public Scanner()
{
Regex regex;
Patterns = new Dictionary<TokenType, Regex>();
Tokens = new List<TokenType>();
LookAheadToken = null;
Skipped = new List<Token>();
SkipList = new List<TokenType>();
SkipList.Add(TokenType.WHITESPACE);
regex = new Regex(@"[\s\t\r\n]+", RegexOptions.Compiled);
Patterns.Add(TokenType.WHITESPACE, regex);
Tokens.Add(TokenType.WHITESPACE);
regex = new Regex(@"[0-9]+", RegexOptions.Compiled);
Patterns.Add(TokenType.NUMBER, regex);
Tokens.Add(TokenType.NUMBER);
regex = new Regex(@"[a-zA-Z0-9_']+", RegexOptions.Compiled);
Patterns.Add(TokenType.IDENTIFIER, regex);
Tokens.Add(TokenType.IDENTIFIER);
regex = new Regex(@"^$", RegexOptions.Compiled);
Patterns.Add(TokenType.EOF, regex);
Tokens.Add(TokenType.EOF);
regex = new Regex(@"parity", RegexOptions.Compiled);
Patterns.Add(TokenType.PARITY, regex);
Tokens.Add(TokenType.PARITY);
regex = new Regex(@"[01]", RegexOptions.Compiled);
Patterns.Add(TokenType.OWNERSTR, regex);
Tokens.Add(TokenType.OWNERSTR);
regex = new Regex(@",", RegexOptions.Compiled);
Patterns.Add(TokenType.COMMA, regex);
Tokens.Add(TokenType.COMMA);
regex = new Regex(";", RegexOptions.Compiled);
Patterns.Add(TokenType.SEMICOLON, regex);
Tokens.Add(TokenType.SEMICOLON);
regex = new Regex(@"""[^""]*""", RegexOptions.Compiled);
Patterns.Add(TokenType.STRING, regex);
Tokens.Add(TokenType.STRING);
}
public void UpdateRange(Token token)
{
if (token.StartPos < startpos) startpos = token.StartPos;
if (token.EndPos > endpos) endpos = token.EndPos;
}
/// <summary>
/// executes a lookahead of the next token
/// and will advance the scan on the input string
/// </summary>
/// <returns></returns>
public Token Scan(params TokenType[] expectedtokens)
{
Token tok = LookAhead(expectedtokens); // temporarely retrieve the lookahead
LookAheadToken = null; // reset lookahead token, so scanning will continue
StartPos = tok.EndPos;
EndPos = tok.EndPos; // set the tokenizer to the new scan position
return tok;
}
/// <summary>
/// returns token with longest best match
/// </summary>
/// <returns></returns>
public Token LookAhead(params TokenType[] expectedtokens)
{
int i;
int startpos = StartPos;
Token tok = null;
List<TokenType> scantokens;
// this prevents double scanning and matching
// increased performance
if (LookAheadToken != null
&& LookAheadToken.Type != TokenType._UNDETERMINED_
&& LookAheadToken.Type != TokenType._NONE_) return LookAheadToken;
// if no scantokens specified, then scan for all of them (= backward compatible)
if (expectedtokens.Length == 0)
scantokens = Tokens;
else {
scantokens = new List<TokenType>(expectedtokens);
scantokens.AddRange(SkipList);
}
do {
int len = -1;
TokenType index = (TokenType)int.MaxValue;
string input = Input.Substring(startpos);
tok = new Token(startpos, EndPos);
for (i = 0; i < scantokens.Count; i++) {
Regex r = Patterns[scantokens[i]];
Match m = r.Match(input);
if (m.Success && m.Index == 0 && ((m.Length > len) || (scantokens[i] < index && m.Length == len))) {
len = m.Length;
index = scantokens[i];
}
}
if (index >= 0 && len >= 0) {
tok.EndPos = startpos + len;
tok.Text = Input.Substring(tok.StartPos, len);
tok.Type = index;
}
else if (tok.StartPos < tok.EndPos - 1) {
tok.Text = Input.Substring(tok.StartPos, 1);
}
if (SkipList.Contains(tok.Type)) {
startpos = tok.EndPos;
Skipped.Add(tok);
}
else {
// only assign to non-skipped tokens
tok.Skipped = Skipped; // assign prior skips to this token
Skipped = new List<Token>(); //reset skips
}
}
while (SkipList.Contains(tok.Type));
LookAheadToken = tok;
return tok;
}
public void Init(string input)
{
this.Input = input;
StartPos = 0;
EndPos = 0;
CurrentLine = 0;
CurrentColumn = 0;
CurrentPosition = 0;
LookAheadToken = null;
}
public Token GetToken(TokenType type)
{
Token t = new Token(this.StartPos, this.EndPos);
t.Type = type;
return t;
}