public Scanner()
{
Regex regex;
Patterns = new Dictionary<TokenType, Regex>();
Tokens = new List<TokenType>();
LookAheadToken = null;
SkipList = new List<TokenType>();
SkipList.Add(TokenType.WHITESPACE);
regex = new Regex(@"[0-9]+", RegexOptions.Compiled);
Patterns.Add(TokenType.NUMBER, regex);
Tokens.Add(TokenType.NUMBER);
regex = new Regex(@"(\+|-)", RegexOptions.Compiled);
Patterns.Add(TokenType.PLUSMINUS, regex);
Tokens.Add(TokenType.PLUSMINUS);
regex = new Regex(@"\*|/", RegexOptions.Compiled);
Patterns.Add(TokenType.MULTDIV, regex);
Tokens.Add(TokenType.MULTDIV);
regex = new Regex(@"\(", RegexOptions.Compiled);
Patterns.Add(TokenType.BROPEN, regex);
Tokens.Add(TokenType.BROPEN);
regex = new Regex(@"\)", RegexOptions.Compiled);
Patterns.Add(TokenType.BRCLOSE, regex);
Tokens.Add(TokenType.BRCLOSE);
regex = new Regex(@"^$", RegexOptions.Compiled);
Patterns.Add(TokenType.EOF, regex);
Tokens.Add(TokenType.EOF);
regex = new Regex(@"\s+", RegexOptions.Compiled);
Patterns.Add(TokenType.WHITESPACE, regex);
Tokens.Add(TokenType.WHITESPACE);
}
public void UpdateRange(Token token)
{
if (token.StartPos < startpos) startpos = token.StartPos;
if (token.EndPos > endpos) endpos = token.EndPos;
}
private List<TokenType> SkipList; // tokens to be skipped
public Scanner()
{
Regex regex;
Patterns = new Dictionary<TokenType, Regex>();
Tokens = new List<TokenType>();
LookAheadToken = null;
Skipped = new List<Token>();
SkipList = new List<TokenType>();
SkipList.Add(TokenType.SPACE);
regex = new Regex(@"(?i)id:", RegexOptions.Compiled);
Patterns.Add(TokenType.IdKeyword1, regex);
Tokens.Add(TokenType.IdKeyword1);
regex = new Regex(@"#", RegexOptions.Compiled);
Patterns.Add(TokenType.IdKeyword2, regex);
Tokens.Add(TokenType.IdKeyword2);
regex = new Regex(@"time:", RegexOptions.Compiled);
Patterns.Add(TokenType.TimeKeyword, regex);
Tokens.Add(TokenType.TimeKeyword);
regex = new Regex(@"status:", RegexOptions.Compiled);
Patterns.Add(TokenType.StatusKeyword1, regex);
Tokens.Add(TokenType.StatusKeyword1);
regex = new Regex(@"state:", RegexOptions.Compiled);
Patterns.Add(TokenType.StatusKeyword2, regex);
Tokens.Add(TokenType.StatusKeyword2);
regex = new Regex(@"comment:", RegexOptions.Compiled);
Patterns.Add(TokenType.CommentKeyword1, regex);
Tokens.Add(TokenType.CommentKeyword1);
regex = new Regex(@"comments:", RegexOptions.Compiled);
Patterns.Add(TokenType.CommentKeyword2, regex);
Tokens.Add(TokenType.CommentKeyword2);
regex = new Regex(@"comm:", RegexOptions.Compiled);
Patterns.Add(TokenType.CommentKeyword3, regex);
Tokens.Add(TokenType.CommentKeyword3);
regex = new Regex(@"cmt:", RegexOptions.Compiled);
Patterns.Add(TokenType.CommentKeyword4, regex);
Tokens.Add(TokenType.CommentKeyword4);
regex = new Regex(@"^$", RegexOptions.Compiled);
Patterns.Add(TokenType.EOF, regex);
Tokens.Add(TokenType.EOF);
regex = new Regex(@"\s+", RegexOptions.Compiled);
Patterns.Add(TokenType.SPACE, regex);
Tokens.Add(TokenType.SPACE);
regex = new Regex(@":", RegexOptions.Compiled);
Patterns.Add(TokenType.Delimiter, regex);
Tokens.Add(TokenType.Delimiter);
regex = new Regex(@"[,\.\?!;]+", RegexOptions.Compiled);
Patterns.Add(TokenType.PUNCTUATIONMARK, regex);
Tokens.Add(TokenType.PUNCTUATIONMARK);
regex = new Regex(@",", RegexOptions.Compiled);
Patterns.Add(TokenType.COMMA, regex);
Tokens.Add(TokenType.COMMA);
regex = new Regex(@"[0-9]+", RegexOptions.Compiled);
Patterns.Add(TokenType.NUMBER, regex);
Tokens.Add(TokenType.NUMBER);
regex = new Regex(@"\d+([\.,]\d+)?", RegexOptions.Compiled);
Patterns.Add(TokenType.DECIMAL, regex);
Tokens.Add(TokenType.DECIMAL);
regex = new Regex(@"(?i)(.(?<!#)(?<!(id|time|status|state|comment|comm|comments|cmt):))+\s", RegexOptions.Compiled);
Patterns.Add(TokenType.ANY_TEXT, regex);
Tokens.Add(TokenType.ANY_TEXT);
regex = new Regex(@"(?i)((?!#)(?!(id|time|status|state|comment|comm|comments|cmt):).)", RegexOptions.Compiled);
Patterns.Add(TokenType.ANY_SYMBOL, regex);
Tokens.Add(TokenType.ANY_SYMBOL);
}
/// <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;
}
/// <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;
}
public Token GetToken(TokenType type)
{
Token t = new Token(this.StartPos, this.EndPos);
t.Type = type;
return t;
}
public void Init(string input)
{
this.Input = input;
StartPos = 0;
EndPos = 0;
CurrentLine = 0;
CurrentColumn = 0;
CurrentPosition = 0;
LookAheadToken = null;
}
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);
SkipList.Add(TokenType.COMMENTLINE);
SkipList.Add(TokenType.COMMENTBLOCK);
regex = new Regex(@"\(", RegexOptions.Compiled);
Patterns.Add(TokenType.BRACKETOPEN, regex);
Tokens.Add(TokenType.BRACKETOPEN);
regex = new Regex(@"\)", RegexOptions.Compiled);
Patterns.Add(TokenType.BRACKETCLOSE, regex);
Tokens.Add(TokenType.BRACKETCLOSE);
regex = new Regex(@"\{[^\}]*\}([^};][^}]*\}+)*;", RegexOptions.Compiled);
Patterns.Add(TokenType.CODEBLOCK, regex);
Tokens.Add(TokenType.CODEBLOCK);
regex = new Regex(@",", RegexOptions.Compiled);
Patterns.Add(TokenType.COMMA, regex);
Tokens.Add(TokenType.COMMA);
regex = new Regex(@"\[", RegexOptions.Compiled);
Patterns.Add(TokenType.SQUAREOPEN, regex);
Tokens.Add(TokenType.SQUAREOPEN);
regex = new Regex(@"\]", RegexOptions.Compiled);
Patterns.Add(TokenType.SQUARECLOSE, regex);
Tokens.Add(TokenType.SQUARECLOSE);
regex = new Regex(@"=", RegexOptions.Compiled);
Patterns.Add(TokenType.ASSIGN, regex);
Tokens.Add(TokenType.ASSIGN);
regex = new Regex(@"\|", RegexOptions.Compiled);
Patterns.Add(TokenType.PIPE, regex);
Tokens.Add(TokenType.PIPE);
regex = new Regex(@";", RegexOptions.Compiled);
Patterns.Add(TokenType.SEMICOLON, regex);
Tokens.Add(TokenType.SEMICOLON);
regex = new Regex(@"(\*|\+|\?)", RegexOptions.Compiled);
Patterns.Add(TokenType.UNARYOPER, regex);
Tokens.Add(TokenType.UNARYOPER);
regex = new Regex(@"[a-zA-Z_][a-zA-Z0-9_]*", RegexOptions.Compiled);
Patterns.Add(TokenType.IDENTIFIER, regex);
Tokens.Add(TokenType.IDENTIFIER);
regex = new Regex(@"[0-9]+", RegexOptions.Compiled);
Patterns.Add(TokenType.INTEGER, regex);
Tokens.Add(TokenType.INTEGER);
regex = new Regex(@"[0-9]*\.[0-9]+", RegexOptions.Compiled);
Patterns.Add(TokenType.DOUBLE, regex);
Tokens.Add(TokenType.DOUBLE);
regex = new Regex(@"(0x[0-9a-fA-F]{6})", RegexOptions.Compiled);
Patterns.Add(TokenType.HEX, regex);
Tokens.Add(TokenType.HEX);
regex = new Regex(@"->", RegexOptions.Compiled);
Patterns.Add(TokenType.ARROW, regex);
Tokens.Add(TokenType.ARROW);
regex = new Regex(@"<%\s*@", RegexOptions.Compiled);
Patterns.Add(TokenType.DIRECTIVEOPEN, regex);
Tokens.Add(TokenType.DIRECTIVEOPEN);
regex = new Regex(@"%>", RegexOptions.Compiled);
Patterns.Add(TokenType.DIRECTIVECLOSE, regex);
Tokens.Add(TokenType.DIRECTIVECLOSE);
regex = new Regex(@"^$", RegexOptions.Compiled);
Patterns.Add(TokenType.EOF, regex);
Tokens.Add(TokenType.EOF);
regex = new Regex(@"@?\""(\""\""|[^\""])*\""", RegexOptions.Compiled);
Patterns.Add(TokenType.STRING, regex);
Tokens.Add(TokenType.STRING);
regex = new Regex(@"\s+", RegexOptions.Compiled);
Patterns.Add(TokenType.WHITESPACE, regex);
Tokens.Add(TokenType.WHITESPACE);
regex = new Regex(@"//[^\n]*\n?", RegexOptions.Compiled);
Patterns.Add(TokenType.COMMENTLINE, regex);
Tokens.Add(TokenType.COMMENTLINE);
regex = new Regex(@"/\*[^*]*\*+(?:[^/*][^*]*\*+)*/", RegexOptions.Compiled);
Patterns.Add(TokenType.COMMENTBLOCK, regex);
Tokens.Add(TokenType.COMMENTBLOCK);
}
/// <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
CurrentLine = tok.Line + (tok.Text.Length - tok.Text.Replace("\n", "").Length);
CurrentFile = tok.File;
return tok;
}
/// <summary>
/// returns token with longest best match
/// </summary>
/// <returns></returns>
public Token LookAhead(params TokenType[] expectedtokens)
{
int i;
int startpos = StartPos;
int endpos = EndPos;
int currentline = CurrentLine;
string currentFile = CurrentFile;
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);
}
// Update the line and column count for error reporting.
tok.File = currentFile;
tok.Line = currentline;
if (tok.StartPos < Input.Length)
tok.Column = tok.StartPos - Input.LastIndexOf('\n', tok.StartPos);
if (SkipList.Contains(tok.Type))
{
startpos = tok.EndPos;
endpos = tok.EndPos;
currentline = tok.Line + (tok.Text.Length - tok.Text.Replace("\n", "").Length);
currentFile = tok.File;
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
}
// Check to see if the parsed token wants to
// alter the file and line number.
if (tok.Type == FileAndLine)
{
var match = Patterns[tok.Type].Match(tok.Text);
var fileMatch = match.Groups["File"];
if (fileMatch.Success)
currentFile = fileMatch.Value;
var lineMatch = match.Groups["Line"];
if (lineMatch.Success)
currentline = int.Parse(lineMatch.Value);
}
}
while (SkipList.Contains(tok.Type));
LookAheadToken = tok;
return tok;
}
public void Init(string input, string fileName = "")
{
this.Input = input;
StartPos = 0;
EndPos = 0;
CurrentFile = fileName;
CurrentLine = 1;
CurrentColumn = 1;
CurrentPosition = 0;
LookAheadToken = null;
}
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(@"\(", RegexOptions.Compiled);
Patterns.Add(TokenType.BROPEN, regex);
Tokens.Add(TokenType.BROPEN);
regex = new Regex(@"\)", RegexOptions.Compiled);
Patterns.Add(TokenType.BRCLOSE, regex);
Tokens.Add(TokenType.BRCLOSE);
regex = new Regex(@"^$", RegexOptions.Compiled);
Patterns.Add(TokenType.EOF, regex);
Tokens.Add(TokenType.EOF);
regex = new Regex("==", RegexOptions.Compiled);
Patterns.Add(TokenType.EQ, regex);
Tokens.Add(TokenType.EQ);
regex = new Regex("!=", RegexOptions.Compiled);
Patterns.Add(TokenType.NEQ, regex);
Tokens.Add(TokenType.NEQ);
regex = new Regex("contains", RegexOptions.Compiled);
Patterns.Add(TokenType.CONTAINS, regex);
Tokens.Add(TokenType.CONTAINS);
regex = new Regex("&&", RegexOptions.Compiled);
Patterns.Add(TokenType.AND, regex);
Tokens.Add(TokenType.AND);
regex = new Regex("\\|\\|", RegexOptions.Compiled);
Patterns.Add(TokenType.OR, regex);
Tokens.Add(TokenType.OR);
regex = new Regex(@"\w+", RegexOptions.Compiled);
Patterns.Add(TokenType.NAME, regex);
Tokens.Add(TokenType.NAME);
regex = new Regex(@"""[^""]*""", RegexOptions.Compiled);
Patterns.Add(TokenType.VALUE, regex);
Tokens.Add(TokenType.VALUE);
regex = new Regex(";", RegexOptions.Compiled);
Patterns.Add(TokenType.SEP, regex);
Tokens.Add(TokenType.SEP);
regex = new Regex("Was", RegexOptions.Compiled);
Patterns.Add(TokenType.WAS, regex);
Tokens.Add(TokenType.WAS);
regex = new Regex("Obsolete", RegexOptions.Compiled);
Patterns.Add(TokenType.OBSOLETE, regex);
Tokens.Add(TokenType.OBSOLETE);
regex = new Regex("Delete", RegexOptions.Compiled);
Patterns.Add(TokenType.DELETE, regex);
Tokens.Add(TokenType.DELETE);
regex = new Regex(@"\s+", RegexOptions.Compiled);
Patterns.Add(TokenType.WHITESPACE, regex);
Tokens.Add(TokenType.WHITESPACE);
}
/// <summary>
/// returns token with longest best match
/// </summary>
/// <returns></returns>
public Token LookAhead(params TokenType[] scantokens)
{
int len;
int index = -1;
int i;
int startpos = StartPos;
Token tok = null;
// 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 (scantokens.Length == 0)
scantokens = Tokens.ToArray();
do
{
len = -1;
string input = Input.Substring(startpos);
tok = new Token(startpos, EndPos);
for (i = 0; i < scantokens.Length; i++)
{
Regex r = Patterns[scantokens[i]];
Match m = r.Match(input);
if (m.Success && m.Index == 0 && m.Length > len)
{
len = m.Length;
index = i;
}
}
if (index >= 0 && len >= 0)
{
tok.EndPos = startpos + len;
tok.Text = Input.Substring(tok.StartPos, len);
tok.Type = scantokens[index];
}
else
{
if (tok.EndPos < Input.Length)
tok.Text = Input.Substring(tok.StartPos, 1);
}
if (SkipList.Contains(tok.Type))
{
startpos = tok.EndPos;
Skipped.Add(tok);
}
}
while (SkipList.Contains(tok.Type));
LookAheadToken = tok;
return tok;
}