internal static FA Parse(LexContext pc, int accept = -1)
{
FA result = null, next = null;
int ich;
pc.EnsureStarted();
while (true)
{
switch (pc.Current)
{
case -1:
#if MINIMIZE
result = result.ToDfa();
result.TrimDuplicates();
#endif
return(result);
case '.':
var dot = FA.Set(new int[] { 0, 0x10ffff }, accept);
if (null == result)
{
result = dot;
}
else
{
result = FA.Concat(new FA[] { result, dot }, accept);
}
pc.Advance();
result = _ParseModifier(result, pc, accept);
break;
case '\\':
pc.Advance();
pc.Expecting();
var isNot = false;
switch (pc.Current)
{
case 'P':
isNot = true;
goto case 'p';
case 'p':
pc.Advance();
pc.Expecting('{');
var uc = new StringBuilder();
int uli = pc.Line;
int uco = pc.Column;
long upo = pc.Position;
while (-1 != pc.Advance() && '}' != pc.Current)
{
uc.Append((char)pc.Current);
}
pc.Expecting('}');
pc.Advance();
int uci = 0;
switch (uc.ToString())
{
case "Pe":
uci = 21;
break;
case "Pc":
uci = 19;
break;
case "Cc":
uci = 14;
break;
case "Sc":
uci = 26;
break;
case "Pd":
uci = 19;
break;
case "Nd":
uci = 8;
break;
case "Me":
uci = 7;
break;
case "Pf":
uci = 23;
break;
case "Cf":
uci = 15;
break;
case "Pi":
uci = 22;
break;
case "Nl":
uci = 9;
break;
case "Zl":
uci = 12;
break;
case "Ll":
uci = 1;
break;
case "Sm":
uci = 25;
break;
case "Lm":
uci = 3;
break;
case "Sk":
uci = 27;
break;
case "Mn":
uci = 5;
break;
case "Ps":
uci = 20;
break;
case "Lo":
uci = 4;
break;
case "Cn":
uci = 29;
break;
case "No":
uci = 10;
break;
case "Po":
uci = 24;
break;
case "So":
uci = 28;
break;
case "Zp":
uci = 13;
break;
case "Co":
uci = 17;
break;
case "Zs":
uci = 11;
break;
case "Mc":
uci = 6;
break;
case "Cs":
uci = 16;
break;
case "Lt":
uci = 2;
break;
case "Lu":
uci = 0;
break;
}
if (isNot)
{
next = FA.Set(CharacterClasses.UnicodeCategories[uci], accept);
}
else
{
next = FA.Set(CharacterClasses.NotUnicodeCategories[uci], accept);
}
break;
case 'd':
next = FA.Set(CharacterClasses.digit, accept);
pc.Advance();
break;
case 'D':
next = FA.Set(RangeUtility.NotRanges(CharacterClasses.digit), accept);
pc.Advance();
break;
case 's':
next = FA.Set(CharacterClasses.space, accept);
pc.Advance();
break;
case 'S':
next = FA.Set(RangeUtility.NotRanges(CharacterClasses.space), accept);
pc.Advance();
break;
case 'w':
next = FA.Set(CharacterClasses.word, accept);
pc.Advance();
break;
case 'W':
next = FA.Set(RangeUtility.NotRanges(CharacterClasses.word), accept);
pc.Advance();
break;
default:
if (-1 != (ich = _ParseEscapePart(pc)))
{
next = FA.Literal(new int[] { ich }, accept);
}
else
{
pc.Expecting(); // throw an error
return(null); // doesn't execute
}
break;
}
next = _ParseModifier(next, pc, accept);
if (null != result)
{
result = FA.Concat(new FA[] { result, next }, accept);
}
else
{
result = next;
}
break;
case ')':
#if MINIMIZE
result = result.ToDfa();
result.TrimDuplicates();
#endif
return(result);
case '(':
pc.Advance();
pc.Expecting();
next = Parse(pc, accept);
pc.Expecting(')');
pc.Advance();
next = _ParseModifier(next, pc, accept);
if (null == result)
{
result = next;
}
else
{
result = FA.Concat(new FA[] { result, next }, accept);
}
break;
case '|':
if (-1 != pc.Advance())
{
next = Parse(pc, accept);
result = FA.Or(new FA[] { result, next }, accept);
}
else
{
result = FA.Optional(result, accept);
}
break;
case '[':
var seti = _ParseSet(pc);
var set = seti.Value;
if (seti.Key)
{
set = RangeUtility.NotRanges(set);
}
next = FA.Set(set, accept);
next = _ParseModifier(next, pc, accept);
if (null == result)
{
result = next;
}
else
{
result = FA.Concat(new FA[] { result, next }, accept);
}
break;
default:
ich = pc.Current;
if (char.IsHighSurrogate((char)ich))
{
if (-1 == pc.Advance())
{
throw new ExpectingException("Expecting low surrogate in Unicode stream", pc.Line, pc.Column, pc.Position, pc.FileOrUrl, "low-surrogate");
}
ich = char.ConvertToUtf32((char)ich, (char)pc.Current);
}
next = FA.Literal(new int[] { ich }, accept);
pc.Advance();
next = _ParseModifier(next, pc, accept);
if (null == result)
{
result = next;
}
else
{
result = FA.Concat(new FA[] { result, next }, accept);
}
break;
}
}
}
/// <summary>
/// Writes a Graphviz dot specification of the specified closure to the specified <see cref="TextWriter"/>
/// </summary>
/// <param name="closure">The closure of all states</param>
/// <param name="writer">The writer</param>
/// <param name="options">A <see cref="DotGraphOptions"/> instance with any options, or null to use the defaults</param>
static void _WriteDotTo(IList <FA> closure, TextWriter writer, DotGraphOptions options = null)
{
if (null == options)
{
options = new DotGraphOptions();
}
string spfx = null == options.StatePrefix ? "q" : options.StatePrefix;
writer.WriteLine("digraph FA {");
writer.WriteLine("rankdir=LR");
writer.WriteLine("node [shape=circle]");
var finals = new List <FA>();
var neutrals = new List <FA>();
var accepting = closure[0].FillAcceptingStates();
foreach (var ffa in closure)
{
if (ffa.IsFinal && !ffa.IsAccepting)
{
finals.Add(ffa);
}
}
int i = 0;
foreach (var ffa in closure)
{
if (!finals.Contains(ffa))
{
if (ffa.IsAccepting)
{
accepting.Add(ffa);
}
else if (ffa.IsNeutral)
{
neutrals.Add(ffa);
}
}
var rngGrps = ffa.FillInputTransitionRangesGroupedByState();
foreach (var rngGrp in rngGrps)
{
var di = closure.IndexOf(rngGrp.Key);
writer.Write(spfx);
writer.Write(i);
writer.Write("->");
writer.Write(spfx);
writer.Write(di.ToString());
writer.Write(" [label=\"");
var sb = new StringBuilder();
var rngs = rngGrp.Value;
var nrngs = RangeUtility.NotRanges(rngs);
var isNot = false;
if (nrngs.Length < rngs.Length || (nrngs.Length == rngs.Length && 0x10ffff == rngs[rngs.Length - 1]))
{
isNot = true;
if (0 != nrngs.Length)
{
sb.Append("^");
}
else
{
sb.Append(".");
}
rngs = nrngs;
}
for (var r = 0; r < rngs.Length; r += 2)
{
_AppendRangeTo(sb, rngs, r);
}
if (isNot || sb.Length != 1 || (char.IsWhiteSpace(sb.ToString(), 0)))
{
writer.Write('[');
writer.Write(_EscapeLabel(sb.ToString()));
writer.Write(']');
}
else
{
writer.Write(_EscapeLabel(sb.ToString()));
}
writer.WriteLine("\"]");
}
// do epsilons
foreach (var fffa in ffa.EpsilonTransitions)
{
writer.Write(spfx);
writer.Write(i);
writer.Write("->");
writer.Write(spfx);
writer.Write(closure.IndexOf(fffa));
writer.WriteLine(" [style=dashed,color=gray]");
}
++i;
}
string delim = "";
i = 0;
foreach (var ffa in closure)
{
writer.Write(spfx);
writer.Write(i);
writer.Write(" [");
writer.Write("label=<");
writer.Write("<TABLE BORDER=\"0\"><TR><TD>");
writer.Write(spfx);
writer.Write("<SUB>");
writer.Write(i);
writer.Write("</SUB></TD></TR>");
if (ffa.IsAccepting)
{
writer.Write("<TR><TD>");
writer.Write(Convert.ToString(ffa.AcceptSymbol).Replace("\"", """));
writer.Write("</TD></TR>");
}
writer.Write("</TABLE>");
writer.Write(">");
bool isfinal = false;
if (accepting.Contains(ffa) || (isfinal = finals.Contains(ffa)))
{
writer.Write(",shape=doublecircle");
}
if (isfinal || neutrals.Contains(ffa))
{
writer.Write(",color=gray");
}
writer.WriteLine("]");
++i;
}
delim = "";
if (0 < accepting.Count)
{
foreach (var ntfa in accepting)
{
writer.Write(delim);
writer.Write(spfx);
writer.Write(closure.IndexOf(ntfa));
delim = ",";
}
writer.WriteLine(" [shape=doublecircle]");
}
delim = "";
if (0 < neutrals.Count)
{
foreach (var ntfa in neutrals)
{
writer.Write(delim);
writer.Write(spfx);
writer.Write(closure.IndexOf(ntfa));
delim = ",";
}
writer.WriteLine(" [color=gray]");
delim = "";
}
delim = "";
if (0 < finals.Count)
{
foreach (var ntfa in finals)
{
writer.Write(delim);
writer.Write(spfx);
writer.Write(closure.IndexOf(ntfa));
delim = ",";
}
writer.WriteLine(" [shape=doublecircle,color=gray]");
}
writer.WriteLine("}");
}
