public override unsafe int FindFirstCharacterToEncode(char *text, int textLength)
{
if (text == null)
{
throw new ArgumentNullException(nameof(text));
}
int idx = 0;
#if NETCOREAPP
if (Sse2.IsSupported || AdvSimd.Arm64.IsSupported)
{
short *startingAddress = (short *)text;
while (textLength - 8 >= idx)
{
Debug.Assert(startingAddress >= text && startingAddress <= (text + textLength - 8));
// Load the next 8 characters.
Vector128 <short> sourceValue;
Vector128 <short> mask;
bool containsNonAsciiChars;
if (Sse2.IsSupported)
{
sourceValue = Sse2.LoadVector128(startingAddress);
mask = Sse2Helper.CreateAsciiMask(sourceValue);
containsNonAsciiChars = Sse2Helper.ContainsNonAsciiByte(mask.AsSByte());
}
else if (AdvSimd.Arm64.IsSupported)
{
sourceValue = AdvSimd.LoadVector128(startingAddress);
mask = AdvSimdHelper.CreateAsciiMask(sourceValue);
containsNonAsciiChars = AdvSimdHelper.ContainsNonAsciiByte(mask.AsSByte());
}
else
{
throw new PlatformNotSupportedException();
}
if (containsNonAsciiChars)
{
// At least one of the following 8 characters is non-ASCII.
int processNextEight = idx + 8;
Debug.Assert(processNextEight <= textLength);
for (; idx < processNextEight; idx++)
{
Debug.Assert((text + idx) <= (text + textLength));
if (!_allowedCharacters.IsCharacterAllowed(*(text + idx)))
{
goto Return;
}
}
startingAddress += 8;
}
else
{
int index;
// Check if any of the 8 characters need to be escaped.
if (Sse2.IsSupported)
{
mask = Sse2Helper.CreateEscapingMask_UnsafeRelaxedJavaScriptEncoder(sourceValue);
index = Sse2Helper.GetIndexOfFirstNonAsciiByte(mask.AsByte());
}
else if (AdvSimd.Arm64.IsSupported)
{
mask = AdvSimdHelper.CreateEscapingMask_UnsafeRelaxedJavaScriptEncoder(sourceValue);
index = AdvSimdHelper.GetIndexOfFirstNonAsciiByte(mask.AsByte());
}
else
{
throw new PlatformNotSupportedException();
}
// If index >= 16, that means none of the 8 characters needed to be escaped.
if (index < 16)
{
// Found at least one character that needs to be escaped, figure out the index of
// the first one found that needed to be escaped within the 8 characters.
Debug.Assert(index % 2 == 0);
idx += index >> 1;
goto Return;
}
idx += 8;
startingAddress += 8;
}
}
// Process the remaining characters.
Debug.Assert(textLength - idx < 8);
}
#endif
for (; idx < textLength; idx++)
{
Debug.Assert((text + idx) <= (text + textLength));
if (!_allowedCharacters.IsCharacterAllowed(*(text + idx)))
{
goto Return;
}
}
idx = -1; // All characters are allowed.
Return:
return(idx);
}
public override unsafe int FindFirstCharacterToEncodeUtf8(ReadOnlySpan <byte> utf8Text)
{
fixed(byte *ptr = utf8Text)
{
int idx = 0;
#if NETCOREAPP
if (Sse2.IsSupported || AdvSimd.Arm64.IsSupported)
{
sbyte *startingAddress = (sbyte *)ptr;
while (utf8Text.Length - 16 >= idx)
{
Debug.Assert(startingAddress >= ptr && startingAddress <= (ptr + utf8Text.Length - 16));
// Load the next 16 bytes.
Vector128 <sbyte> sourceValue;
bool containsNonAsciiBytes;
// Check for ASCII text. Any byte that's not in the ASCII range will already be negative when
// casted to signed byte.
if (Sse2.IsSupported)
{
sourceValue = Sse2.LoadVector128(startingAddress);
containsNonAsciiBytes = Sse2Helper.ContainsNonAsciiByte(sourceValue);
}
else
{
sourceValue = AdvSimd.LoadVector128(startingAddress);
containsNonAsciiBytes = AdvSimdHelper.ContainsNonAsciiByte(sourceValue);
}
if (containsNonAsciiBytes)
{
// At least one of the following 16 bytes is non-ASCII.
int processNextSixteen = idx + 16;
Debug.Assert(processNextSixteen <= utf8Text.Length);
while (idx < processNextSixteen)
{
Debug.Assert((ptr + idx) <= (ptr + utf8Text.Length));
if (UnicodeUtility.IsAsciiCodePoint(ptr[idx]))
{
if (!_allowedCharacters.IsUnicodeScalarAllowed(ptr[idx]))
{
goto Return;
}
idx++;
}
else
{
OperationStatus opStatus = UnicodeHelpers.DecodeScalarValueFromUtf8(utf8Text.Slice(idx), out uint nextScalarValue, out int utf8BytesConsumedForScalar);
Debug.Assert(nextScalarValue <= int.MaxValue);
if (opStatus != OperationStatus.Done || WillEncode((int)nextScalarValue))
{
goto Return;
}
Debug.Assert(opStatus == OperationStatus.Done);
idx += utf8BytesConsumedForScalar;
}
}
startingAddress = (sbyte *)ptr + idx;
}
else
{
// Check if any of the 16 bytes need to be escaped.
int index;
if (Sse2.IsSupported)
{
Vector128 <sbyte> mask = Sse2Helper.CreateEscapingMask_UnsafeRelaxedJavaScriptEncoder(sourceValue);
index = Sse2Helper.GetIndexOfFirstNonAsciiByte(mask.AsByte());
}
else
{
Vector128 <sbyte> mask = AdvSimdHelper.CreateEscapingMask_UnsafeRelaxedJavaScriptEncoder(sourceValue);
index = AdvSimdHelper.GetIndexOfFirstNonAsciiByte(mask.AsByte());
}
// If index >= 16, that means none of the 16 bytes needed to be escaped.
if (index < 16)
{
// Found at least one byte that needs to be escaped, figure out the index of
// the first one found that needed to be escaped within the 16 bytes.
idx += index;
goto Return;
}
idx += 16;
startingAddress += 16;
}
}
// Process the remaining bytes.
Debug.Assert(utf8Text.Length - idx < 16);
}
#endif
while (idx < utf8Text.Length)
{
Debug.Assert((ptr + idx) <= (ptr + utf8Text.Length));
if (UnicodeUtility.IsAsciiCodePoint(ptr[idx]))
{
if (!_allowedCharacters.IsUnicodeScalarAllowed(ptr[idx]))
{
goto Return;
}
idx++;
}
else
{
OperationStatus opStatus = UnicodeHelpers.DecodeScalarValueFromUtf8(utf8Text.Slice(idx), out uint nextScalarValue, out int utf8BytesConsumedForScalar);
Debug.Assert(nextScalarValue <= int.MaxValue);
if (opStatus != OperationStatus.Done || WillEncode((int)nextScalarValue))
{
goto Return;
}
Debug.Assert(opStatus == OperationStatus.Done);
idx += utf8BytesConsumedForScalar;
}
}
Debug.Assert(idx == utf8Text.Length);
idx = -1; // All bytes are allowed.
Return:
return(idx);
}
}
