/// <summary>
/// Pattern matches through a string of text
/// </summary>
/// <param name="context">The parse context to search</param>
/// <returns>A <see cref="CharFAMatch"/> that contains the match information, or null if the match is not found.</returns>
public CharFAMatch Match(ParseContext context)
{
context.EnsureStarted();
var line = context.Line;
var column = context.Column;
var position = context.Position;
var l = context.CaptureBuffer.Length;
var success = false;
// keep going until we find something or reach the end
while (-1 != context.Current && !(success = _DoMatch(context)))
{
line = context.Line;
column = context.Column;
position = context.Position;
l = context.CaptureBuffer.Length;
}
if (success)
{
return(new CharFAMatch(
line,
column,
position,
context.GetCapture(l)));
}
return(null);
}
static Cfg _Parse(ParseContext pc)
{
var result = new Cfg();
pc.EnsureStarted();
while (-1 != pc.Current)
{
result.Rules.Add(_ParseRule(pc));
}
return(result);
}
static bool _SkipWhiteSpace(ParseContext pc)
{
pc.EnsureStarted();
if (-1 == pc.Current || '\n' == pc.Current || !char.IsWhiteSpace((char)pc.Current))
{
return(false);
}
while (-1 != pc.Advance() && '\n' != pc.Current && char.IsWhiteSpace((char)pc.Current))
{
;
}
return(true);
}
static bool _DoMatchDfa(CharDfaEntry[] dfaTable, ParseContext context)
{
// track our current state
var state = 0;
// prepare the parse context
context.EnsureStarted();
while (true)
{
// if no more input
if (-1 == context.Current)
{
// if we accept, return that
return(-1 != dfaTable[state].AcceptSymbolId);
}
// move by current character
var newState = MoveDfa(dfaTable, state, (char)context.Current);
// we couldn't match anything
if (-1 == newState)
{
// if we accept, return that
if (-1 != dfaTable[state].AcceptSymbolId)
{
return(true);
}
// otherwise error
// advance the input
context.Advance();
return(false);
}
// store the current character
context.CaptureCurrent();
// advance the input
context.Advance();
// iterate to our next states
state = newState;
}
}
static CharFA <TAccept> _Parse(ParseContext pc, TAccept accept)
{
CharFA <TAccept> result = new CharFA <TAccept>(true, accept);
CharFA <TAccept> f, next;
int ch;
pc.EnsureStarted();
var current = result;
while (true)
{
switch (pc.Current)
{
case -1:
return(result);
case '.':
pc.Advance();
f = current.FirstAcceptingState as CharFA <TAccept>;
current = Set(new CharRange[] { new CharRange(char.MinValue, char.MaxValue) }, accept);
switch (pc.Current)
{
case '*':
current = Kleene(current, accept);
pc.Advance();
break;
case '+':
current = Repeat(current, accept);
pc.Advance();
break;
case '?':
current = Optional(current, accept);
pc.Advance();
break;
}
f.IsAccepting = false;
f.EpsilonTransitions.Add(current);
break;
case '\\':
if (-1 != (ch = _ParseEscape(pc)))
{
next = null;
switch (pc.Current)
{
case '*':
next = new CharFA <TAccept>();
next.Transitions.Add((char)ch, new CharFA <TAccept>(true, accept));
next = Kleene(next, accept);
pc.Advance();
break;
case '+':
next = new CharFA <TAccept>();
next.Transitions.Add((char)ch, new CharFA <TAccept>(true, accept));
next = Repeat(next, accept);
pc.Advance();
break;
case '?':
next = new CharFA <TAccept>();
next.Transitions.Add((char)ch, new CharFA <TAccept>(true, accept));
next = Optional(next, accept);
pc.Advance();
break;
default:
current = current.FirstAcceptingState as CharFA <TAccept>;
current.IsAccepting = false;
current.Transitions.Add((char)ch, new CharFA <TAccept>(true, accept));
break;
}
if (null != next)
{
current = current.FirstAcceptingState as CharFA <TAccept>;
current.IsAccepting = false;
current.EpsilonTransitions.Add(next);
current = next;
}
}
else
{
pc.Expecting(); // throw an error
return(null); // doesn't execute
}
break;
case ')':
return(result);
case '(':
pc.Advance();
pc.Expecting();
f = current.FirstAcceptingState as CharFA <TAccept>;
current = _Parse(pc, accept);
pc.Expecting(')');
pc.Advance();
switch (pc.Current)
{
case '*':
current = Kleene(current, accept);
pc.Advance();
break;
case '+':
current = Repeat(current, accept);
pc.Advance();
break;
case '?':
current = Optional(current, accept);
pc.Advance();
break;
}
var ff = f.FirstAcceptingState;
ff.EpsilonTransitions.Add(current);
ff.IsAccepting = false;
break;
case '|':
if (-1 != pc.Advance())
{
current = _Parse(pc, accept);
result = Or(new CharFA <TAccept>[] { result as CharFA <TAccept>, current as CharFA <TAccept> }, accept);
}
else
{
current = current.FirstAcceptingState as CharFA <TAccept>;
result = Optional(result, accept);
}
break;
case '[':
pc.ClearCapture();
pc.Advance();
pc.Expecting();
bool not = false;
if ('^' == pc.Current)
{
not = true;
pc.Advance();
pc.Expecting();
}
pc.TryReadUntil(']', '\\', false);
pc.Expecting(']');
pc.Advance();
var r = (!not && "." == pc.Capture) ?
new CharRange[] { new CharRange(char.MinValue, char.MaxValue) } :
_ParseRanges(pc.Capture, true);
if (not)
{
r = CharRange.NotRanges(r);
}
f = current.FirstAcceptingState as CharFA <TAccept>;
current = Set(r, accept);
switch (pc.Current)
{
case '*':
current = Kleene(current, accept);
pc.Advance();
break;
case '+':
current = Repeat(current, accept);
pc.Advance();
break;
case '?':
current = Optional(current, accept);
pc.Advance();
break;
}
f.IsAccepting = false;
f.EpsilonTransitions.Add(current);
break;
default:
ch = pc.Current;
pc.Advance();
next = null;
switch (pc.Current)
{
case '*':
next = new CharFA <TAccept>();
next.Transitions.Add((char)ch, new CharFA <TAccept>(true, accept));
next = Kleene(next, accept);
pc.Advance();
break;
case '+':
next = new CharFA <TAccept>();
next.Transitions.Add((char)ch, new CharFA <TAccept>(true, accept));
next = Repeat(next, accept);
pc.Advance();
break;
case '?':
next = new CharFA <TAccept>();
next.Transitions.Add((char)ch, new CharFA <TAccept>(true, accept));
next = Optional(next, accept);
pc.Advance();
break;
default:
current = current.FirstAcceptingState as CharFA <TAccept>;
current.IsAccepting = false;
current.Transitions.Add((char)ch, new CharFA <TAccept>(true, accept));
break;
}
if (null != next)
{
current = current.FirstAcceptingState as CharFA <TAccept>;
current.IsAccepting = false;
current.EpsilonTransitions.Add(next);
current = next;
}
break;
}
}
}
internal static RegexExpression Parse(ParseContext pc)
{
RegexExpression result = null, next = null;
int ich;
pc.EnsureStarted();
var line = pc.Line;
var column = pc.Column;
var position = pc.Position;
while (true)
{
switch (pc.Current)
{
case -1:
return(result);
case '.':
var nset = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetRangeEntry(char.MinValue, char.MaxValue) }, false);
nset.SetLocation(line, column, position);
if (null == result)
{
result = nset;
}
else
{
result = new RegexConcatExpression(result, nset);
result.SetLocation(line, column, position);
}
pc.Advance();
result = _ParseModifier(result, pc);
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
case '\\':
if (-1 != (ich = _ParseEscape(pc)))
{
next = new RegexLiteralExpression((char)ich);
next.SetLocation(line, column, position);
next = _ParseModifier(next, pc);
if (null != result)
{
result = new RegexConcatExpression(result, next);
result.SetLocation(line, column, position);
}
else
{
result = next;
}
}
else
{
pc.Expecting(); // throw an error
return(null); // doesn't execute
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
case ')':
return(result);
case '(':
pc.Advance();
pc.Expecting();
next = Parse(pc);
pc.Expecting(')');
pc.Advance();
next = _ParseModifier(next, pc);
if (null == result)
{
result = next;
}
else
{
result = new RegexConcatExpression(result, next);
result.SetLocation(line, column, position);
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
case '|':
if (-1 != pc.Advance())
{
next = Parse(pc);
result = new RegexOrExpression(result, next);
result.SetLocation(line, column, position);
}
else
{
result = new RegexOrExpression(result, null);
result.SetLocation(line, column, position);
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
case '[':
pc.ClearCapture();
pc.Advance();
pc.Expecting();
bool not = false;
if ('^' == pc.Current)
{
not = true;
pc.Advance();
pc.Expecting();
}
var ranges = _ParseRanges(pc);
if (ranges.Count == 0)
{
System.Diagnostics.Debugger.Break();
}
pc.Expecting(']');
pc.Advance();
next = new RegexCharsetExpression(ranges, not);
next.SetLocation(line, column, position);
next = _ParseModifier(next, pc);
if (null == result)
{
result = next;
}
else
{
result = new RegexConcatExpression(result, next);
result.SetLocation(pc.Line, pc.Column, pc.Position);
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
default:
ich = pc.Current;
next = new RegexLiteralExpression((char)ich);
next.SetLocation(line, column, position);
pc.Advance();
next = _ParseModifier(next, pc);
if (null == result)
{
result = next;
}
else
{
result = new RegexConcatExpression(result, next);
result.SetLocation(line, column, position);
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
}
}
}
static IList <RegexCharsetEntry> _ParseRanges(ParseContext pc)
{
pc.EnsureStarted();
var result = new List <RegexCharsetEntry>();
RegexCharsetEntry next = null;
bool readDash = false;
while (-1 != pc.Current && ']' != pc.Current)
{
switch (pc.Current)
{
case '[': // char class
if (null != next)
{
result.Add(next);
if (readDash)
{
result.Add(new RegexCharsetCharEntry('-'));
}
result.Add(new RegexCharsetCharEntry('-'));
}
pc.Advance();
pc.Expecting(':');
pc.Advance();
var l = pc.CaptureBuffer.Length;
pc.TryReadUntil(':', false);
var n = pc.GetCapture(l);
pc.Advance();
pc.Expecting(']');
pc.Advance();
result.Add(new RegexCharsetClassEntry(n));
readDash = false;
next = null;
break;
case '\\':
//pc.Advance();
//pc.Expecting();
var ch = (char)_ParseEscape(pc);
if (null == next)
{
next = new RegexCharsetCharEntry(ch);
}
else
{
if (readDash)
{
result.Add(new RegexCharsetRangeEntry(((RegexCharsetCharEntry)next).Value, ch));
next = null;
readDash = false;
}
else
{
result.Add(next);
next = new RegexCharsetCharEntry(ch);
}
}
break;
case '-':
pc.Advance();
if (null == next)
{
next = new RegexCharsetCharEntry('-');
readDash = false;
}
else
{
if (readDash)
{
result.Add(next);
}
readDash = true;
}
break;
default:
if (null == next)
{
next = new RegexCharsetCharEntry((char)pc.Current);
}
else
{
if (readDash)
{
result.Add(new RegexCharsetRangeEntry(((RegexCharsetCharEntry)next).Value, (char)pc.Current));
next = null;
readDash = false;
}
else
{
result.Add(next);
next = new RegexCharsetCharEntry((char)pc.Current);
}
}
pc.Advance();
break;
}
}
if (null != next)
{
result.Add(next);
if (readDash)
{
next = new RegexCharsetCharEntry('-');
result.Add(next);
}
}
return(result);
}