public void AddMatch(string str, bool caseInsensitive, TokenPattern value)
{
DFAState state;
char c = str[0];
int start = 0;
if (caseInsensitive)
{
c = Char.ToLower(c);
}
if (c < 128)
{
state = _ascii[c];
if (state == null)
{
state = _ascii[c] = new DFAState();
}
start++;
}
else
{
state = _nonAscii;
}
for (int i = start; i < str.Length; i++)
{
var next = state.Tree.Find(str[i], caseInsensitive);
if (next == null)
{
next = new DFAState();
state.Tree.Add(str[i], caseInsensitive, next);
}
state = next;
}
state.Value = value;
}
protected virtual Token NewToken(TokenPattern pattern,
string image,
int line,
int column)
{
return(new Token(pattern, image, line, column));
}
public void Update(int length, TokenPattern pattern)
{
if (this._length < length)
{
this._length = length;
this._pattern = pattern;
}
}
/**
* nfa - true to attempt as an nfa pattern for regexp. This handles most things except the complex repeates, ie {1,4}
*/
public void AddPattern(TokenPattern pattern, bool nfa = true)
{
switch (pattern.Type)
{
case TokenPattern.PatternType.STRING:
try
{
_stringDfaMatcher.AddPattern(pattern);
}
catch (Exception e)
{
throw new ParserCreationException(
ParserCreationException.ErrorType.INVALID_TOKEN,
pattern.Name,
"error adding string token: " +
e.Message);
}
break;
case TokenPattern.PatternType.REGEXP:
if (nfa)
{
try
{
_nfaMatcher.AddPattern(pattern);
}
catch (Exception)
{
nfa = false;
}
}
if (!nfa)
{
try
{
_regExpMatcher.AddPattern(pattern);
}
catch (Exception e)
{
throw new ParserCreationException(
ParserCreationException.ErrorType.INVALID_TOKEN,
pattern.Name,
"regular expression contains error(s): " +
e.Message);
}
}
break;
default:
throw new ParserCreationException(
ParserCreationException.ErrorType.INVALID_TOKEN,
pattern.Name,
"pattern type " + pattern.Type +
" is undefined");
}
}
public override void Match(ReaderBuffer buffer, TokenMatch match)
{
TokenPattern res = _automaton.Match(buffer, IgnoreCase);
if (res != null)
{
match.Update(res.Pattern.Length, res);
}
}
public void AddRegExpMatch(string pattern,
bool ignoreCase,
TokenPattern value)
{
TokenRegExpParser parser = new TokenRegExpParser(pattern, ignoreCase);
string debug = "DFA regexp; " + parser.GetDebugInfo();
var isAscii = parser.Start.IsAsciiOutgoing();
for (int i = 0; isAscii && i < 128; i++)
{
bool match = false;
for (int j = 0; j < parser.Start.Outgoing.Length; j++)
{
if (parser.Start.Outgoing[j].Match((char)i))
{
if (match)
{
isAscii = false;
break;
}
match = true;
}
}
if (match && _initialChar[i] != null)
{
isAscii = false;
}
}
if (parser.Start.Incoming.Length > 0)
{
_initial.AddOut(new NFAEpsilonTransition(parser.Start));
debug += ", uses initial epsilon";
}
else if (isAscii && !ignoreCase)
{
for (int i = 0; isAscii && i < 128; i++)
{
for (int j = 0; j < parser.Start.Outgoing.Length; j++)
{
if (parser.Start.Outgoing[j].Match((char)i))
{
_initialChar[i] = parser.Start.Outgoing[j].State;
}
}
}
debug += ", uses ASCII lookup";
}
else
{
parser.Start.MergeInto(_initial);
debug += ", uses initial state";
}
parser.End.Value = value;
value.DebugInfo = debug;
}
public void AddPattern(TokenPattern pattern)
{
switch (pattern.Type)
{
case TokenPattern.PatternType.STRING:
try
{
_stringDfaMatcher.AddPattern(pattern);
}
catch (Exception e)
{
throw new ParserCreationException(
ParserCreationException.ErrorType.INVALID_TOKEN,
pattern.Name,
"error adding string token: " +
e.Message);
}
break;
case TokenPattern.PatternType.REGEXP:
try
{
//Because of a bug in nfaMatcher's treatment
//of repeath specifiers we resort to effectively always using the
//regExp matcher.
//Under some conditions, the process of throwing the
//exception from the nfaMatcher and catching it
//incurs an unacceptable slowdown in performance
_regExpMatcher.AddPattern(pattern);
}
catch (Exception)
{
try
{
_nfaMatcher.AddPattern(pattern);
}
catch (Exception e)
{
throw new ParserCreationException(
ParserCreationException.ErrorType.INVALID_TOKEN,
pattern.Name,
"regular expression contains error(s): " +
e.Message);
}
}
break;
default:
throw new ParserCreationException(
ParserCreationException.ErrorType.INVALID_TOKEN,
pattern.Name,
"pattern type " + pattern.Type +
" is undefined");
}
}
public override void AddPattern(TokenPattern pattern)
{
if (pattern.Type == TokenPattern.PatternType.STRING)
{
_automaton.AddTextMatch(pattern.Pattern, IgnoreCase, pattern);
}
else
{
_automaton.AddRegExpMatch(pattern.Pattern, IgnoreCase, pattern);
}
base.AddPattern(pattern);
}
public void Update(int length, TokenPattern pattern)
{
if (this._length < length)
{
this._length = length;
this._pattern = pattern;
}
else if (this._length == length && this._pattern.Id > pattern.Id)
{
this._length = length;
this._pattern = pattern;
}
}
public void AddPattern(TokenPattern pattern)
{
switch (pattern.Type)
{
case TokenPattern.PatternType.STRING:
try
{
_stringDfaMatcher.AddPattern(pattern);
}
catch (Exception e)
{
throw new ParserCreationException(
ParserCreationException.ErrorType.INVALID_TOKEN,
pattern.Name,
"error adding string token: " +
e.Message);
}
break;
case TokenPattern.PatternType.REGEXP:
try
{
//See comments in NetStandard version for why the regExp matcher is tried first
_regExpMatcher.AddPattern(pattern);
}
catch (Exception)
{
try
{
_nfaMatcher.AddPattern(pattern);
}
catch (Exception e)
{
throw new ParserCreationException(
ParserCreationException.ErrorType.INVALID_TOKEN,
pattern.Name,
"regular expression contains error(s): " +
e.Message);
}
}
break;
default:
throw new ParserCreationException(
ParserCreationException.ErrorType.INVALID_TOKEN,
pattern.Name,
"pattern type " + pattern.Type +
" is undefined");
}
}
public Token(TokenPattern pattern, string image, int line, int col)
{
this._pattern = pattern;
this._image = image;
this._startLine = line;
this._startColumn = col;
this._endLine = line;
this._endColumn = col + image.Length - 1;
for (int pos = 0; image.IndexOf('\n', pos) >= 0;)
{
pos = image.IndexOf('\n', pos) + 1;
this._endLine++;
_endColumn = image.Length - pos;
}
}
public TokenPattern Match(ReaderBuffer buffer, bool caseInsensitive)
{
TokenPattern result = null;
DFAState state;
int pos = 0;
var c = buffer.Peek(0);
if (c < 0)
{
return(null);
}
if (caseInsensitive)
{
c = Char.ToLower((char)c);
}
if (c < 128)
{
state = _ascii[c];
if (state == null)
{
return(null);
}
else if (state.Value != null)
{
result = state.Value;
}
pos++;
}
else
{
state = _nonAscii;
}
while ((c = buffer.Peek(pos)) >= 0)
{
state = state.Tree.Find((char)c, caseInsensitive);
if (state == null)
{
break;
}
else if (state.Value != null)
{
result = state.Value;
}
pos++;
}
return(result);
}
public override void AddPattern(TokenPattern pattern)
{
REHandler re;
try
{
re = new GrammaticaRE(pattern.Pattern, IgnoreCase);
pattern.DebugInfo = "Grammatica regexp\n" + re;
}
catch (Exception)
{
re = new SystemRE(pattern.Pattern, IgnoreCase);
pattern.DebugInfo = "native .NET regexp";
}
Array.Resize(ref _regExps, _regExps.Length + 1);
_regExps[_regExps.Length - 1] = re;
base.AddPattern(pattern);
}
public void AddTextMatch(string str, bool ignoreCase, TokenPattern value)
{
NFAState state;
char ch = str[0];
if (ch < 128 && !ignoreCase)
{
state = _initialChar[ch];
if (state == null)
{
state = _initialChar[ch] = new NFAState();
}
}
else
{
state = _initial.AddOut(ch, ignoreCase, null);
}
for (int i = 1; i < str.Length; i++)
{
state = state.AddOut(str[i], ignoreCase, null);
}
state.Value = value;
}
public void Clear()
{
_length = 0;
_pattern = null;
}
public override void AddPattern(TokenPattern pattern)
{
_automaton.AddMatch(pattern.Pattern, IgnoreCase, pattern);
base.AddPattern(pattern);
}
public virtual void AddPattern(TokenPattern pattern)
{
Array.Resize(ref Patterns, Patterns.Length + 1);
Patterns[Patterns.Length - 1] = pattern;
}