//
// Check if highSurr and lowSurr are a surrogate pair then
// it checks if the combined char is in the range
// Takes in isQuery because because iri restrictions for query are different
//
internal static bool CheckIriUnicodeRange(char highSurr, char lowSurr, ref bool surrogatePair, bool isQuery)
{
bool inRange = false;
surrogatePair = false;
Debug.Assert(Char.IsHighSurrogate(highSurr));
if (Char.IsSurrogatePair(highSurr, lowSurr))
{
surrogatePair = true;
char[] chars = new char[2] {
highSurr, lowSurr
};
string surrPair = new string(chars);
if (((string.CompareOrdinal(surrPair, "\U00010000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0001FFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U00020000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0002FFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U00030000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0003FFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U00040000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0004FFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U00050000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0005FFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U00060000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0006FFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U00070000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0007FFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U00080000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0008FFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U00090000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0009FFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U000A0000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U000AFFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U000B0000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U000BFFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U000C0000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U000CFFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U000D0000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U000DFFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U000E1000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U000EFFFD") <= 0)) ||
(isQuery &&
(((string.CompareOrdinal(surrPair, "\U000F0000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U000FFFFD") <= 0)) ||
((string.CompareOrdinal(surrPair, "\U00100000") >= 0) &&
(string.CompareOrdinal(surrPair, "\U0010FFFD") <= 0)))))
{
inRange = true;
}
}
return(inRange);
}
public static int ConvertToUtf32(string s, int index)
{
CheckParameter(s, index);
if (!Char.IsSurrogate(s [index]))
{
return(s [index]);
}
if (!Char.IsHighSurrogate(s [index]) ||
index == s.Length - 1 ||
!Char.IsLowSurrogate(s [index + 1]))
{
throw new ArgumentException(String.Format("The string contains invalid surrogate pair character at {0}", index));
}
return(ConvertToUtf32(s [index], s [index + 1]));
}
public static int ConvertToUtf32(string s, int index)
{
if (s == null)
{
throw new ArgumentNullException("s");
}
if (index < 0 || index >= s.Length)
{
throw new ArgumentOutOfRangeException("index");
}
if (!Char.IsSurrogate(s [index]))
{
return(s [index]);
}
if (!Char.IsHighSurrogate(s [index]) ||
index == s.Length - 1 ||
!Char.IsLowSurrogate(s [index + 1]))
{
throw new ArgumentException(String.Format("The string contains invalid surrogate pair character at {0}", index));
}
return(ConvertToUtf32(s [index], s [index + 1]));
}
//
// Need to check for invalid utf sequences that may not have given any chars.
// We got the unescaped chars, we then reencode them and match off the bytes
// to get the invalid sequence bytes that we just copy off
//
internal static unsafe void MatchUTF8Sequence(char *pDest, char[] dest, ref int destOffset, char[] unescapedChars,
int charCount, byte[] bytes, int byteCount, bool isQuery, bool iriParsing)
{
int count = 0;
fixed(char *unescapedCharsPtr = unescapedChars)
{
for (int j = 0; j < charCount; ++j)
{
bool isHighSurr = Char.IsHighSurrogate(unescapedCharsPtr[j]);
byte[] encodedBytes = Encoding.UTF8.GetBytes(unescapedChars, j, isHighSurr ? 2 : 1);
int encodedBytesLength = encodedBytes.Length;
// we have to keep unicode chars outside Iri range escaped
bool inIriRange = false;
if (iriParsing)
{
if (!isHighSurr)
{
inIriRange = IriHelper.CheckIriUnicodeRange(unescapedChars[j], isQuery);
}
else
{
bool surrPair = false;
inIriRange = IriHelper.CheckIriUnicodeRange(unescapedChars[j], unescapedChars[j + 1],
ref surrPair, isQuery);
}
}
while (true)
{
// Escape any invalid bytes that were before this character
while (bytes[count] != encodedBytes[0])
{
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
EscapeAsciiChar((char)bytes[count++], dest, ref destOffset);
}
// check if all bytes match
bool allBytesMatch = true;
int k = 0;
for (; k < encodedBytesLength; ++k)
{
if (bytes[count + k] != encodedBytes[k])
{
allBytesMatch = false;
break;
}
}
if (allBytesMatch)
{
count += encodedBytesLength;
if (iriParsing)
{
if (!inIriRange)
{
// need to keep chars not allowed as escaped
for (int l = 0; l < encodedBytes.Length; ++l)
{
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
EscapeAsciiChar((char)encodedBytes[l], dest, ref destOffset);
}
}
else if (!Uri.IsBidiControlCharacter(unescapedCharsPtr[j]))
{
//copy chars
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = unescapedCharsPtr[j];
if (isHighSurr)
{
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = unescapedCharsPtr[j + 1];
}
}
}
else
{
//copy chars
Debug.Assert(dest.Length > destOffset);
pDest[destOffset++] = unescapedCharsPtr[j];
if (isHighSurr)
{
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = unescapedCharsPtr[j + 1];
}
}
break; // break out of while (true) since we've matched this char bytes
}
else
{
// copy bytes till place where bytes dont match
for (int l = 0; l < k; ++l)
{
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
EscapeAsciiChar((char)bytes[count++], dest, ref destOffset);
}
}
}
if (isHighSurr)
{
j++;
}
}
}
// Include any trailing invalid sequences
while (count < byteCount)
{
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
EscapeAsciiChar((char)bytes[count++], dest, ref destOffset);
}
}
public static Boolean IsHighSurrogate(this Char c)
{
return(Char.IsHighSurrogate(c));
}
/// <summary> /// Indicates whether the specified Char object is a high surrogate. /// </summary> /// <param name="Char">The Unicode character to evaluate.</param> /// <returns>true if the numeric value of the <paramref name="Char"/> parameter ranges from U+D800 through U+DBFF; otherwise, false.</returns> public static Boolean IsHighSurrogate(this Char Char) => Char.IsHighSurrogate(Char);
//
// IRI normalization for strings containing characters that are not allowed or
// escaped characters that should be unescaped in the context of the specified Uri component.
//
internal static unsafe string EscapeUnescapeIri(char *pInput, int start, int end, UriComponents component)
{
char[] dest = new char[end - start];
byte[] bytes = null;
// Pin the array to do pointer accesses
GCHandle destHandle = GCHandle.Alloc(dest, GCHandleType.Pinned);
char * pDest = (char *)destHandle.AddrOfPinnedObject();
const int percentEncodingLen = 3; // Escaped UTF-8 will take 3 chars: %AB.
const int bufferCapacityIncrease = 30 * percentEncodingLen;
int bufferRemaining = 0;
int next = start;
int destOffset = 0;
char ch;
bool escape = false;
bool surrogatePair = false;
for (; next < end; ++next)
{
escape = false;
surrogatePair = false;
if ((ch = pInput[next]) == '%')
{
if (next + 2 < end)
{
ch = UriHelper.EscapedAscii(pInput[next + 1], pInput[next + 2]);
// Do not unescape a reserved char
if (ch == Uri.c_DummyChar || ch == '%' || CheckIsReserved(ch, component) || UriHelper.IsNotSafeForUnescape(ch))
{
// keep as is
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = pInput[next++];
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = pInput[next++];
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = pInput[next];
continue;
}
else if (ch <= '\x7F')
{
Debug.Assert(ch < 0xFF, "Expecting ASCII character.");
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
//ASCII
pDest[destOffset++] = ch;
next += 2;
continue;
}
else
{
// possibly utf8 encoded sequence of unicode
// check if safe to unescape according to Iri rules
Debug.Assert(ch < 0xFF, "Expecting ASCII character.");
int startSeq = next;
int byteCount = 1;
// lazy initialization of max size, will reuse the array for next sequences
if ((object)bytes == null)
{
bytes = new byte[end - next];
}
bytes[0] = (byte)ch;
next += 3;
while (next < end)
{
// Check on exit criterion
if ((ch = pInput[next]) != '%' || next + 2 >= end)
{
break;
}
// already made sure we have 3 characters in str
ch = UriHelper.EscapedAscii(pInput[next + 1], pInput[next + 2]);
//invalid hex sequence ?
if (ch == Uri.c_DummyChar)
{
break;
}
// character is not part of a UTF-8 sequence ?
else if (ch < '\x80')
{
break;
}
else
{
//a UTF-8 sequence
bytes[byteCount++] = (byte)ch;
next += 3;
}
Debug.Assert(ch < 0xFF, "Expecting ASCII character.");
}
next--; // for loop will increment
// Using encoder with no replacement fall-back will skip all invalid UTF-8 sequences.
Encoding noFallbackCharUTF8 = (Encoding)Encoding.UTF8.Clone();
noFallbackCharUTF8.EncoderFallback = new EncoderReplacementFallback("");
noFallbackCharUTF8.DecoderFallback = new DecoderReplacementFallback("");
char[] unescapedChars = new char[bytes.Length];
int charCount = noFallbackCharUTF8.GetChars(bytes, 0, byteCount, unescapedChars, 0);
if (charCount != 0)
{
// If invalid sequences were present in the original escaped string, we need to
// copy the escaped versions of those sequences.
// Decoded Unicode values will be kept only when they are allowed by the URI/IRI RFC
// rules.
UriHelper.MatchUTF8Sequence(pDest, dest, ref destOffset, unescapedChars, charCount, bytes,
byteCount, component == UriComponents.Query, true);
}
else
{
// copy escaped sequence as is
for (int i = startSeq; i <= next; ++i)
{
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = pInput[i];
}
}
}
}
else
{
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = pInput[next];
}
}
else if (ch > '\x7f')
{
// unicode
char ch2;
if ((Char.IsHighSurrogate(ch)) && (next + 1 < end))
{
ch2 = pInput[next + 1];
escape = !CheckIriUnicodeRange(ch, ch2, ref surrogatePair, component == UriComponents.Query);
if (!escape)
{
// copy the two chars
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = pInput[next++];
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = pInput[next];
}
}
else
{
if (CheckIriUnicodeRange(ch, component == UriComponents.Query))
{
if (!Uri.IsBidiControlCharacter(ch) || !UriParser.DontKeepUnicodeBidiFormattingCharacters)
{
// copy it
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = pInput[next];
}
}
else
{
// escape it
escape = true;
}
}
}
else
{
// just copy the character
Debug.Assert(dest.Length > destOffset, "Buffer overrun detected");
pDest[destOffset++] = pInput[next];
}
if (escape)
{
const int maxNumberOfBytesEncoded = 4;
if (bufferRemaining < maxNumberOfBytesEncoded * percentEncodingLen)
{
int newBufferLength = 0;
checked
{
// may need more memory since we didn't anticipate escaping
newBufferLength = dest.Length + bufferCapacityIncrease;
bufferRemaining += bufferCapacityIncrease;
}
char[] newDest = new char[newBufferLength];
fixed(char *pNewDest = newDest)
{
#if !UT_PUBLIC_DEPENDS
Buffer.Memcpy((byte *)pNewDest, (byte *)pDest, destOffset * sizeof(char));
#else
for (int idx = 0; idx < destOffset; idx++)
{
pNewDest[idx] = pDest[idx];
}
#endif
}
if (destHandle.IsAllocated)
{
destHandle.Free();
}
dest = newDest;
// re-pin new dest[] array
destHandle = GCHandle.Alloc(dest, GCHandleType.Pinned);
pDest = (char *)destHandle.AddrOfPinnedObject();
}
byte[] encodedBytes = new byte[maxNumberOfBytesEncoded];
fixed(byte *pEncodedBytes = encodedBytes)
{
int encodedBytesCount = Encoding.UTF8.GetBytes(pInput + next, surrogatePair ? 2 : 1, pEncodedBytes, maxNumberOfBytesEncoded);
Debug.Assert(encodedBytesCount <= maxNumberOfBytesEncoded, "UTF8 encoder should not exceed specified byteCount");
bufferRemaining -= encodedBytesCount * percentEncodingLen;
for (int count = 0; count < encodedBytesCount; ++count)
{
UriHelper.EscapeAsciiChar((char)encodedBytes[count], dest, ref destOffset);
}
}
}
}
if (destHandle.IsAllocated)
{
destHandle.Free();
}
Debug.Assert(destOffset <= dest.Length, "Buffer overrun detected");
return(new string(dest, 0, destOffset));
}
