// Detects a Unicode encoding
private int DetectUnicodeEncoding()
{
int mode = this.mode;
int c1 = this.stream.ReadByte();
int c2;
if (c1 < 0)
{
return(-1);
}
Utf8Reader utf8reader;
switch (mode)
{
case 0:
// UTF-8 only
utf8reader = new Utf8Reader(this.stream, this.errorThrow);
this.reader = utf8reader;
utf8reader.Unget(c1);
c1 = utf8reader.ReadChar();
if (c1 == 0xfeff && !this.dontSkipUtf8Bom)
{
// Skip BOM
c1 = utf8reader.ReadChar();
}
return(c1);
case 1:
case 3:
c2 = this.DetectUtf8OrUtf16(c1);
if (c2 >= -1)
{
return(c2);
}
break;
case 2:
case 4:
// UTF-8, UTF-16, or UTF-32
c2 = this.DetectUtf8Or16Or32(c1);
if (c2 >= -1)
{
return(c2);
}
break;
}
// Default case: assume UTF-8
utf8reader = new Utf8Reader(this.stream, this.errorThrow);
this.reader = utf8reader;
utf8reader.Unget(c1);
c1 = utf8reader.ReadChar();
if (!this.dontSkipUtf8Bom && c1 == 0xfeff)
{
// Skip BOM
c1 = utf8reader.ReadChar();
}
return(c1);
}
// Detects a Unicode encoding
private int DetectUnicodeEncoding()
{
int mode = this.mode;
int c1 = this.stream.ReadByte();
int c2;
if (c1 < 0)
{
return(-1);
}
Utf8Reader utf8reader;
if (mode == 0)
{
// UTF-8 only
utf8reader = new Utf8Reader(this.stream, this.errorThrow);
this.reader = utf8reader;
c1 = utf8reader.ReadChar();
if (c1 == 0xfeff)
{
// Skip BOM
c1 = utf8reader.ReadChar();
}
return(c1);
}
else if (mode == 1 || mode == 3)
{
c2 = this.DetectUtf8OrUtf16(c1);
if (c2 >= -1)
{
return(c2);
}
}
else if (mode == 2 || mode == 4)
{
// UTF-8, UTF-16, or UTF-32
c2 = this.DetectUtf8Or16Or32(c1);
if (c2 >= -1)
{
return(c2);
}
}
// Default case: assume UTF-8
utf8reader = new Utf8Reader(this.stream, this.errorThrow);
this.reader = utf8reader;
utf8reader.Unget(c1);
c1 = utf8reader.ReadChar();
if (!this.dontSkipUtf8Bom && c1 == 0xfeff)
{
// Skip BOM
c1 = utf8reader.ReadChar();
}
return(c1);
}
private int DetectUtf8OrUtf16(int c1)
{
int mode = this.mode;
int c2;
if (c1 == 0xff || c1 == 0xfe)
{
c2 = this.stream.ReadByte();
bool bigEndian = c1 == 0xfe;
int otherbyte = bigEndian ? 0xff : 0xfe;
if (c2 == otherbyte)
{
var newReader = new Utf16Reader(
this.stream,
bigEndian,
this.errorThrow);
this.reader = newReader;
return(newReader.ReadChar());
}
// Assume UTF-8 here, so the 0xff or 0xfe is invalid
if (this.errorThrow)
{
throw new InvalidOperationException("Invalid Unicode stream");
}
else
{
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return(0xfffd);
}
}
else if (mode == 1)
{
if (c1 >= 0x01 && c1 <= 0x7f)
{
// Nonzero ASCII character
c2 = this.stream.ReadByte();
if (c2 == 0)
{
// NZA 0, so UTF-16LE
var newReader = new Utf16Reader(
this.stream,
false,
this.errorThrow);
this.reader = newReader;
}
else
{
// NZA NZ
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
}
return(c1);
}
else if (c1 == 0)
{
// Zero
c2 = this.stream.ReadByte();
if (c2 >= 0x01 && c2 <= 0x7f)
{
// 0 NZA, so UTF-16BE
var newReader = new Utf16Reader(this.stream, true, this.errorThrow);
this.reader = newReader;
return(c2);
}
else
{
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return(c1);
}
}
}
// Use default of UTF-8
return(-2);
}
private int DetectUtf8Or16Or32(int c1)
{
int c2, c3, c4;
if (c1 == 0xff || c1 == 0xfe)
{
// Start of a possible byte-order mark
// FF FE 0 0 --> UTF-32LE
// FF FE ... --> UTF-16LE
// FE FF --> UTF-16BE
c2 = this.stream.ReadByte();
bool bigEndian = c1 == 0xfe;
int otherbyte = bigEndian ? 0xff : 0xfe;
if (c2 == otherbyte)
{
c3 = this.stream.ReadByte();
c4 = this.stream.ReadByte();
if (!bigEndian && c3 == 0 && c4 == 0)
{
this.reader = new Utf32Reader(this.stream, false, this.errorThrow);
return(this.reader.ReadChar());
}
else
{
var newReader = new Utf16Reader(
this.stream,
bigEndian,
this.errorThrow);
newReader.Unget(c3, c4);
this.reader = newReader;
return(newReader.ReadChar());
}
}
// Assume UTF-8 here, so the 0xff or 0xfe is invalid
if (this.errorThrow)
{
throw new InvalidOperationException("Invalid Unicode stream");
}
else
{
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return(0xfffd);
}
}
else if (c1 == 0 && this.mode == 4)
{
// Here, the relevant cases are:
// 0 0 0 NZA --> UTF-32BE (if mode is 4)
// 0 0 FE FF --> UTF-32BE
// Anything else is treated as UTF-8
c2 = this.stream.ReadByte();
c3 = this.stream.ReadByte();
c4 = this.stream.ReadByte();
if (c2 == 0 &&
((c3 == 0xfe && c4 == 0xff) ||
(c3 == 0 && c4 >= 0x01 && c4 <= 0x7f)))
{
this.reader = new Utf32Reader(this.stream, true, this.errorThrow);
return(c3 == 0 ? c4 : this.reader.ReadChar());
}
else
{
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.UngetThree(c2, c3, c4);
this.reader = utf8reader;
return(c1);
}
}
else if (this.mode == 2)
{
if (c1 >= 0x01 && c1 <= 0x7f)
{
// Nonzero ASCII character
c2 = this.stream.ReadByte();
if (c2 == 0)
{
// NZA 0, so UTF-16LE or UTF-32LE
c3 = this.stream.ReadByte();
c4 = this.stream.ReadByte();
if (c3 == 0 && c4 == 0)
{
this.reader = new Utf32Reader(
this.stream,
false,
this.errorThrow);
return(c1);
}
else
{
var newReader = new Utf16Reader(
this.stream,
false,
this.errorThrow);
newReader.Unget(c3, c4);
this.reader = newReader;
return(c1);
}
}
else
{
// NZA NZ, so UTF-8
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return(c1);
}
}
else if (c1 == 0)
{
// Zero
c2 = this.stream.ReadByte();
if (c2 >= 0x01 && c2 <= 0x7f)
{
// 0 NZA, so UTF-16BE
var newReader = new Utf16Reader(this.stream, true, this.errorThrow);
this.reader = newReader;
return(c2);
}
else if (c2 == 0)
{
// 0 0, so maybe UTF-32BE
c3 = this.stream.ReadByte();
c4 = this.stream.ReadByte();
if (c3 == 0 && c4 >= 0x01 && c4 <= 0x7f)
{
// 0 0 0 NZA
this.reader = new Utf32Reader(this.stream, true, this.errorThrow);
return(c4);
}
else if (c3 == 0xfe && c4 == 0xff)
{
// 0 0 FE FF
this.reader = new Utf32Reader(this.stream, true, this.errorThrow);
return(this.reader.ReadChar());
}
else
{
// 0 0 ...
var newReader = new Utf8Reader(this.stream, this.errorThrow);
newReader.UngetThree(c2, c3, c4);
this.reader = newReader;
return(c1);
}
}
else
{
// 0 NonAscii, so UTF-8
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return(c1);
}
}
}
// Use default of UTF-8
return(-2);
}
private int DetectUtf8OrUtf16(int c1)
{
int mode = this.mode;
int c2;
if (c1 == 0xff || c1 == 0xfe) {
c2 = this.stream.ReadByte();
bool bigEndian = c1 == 0xfe;
int otherbyte = bigEndian ? 0xff : 0xfe;
if (c2 == otherbyte) {
var newReader = new Utf16Reader(
this.stream,
bigEndian,
this.errorThrow);
this.reader = newReader;
return newReader.ReadChar();
}
// Assume UTF-8 here, so the 0xff or 0xfe is invalid
if (this.errorThrow) {
throw new InvalidOperationException("Invalid Unicode stream");
} else {
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return 0xfffd;
}
} else if (mode == 1) {
if (c1 >= 0x01 && c1 <= 0x7f) {
// Nonzero ASCII character
c2 = this.stream.ReadByte();
if (c2 == 0) {
// NZA 0, so UTF-16LE
var newReader = new Utf16Reader(
this.stream,
false,
this.errorThrow);
this.reader = newReader;
} else {
// NZA NZ
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
}
return c1;
} else if (c1 == 0) {
// Zero
c2 = this.stream.ReadByte();
if (c2 >= 0x01 && c2 <= 0x7f) {
// 0 NZA, so UTF-16BE
var newReader = new Utf16Reader(this.stream, true, this.errorThrow);
this.reader = newReader;
return c2;
} else {
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return c1;
}
}
}
// Use default of UTF-8
return -2;
}
private int DetectUtf8Or16Or32(int c1)
{
int c2, c3, c4;
if (c1 == 0xff || c1 == 0xfe) {
// Start of a possible byte-order mark
// FF FE 0 0 --> UTF-32LE
// FF FE ... --> UTF-16LE
// FE FF --> UTF-16BE
c2 = this.stream.ReadByte();
bool bigEndian = c1 == 0xfe;
int otherbyte = bigEndian ? 0xff : 0xfe;
if (c2 == otherbyte) {
c3 = this.stream.ReadByte();
c4 = this.stream.ReadByte();
if (!bigEndian && c3 == 0 && c4 == 0) {
this.reader = new Utf32Reader(this.stream, false, this.errorThrow);
return this.reader.ReadChar();
} else {
var newReader = new Utf16Reader(
this.stream,
bigEndian,
this.errorThrow);
newReader.Unget(c3, c4);
this.reader = newReader;
return newReader.ReadChar();
}
}
// Assume UTF-8 here, so the 0xff or 0xfe is invalid
if (this.errorThrow) {
throw new InvalidOperationException("Invalid Unicode stream");
} else {
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return 0xfffd;
}
} else if (c1 == 0 && this.mode == 4) {
// Here, the relevant cases are:
// 0 0 0 NZA --> UTF-32BE (if mode is 4)
// 0 0 FE FF --> UTF-32BE
// Anything else is treated as UTF-8
c2 = this.stream.ReadByte();
c3 = this.stream.ReadByte();
c4 = this.stream.ReadByte();
if (c2 == 0 &&
((c3 == 0xfe && c4 == 0xff) ||
(c3 == 0 && c4 >= 0x01 && c4 <= 0x7f))) {
this.reader = new Utf32Reader(this.stream, true, this.errorThrow);
return c3 == 0 ? c4 : this.reader.ReadChar();
} else {
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.UngetThree(c2, c3, c4);
this.reader = utf8reader;
return c1;
}
} else if (this.mode == 2) {
if (c1 >= 0x01 && c1 <= 0x7f) {
// Nonzero ASCII character
c2 = this.stream.ReadByte();
if (c2 == 0) {
// NZA 0, so UTF-16LE or UTF-32LE
c3 = this.stream.ReadByte();
c4 = this.stream.ReadByte();
if (c3 == 0 && c4 == 0) {
this.reader = new Utf32Reader(
this.stream,
false,
this.errorThrow);
return c1;
} else {
var newReader = new Utf16Reader(
this.stream,
false,
this.errorThrow);
newReader.Unget(c3, c4);
this.reader = newReader;
return c1;
}
} else {
// NZA NZ, so UTF-8
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return c1;
}
} else if (c1 == 0) {
// Zero
c2 = this.stream.ReadByte();
if (c2 >= 0x01 && c2 <= 0x7f) {
// 0 NZA, so UTF-16BE
var newReader = new Utf16Reader(this.stream, true, this.errorThrow);
this.reader = newReader;
return c2;
} else if (c2 == 0) {
// 0 0, so maybe UTF-32BE
c3 = this.stream.ReadByte();
c4 = this.stream.ReadByte();
if (c3 == 0 && c4 >= 0x01 && c4 <= 0x7f) {
// 0 0 0 NZA
this.reader = new Utf32Reader(this.stream, true, this.errorThrow);
return c4;
} else if (c3 == 0xfe && c4 == 0xff) {
// 0 0 FE FF
this.reader = new Utf32Reader(this.stream, true, this.errorThrow);
return this.reader.ReadChar();
} else {
// 0 0 ...
var newReader = new Utf8Reader(this.stream, this.errorThrow);
newReader.UngetThree(c2, c3, c4);
this.reader = newReader;
return c1;
}
} else {
// 0 NonAscii, so UTF-8
var utf8reader = new Utf8Reader(this.stream, this.errorThrow);
utf8reader.Unget(c2);
this.reader = utf8reader;
return c1;
}
}
}
// Use default of UTF-8
return -2;
}
// Detects a Unicode encoding
private int DetectUnicodeEncoding()
{
int mode = this.mode;
int c1 = this.stream.ReadByte();
int c2;
if (c1 < 0) {
return -1;
}
Utf8Reader utf8reader;
if (mode == 0) {
// UTF-8 only
utf8reader = new Utf8Reader(this.stream, this.errorThrow);
this.reader = utf8reader;
c1 = utf8reader.ReadChar();
if (c1 == 0xfeff) {
// Skip BOM
c1 = utf8reader.ReadChar();
}
return c1;
} else if (mode == 1 || mode == 3) {
c2 = this.DetectUtf8OrUtf16(c1);
if (c2 >= -1) {
return c2;
}
} else if (mode == 2 || mode == 4) {
// UTF-8, UTF-16, or UTF-32
c2 = this.DetectUtf8Or16Or32(c1);
if (c2 >= -1) {
return c2;
}
}
// Default case: assume UTF-8
utf8reader = new Utf8Reader(this.stream, this.errorThrow);
this.reader = utf8reader;
utf8reader.Unget(c1);
c1 = utf8reader.ReadChar();
if (!this.dontSkipUtf8Bom && c1 == 0xfeff) {
// Skip BOM
c1 = utf8reader.ReadChar();
}
return c1;
}