AllowChars("ab").AllowChars('\0', '&', '\uFFFF', 'd');
UrlEncoder encoder = new UrlEncoder(filter);
// Act & assert
Assert.Equal("a", encoder.UrlEncode("a"));
Assert.Equal("b", encoder.UrlEncode("b"));
Assert.Equal("%63", encoder.UrlEncode("c"));
Assert.Equal("d", encoder.UrlEncode("d"));
Assert.Equal("%00", encoder.UrlEncode("\0")); // we still always encode control chars
Assert.Equal("%26", encoder.UrlEncode("&")); // we still always encode HTML-special chars
Assert.Equal("%EF%BF%BF", encoder.UrlEncode("\uFFFF")); // we still always encode non-chars and other forbidden chars
}
public void Ctor_WithCodePointFilter()
{
// Arrange
var filter = new CodePointFilter().AllowChars("ab").AllowChars('\0', '&', '\uFFFF', 'd');
HtmlEncoder encoder = new HtmlEncoder(filter);
// Act & assert
Assert.Equal("a", encoder.HtmlEncode("a"));
Assert.Equal("b", encoder.HtmlEncode("b"));
Assert.Equal("c", encoder.HtmlEncode("c"));
Assert.Equal("d", encoder.HtmlEncode("d"));
Assert.Equal("�", encoder.HtmlEncode("\0")); // we still always encode control chars
Assert.Equal("&", encoder.HtmlEncode("&")); // we still always encode HTML-special chars
Assert.Equal("", encoder.HtmlEncode("\uFFFF")); // we still always encode non-chars and other forbidden chars
}
public void AllowChars_String()
{
// Arrange
var filter = new CodePointFilter();
// Act
var retVal = filter.AllowChars("\u0100\u0102");
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
Assert.True(filter.IsCharacterAllowed('\u0100'));
Assert.False(filter.IsCharacterAllowed('\u0101'));
Assert.True(filter.IsCharacterAllowed('\u0102'));
Assert.False(filter.IsCharacterAllowed('\u0103'));
}
public void Ctor_WithCodePointFilter()
{
// Arrange
var filter = new CodePointFilter().AllowChars("ab").AllowChars('\0', '&', '\uFFFF', 'd');
JavaScriptStringEncoder encoder = new JavaScriptStringEncoder(filter);
// Act & assert
Assert.Equal("a", encoder.JavaScriptStringEncode("a"));
Assert.Equal("b", encoder.JavaScriptStringEncode("b"));
Assert.Equal(@"\u0063", encoder.JavaScriptStringEncode("c"));
Assert.Equal("d", encoder.JavaScriptStringEncode("d"));
Assert.Equal(@"\u0000", encoder.JavaScriptStringEncode("\0")); // we still always encode control chars
Assert.Equal(@"\u0026", encoder.JavaScriptStringEncode("&")); // we still always encode HTML-special chars
Assert.Equal(@"\uFFFF", encoder.JavaScriptStringEncode("\uFFFF")); // we still always encode non-chars and other forbidden chars
}
public void Ctor_WithCodePointFilter()
{
// Arrange
var filter = new CodePointFilter().AllowChars("ab").AllowChars('\0', '&', '\uFFFF', 'd');
HtmlEncoder encoder = new HtmlEncoder(filter);
// Act & assert
Assert.Equal("a", encoder.HtmlEncode("a"));
Assert.Equal("b", encoder.HtmlEncode("b"));
Assert.Equal("c", encoder.HtmlEncode("c"));
Assert.Equal("d", encoder.HtmlEncode("d"));
Assert.Equal("�", encoder.HtmlEncode("\0")); // we still always encode control chars
Assert.Equal("&", encoder.HtmlEncode("&")); // we still always encode HTML-special chars
Assert.Equal("", encoder.HtmlEncode("\uFFFF")); // we still always encode non-chars and other forbidden chars
}
public void AllowFilter()
{
// Arrange
var filter = new CodePointFilter(UnicodeRanges.BasicLatin);
// Act
var retVal = filter.AllowFilter(new OddCodePointFilter());
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
for (int i = 0; i <= 0x007F; i++)
{
Assert.True(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x0080; i <= Char.MaxValue; i++)
{
Assert.Equal((i % 2) == 1, filter.IsCharacterAllowed((char)i));
}
}
public void ForbidRange()
{
// Arrange
var filter = new CodePointFilter(new OddCodePointFilter());
// Act
var retVal = filter.ForbidRange(UnicodeRanges.Specials);
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
for (int i = 0; i <= 0xFFEF; i++)
{
Assert.Equal((i % 2) == 1, filter.IsCharacterAllowed((char)i));
}
for (int i = 0xFFF0; i <= Char.MaxValue; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
}
/// <summary>
/// Instantiates the filter by cloning the allow list of another <see cref="ICodePointFilter"/>.
/// </summary>
public CodePointFilter(ICodePointFilter other)
{
if (other == null)
{
throw new ArgumentNullException(nameof(other));
}
CodePointFilter otherAsCodePointFilter = other as CodePointFilter;
if (otherAsCodePointFilter != null)
{
_allowedCharsBitmap = otherAsCodePointFilter.GetAllowedCharsBitmap();
}
else
{
_allowedCharsBitmap = AllowedCharsBitmap.CreateNew();
AllowFilter(other);
}
}
public void GetAllowedCodePoints()
{
// Arrange
var expected = Enumerable.Range(UnicodeRanges.BasicLatin.FirstCodePoint, UnicodeRanges.BasicLatin.RangeSize)
.Concat(Enumerable.Range(UnicodeRanges.Specials.FirstCodePoint, UnicodeRanges.Specials.RangeSize))
.Except(new int[] { 'x' })
.OrderBy(i => i)
.ToArray();
var filter = new CodePointFilter(UnicodeRanges.BasicLatin, UnicodeRanges.Specials);
filter.ForbidChar('x');
// Act
var retVal = filter.GetAllowedCodePoints().OrderBy(i => i).ToArray();
// Assert
Assert.Equal <int>(expected, retVal);
}
public void Clear()
{
// Arrange
var filter = new CodePointFilter();
for (int i = 1; i <= Char.MaxValue; i++)
{
filter.AllowChar((char)i);
}
// Act
var retVal = filter.Clear();
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
for (int i = 0; i <= Char.MaxValue; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
}
/// <summary>
/// Instantiates an encoder using a custom allow list of characters.
/// </summary>
protected UnicodeEncoderBase(CodePointFilter filter, int maxOutputCharsPerInputChar)
{
_maxOutputCharsPerInputChar = maxOutputCharsPerInputChar;
_allowedCharsBitmap = filter.GetAllowedCharsBitmap();
// Forbid characters that are special in HTML.
// Even though this is a common encoder used by everybody (including URL
// and JavaScript strings), it's unfortunately common for developers to
// forget to HTML-encode a string once it has been URL-encoded or
// JavaScript string-escaped, so this offers extra protection.
ForbidCharacter('<');
ForbidCharacter('>');
ForbidCharacter('&');
ForbidCharacter('\''); // can be used to escape attributes
ForbidCharacter('\"'); // can be used to escape attributes
ForbidCharacter('+'); // technically not HTML-specific, but can be used to perform UTF7-based attacks
// Forbid codepoints which aren't mapped to characters or which are otherwise always disallowed
// (includes categories Cc, Cs, Co, Cn, Zs [except U+0020 SPACE], Zl, Zp)
_allowedCharsBitmap.ForbidUndefinedCharacters();
}
/// <summary>
/// Instantiates an encoder using a custom allow list of characters.
/// </summary>
protected UnicodeEncoderBase(CodePointFilter filter, int maxOutputCharsPerInputChar)
{
_maxOutputCharsPerInputChar = maxOutputCharsPerInputChar;
_allowedCharsBitmap = filter.GetAllowedCharsBitmap();
// Forbid characters that are special in HTML.
// Even though this is a common encoder used by everybody (including URL
// and JavaScript strings), it's unfortunately common for developers to
// forget to HTML-encode a string once it has been URL-encoded or
// JavaScript string-escaped, so this offers extra protection.
ForbidCharacter('<');
ForbidCharacter('>');
ForbidCharacter('&');
ForbidCharacter('\''); // can be used to escape attributes
ForbidCharacter('\"'); // can be used to escape attributes
ForbidCharacter('+'); // technically not HTML-specific, but can be used to perform UTF7-based attacks
// Forbid codepoints which aren't mapped to characters or which are otherwise always disallowed
// (includes categories Cc, Cs, Co, Cn, Zs [except U+0020 SPACE], Zl, Zp)
_allowedCharsBitmap.ForbidUndefinedCharacters();
}
public void AllowRange()
{
// Arrange
var filter = new CodePointFilter();
// Act
var retVal = filter.AllowRange(UnicodeRanges.LatinExtendedA);
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
for (int i = 0; i < 0x0100; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x0100; i <= 0x017F; i++)
{
Assert.True(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x0180; i <= Char.MaxValue; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
}
// We pass a (known bad) value of 1 for 'max output chars per input char',
// which also tests that the code behaves properly even if the original
// estimate is incorrect.
public CustomUnicodeEncoderBase(CodePointFilter filter)
: base(filter, maxOutputCharsPerInputChar: 1)
{
}
public void AllowRanges()
{
// Arrange
var filter = new CodePointFilter();
// Act
var retVal = filter.AllowRanges(UnicodeRanges.LatinExtendedA, UnicodeRanges.LatinExtendedC);
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
for (int i = 0; i < 0x0100; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x0100; i <= 0x017F; i++)
{
Assert.True(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x0180; i < 0x2C60; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x2C60; i <= 0x2C7F; i++)
{
Assert.True(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x2C80; i <= Char.MaxValue; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
}
public void Clear()
{
// Arrange
var filter = new CodePointFilter();
for (int i = 1; i <= Char.MaxValue; i++)
{
filter.AllowChar((char)i);
}
// Act
var retVal = filter.Clear();
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
for (int i = 0; i <= Char.MaxValue; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
}
public void ForbidRanges()
{
// Arrange
var filter = new CodePointFilter(new OddCodePointFilter());
// Act
var retVal = filter.ForbidRanges(UnicodeRanges.BasicLatin, UnicodeRanges.Specials);
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
for (int i = 0; i <= 0x007F; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
for (int i = 0x0080; i <= 0xFFEF; i++)
{
Assert.Equal((i % 2) == 1, filter.IsCharacterAllowed((char)i));
}
for (int i = 0xFFF0; i <= Char.MaxValue; i++)
{
Assert.False(filter.IsCharacterAllowed((char)i));
}
}
public void ForbidChars_String()
{
// Arrange
var filter = new CodePointFilter(UnicodeRanges.BasicLatin);
// Act
var retVal = filter.ForbidChars("xz");
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
Assert.True(filter.IsCharacterAllowed('w'));
Assert.False(filter.IsCharacterAllowed('x'));
Assert.True(filter.IsCharacterAllowed('y'));
Assert.False(filter.IsCharacterAllowed('z'));
}
// We pass a (known bad) value of 1 for 'max output chars per input char',
// which also tests that the code behaves properly even if the original
// estimate is incorrect.
public CustomUnicodeEncoderBase(CodePointFilter filter)
: base(filter, maxOutputCharsPerInputChar: 1)
{
}
public void GetAllowedCodePoints()
{
// Arrange
var expected = Enumerable.Range(UnicodeRanges.BasicLatin.FirstCodePoint, UnicodeRanges.BasicLatin.RangeSize)
.Concat(Enumerable.Range(UnicodeRanges.Specials.FirstCodePoint, UnicodeRanges.Specials.RangeSize))
.Except(new int[] { 'x' })
.OrderBy(i => i)
.ToArray();
var filter = new CodePointFilter(UnicodeRanges.BasicLatin, UnicodeRanges.Specials);
filter.ForbidChar('x');
// Act
var retVal = filter.GetAllowedCodePoints().OrderBy(i => i).ToArray();
// Assert
Assert.Equal<int>(expected, retVal);
}
internal HtmlUnicodeEncoder(CodePointFilter filter)
: base(filter, MaxOutputCharsPerInputChar)
{
}
internal JavaScriptStringUnicodeEncoder(CodePointFilter filter)
: base(filter, MaxOutputCharsPerInputChar)
{
// The only interesting characters above and beyond what the base encoder
// already covers are the solidus and reverse solidus.
ForbidCharacter('\\');
ForbidCharacter('/');
}
public void AllowChar()
{
// Arrange
var filter = new CodePointFilter();
// Act
var retVal = filter.AllowChar('\u0100');
// Assert
Assert.Same(filter, retVal); // returns 'this' instance
Assert.True(filter.IsCharacterAllowed('\u0100'));
Assert.False(filter.IsCharacterAllowed('\u0101'));
}
internal HtmlUnicodeEncoder(CodePointFilter filter)
: base(filter, MaxOutputCharsPerInputChar)
{
}
internal UrlUnicodeEncoder(CodePointFilter filter)
: base(filter, MaxOutputCharsPerInputChar)
{
// Per RFC 3987, Sec. 2.2, we want encodings that are safe for
// four particular components: 'isegment', 'ipath-noscheme',
// 'iquery', and 'ifragment'. The relevant definitions are below.
//
// ipath-noscheme = isegment-nz-nc *( "/" isegment )
//
// isegment = *ipchar
//
// isegment-nz-nc = 1*( iunreserved / pct-encoded / sub-delims
// / "@" )
// ; non-zero-length segment without any colon ":"
//
// ipchar = iunreserved / pct-encoded / sub-delims / ":"
// / "@"
//
// iquery = *( ipchar / iprivate / "/" / "?" )
//
// ifragment = *( ipchar / "/" / "?" )
//
// iunreserved = ALPHA / DIGIT / "-" / "." / "_" / "~" / ucschar
//
// ucschar = %xA0-D7FF / %xF900-FDCF / %xFDF0-FFEF
// / %x10000-1FFFD / %x20000-2FFFD / %x30000-3FFFD
// / %x40000-4FFFD / %x50000-5FFFD / %x60000-6FFFD
// / %x70000-7FFFD / %x80000-8FFFD / %x90000-9FFFD
// / %xA0000-AFFFD / %xB0000-BFFFD / %xC0000-CFFFD
// / %xD0000-DFFFD / %xE1000-EFFFD
//
// pct-encoded = "%" HEXDIG HEXDIG
//
// sub-delims = "!" / "$" / "&" / "'" / "(" / ")"
// / "*" / "+" / "," / ";" / "="
//
// The only common characters between these four components are the
// intersection of 'isegment-nz-nc' and 'ipchar', which is really
// just 'isegment-nz-nc' (colons forbidden).
//
// From this list, the base encoder already forbids "&", "'", "+",
// and we'll additionally forbid "=" since it has special meaning
// in x-www-form-urlencoded representations.
//
// This means that the full list of allowed characters from the
// Basic Latin set is:
// ALPHA / DIGIT / "-" / "." / "_" / "~" / "!" / "$" / "(" / ")" / "*" / "," / ";" / "@"
const string forbiddenChars = @" #%/:=?[\]^`{|}"; // chars from Basic Latin which aren't already disallowed by the base encoder
foreach (char c in forbiddenChars)
{
ForbidCharacter(c);
}
// Specials (U+FFF0 .. U+FFFF) are forbidden by the definition of 'ucschar' above
for (int i = 0; i < 16; i++)
{
ForbidCharacter((char)(0xFFF0 | i));
}
// Supplementary characters are forbidden anyway by the base encoder
}