private static Token ParseWhitespace(TokenizerContext context)
{
var originalPos = context.Position;
var len = context.File.Content.Length;
var position = context.CreatePosition();
while (context.Position < len) {
var ch = context.File.Content[context.Position++];
if (ch == '\r' || ch == '\n') {
if (context.Position < len) {
var nextCh = context.File.Content[context.Position];
if ((nextCh == '\r' || nextCh == '\n') && nextCh != ch) {
context.Position++;
}
}
context.IncrementLine();
} else if (ch == ' ' || ch == '\t') {
} else {
context.Position--;
break;
}
}
var res = new Token {
Type = TokenType.Whitespace,
StringValue = context.File.Content.Substring(originalPos, context.Position - originalPos),
Position = position
};
return res;
}
private static Token ParseToken(TokenizerContext context)
{
var position = context.CreatePosition();
var len = context.File.Content.Length;
var ch = context.File.Content[context.Position];
var preview = (char)(0xffff);
if (context.Position + 1 < len) {
preview = context.File.Content[context.Position + 1];
}
if (ch == '>' && preview == '>') {
context.Position++;
return new Token {
Type = TokenType.RightShift,
StringValue = ">",
Position = position
};
}
if (ch == '<' && preview == '<') {
context.Position++;
return new Token {
Type = TokenType.LeftShift,
StringValue = "<",
Position = position
};
}
if (ch == '/' && preview == '/') {
return ParseSingleLineComment(context);
}
if (ch == '/' && preview == '*') {
return ParseMultiLineComment(context);
}
if (ch == '.' && (preview >= '0' && preview <= '9')) {
return ParseNumberToken(context);
}
switch (ch) {
case '\r':
case '\n':
case '\t':
case ' ':
return ParseWhitespace(context);
case '"':
case '\'':
return ParseString(context);
case '-':
var literal = ParseLiteral(context);
if (literal.StringValue != "-") return literal;
return new Token {
Position = literal.Position,
Type = TokenType.Minus,
StringValue = literal.StringValue
};
default:
if (IsPunctuation(ch)) {
context.Position++;
return new Token {
Type = GetPunctuationTokenType(ch),
StringValue = ch.ToString(),
Position = position
};
}
if (char.IsDigit(ch)) {
return ParseNumberToken(context);
}
return ParseLiteral(context);
}
}
private static Token ParseString(TokenizerContext context)
{
var quoteCh = context.File.Content[context.Position++];
var token = new StringBuilder();
var len = context.File.Content.Length;
var position = context.CreatePosition();
while (context.Position < len) {
var ch = context.File.Content[context.Position++];
if (ch == '\r' || ch == '\n') {
throw new TokenException("missing end quote", new Token { Position = position });
}
if (ch == '\\') {
if (context.Position >= len) throw new TokenException("missing end quote", new Token { Position = position });
ch = context.File.Content[context.Position++];
switch (ch) {
case 'n':
token.Append('\n');
break;
case 'r':
token.Append('\r');
break;
case 'a':
token.Append('\a');
break;
case 'b':
token.Append('\b');
break;
case 'f':
token.Append('\f');
break;
case 't':
token.Append('\t');
break;
case 'v':
token.Append('\v');
break;
case '\\':
token.Append('\\');
break;
case '\'':
token.Append('\'');
break;
case '\"':
token.Append('\"');
break;
case '0':
token.Append('\0');
break;
default:
throw new TokenException("invalid escape character " + ch, new Token { Position = position });
}
} else if (ch == quoteCh) {
if (context.Position >= len)
return new Token {
Position = position,
Type = TokenType.String,
StringValue = token.ToString()
};
var next = context.File.Content[context.Position];
if (next == quoteCh) {
token.Append(ch);
context.Position++;
} else {
return new Token {
Position = position,
Type = TokenType.String,
StringValue = token.ToString()
};
}
} else {
token.Append(ch);
}
}
throw new TokenException("missing end quote", new Token { Position = position });
}
private static Token ParseNumberToken(TokenizerContext context)
{
var originalPosition = context.Position;
var len = context.File.Content.Length;
var position = context.CreatePosition();
var val = 0.0;
var dotExp = 0;
while (context.Position < len) {
var ch = context.File.Content[context.Position++];
if (ch >= '0' && ch <= '9') {
val = val * 10 + (ch - '0');
if (dotExp > 0) dotExp *= 10;
} else if (dotExp == 0 && ch == '.') {
dotExp = 1;
} else {
context.Position--;
break;
}
}
if (dotExp > 0) val /= dotExp;
if (context.Position + 1 < len) {
var expPrev = context.File.Content[context.Position];
var postExpCh = context.File.Content[context.Position + 1];
if (expPrev == 'e' || expPrev == 'E') {
if (postExpCh == '+' || postExpCh == '-' || (postExpCh >= '0' && postExpCh <= '9')) {
context.Position++;
if (postExpCh == '+' || postExpCh == '-') context.Position++;
var exp = 0.0;
var expChars = 0;
while (context.Position < len) {
var ch = context.File.Content[context.Position];
if (ch >= '0' && ch <= '9') {
context.Position++;
exp = exp * 10 + (ch - '0');
expChars++;
} else {
break;
}
}
if (expChars == 0) {
throw new TokenException("exponent value expected", new Token {
Position = position,
Type = TokenType.Number
});
}
if (postExpCh == '-') exp = -exp;
val = val * Math.Pow(10, exp);
}
}
}
return new Token {
Type = TokenType.Number,
Position = position,
StringValue = context.File.Content.Substring(originalPosition, context.Position - originalPosition),
NumberValue = val
};
}
private static Token ParseMultiLineComment(TokenizerContext context)
{
var originalPos = context.Position;
var len = context.File.Content.Length;
var position = context.CreatePosition();
while (context.Position < len - 1) {
var ch = context.File.Content[context.Position++];
if (ch == '\r' || ch == '\n') {
if (context.Position < len) {
var nextCh = context.File.Content[context.Position];
if ((nextCh == '\r' || nextCh == '\n') && nextCh != ch) {
context.Position++;
}
}
context.IncrementLine();
} else if (ch == '*' && context.File.Content[context.Position] == '/') {
context.Position++;
return new Token {
Type = TokenType.MultiLineComment,
StringValue = context.File.Content.Substring(originalPos, context.Position - originalPos),
Position = position
};
}
}
throw new TokenException("unclosed multiline comment", new Token {
Type = TokenType.MultiLineComment,
StringValue = context.File.Content.Substring(originalPos, context.Position - originalPos),
Position = position
});
}
private static Token ParseLiteral(TokenizerContext context)
{
var originalPosition = context.Position;
var len = context.File.Content.Length;
var position = context.CreatePosition();
if (context.Position < len && context.File.Content[context.Position] == '-') {
context.Position++;
}
if (context.Position < len) {
var ch = context.File.Content[context.Position];
if ((ch >= 'a' && ch <= 'z') || (ch >= 'A' && ch <= 'Z') || ch == '_') {
context.Position++;
while (context.Position < len) {
ch = context.File.Content[context.Position];
if ((ch >= 'a' && ch <= 'z') || (ch >= 'A' && ch <= 'Z') || (ch >= '0' && ch <= '9') || ch == '_' || ch == '-') {
context.Position++;
} else {
break;
}
}
}
}
return new Token {
Position = position,
Type = TokenType.Literal,
StringValue = context.File.Content.Substring(originalPosition, context.Position - originalPosition)
};
}