public static NfaPair NewNLPair(Spec spec)
{
NfaPair nfaPair = Alloc.NewNfaPair();
nfaPair.end = Alloc.NewNfa(spec);
nfaPair.start = Alloc.NewNfa(spec);
Nfa start = nfaPair.start;
start.Next = Alloc.NewNfa(spec);
Nfa next = start.Next;
next.Edge = '';
next.SetCharSet(new CharSet());
next.GetCharSet().add(10);
next.Next = nfaPair.end;
start.Sibling = Alloc.NewNfa(spec);
Nfa sibling = start.Sibling;
sibling.Edge = '\r';
sibling.Next = Alloc.NewNfa(spec);
Nfa next2 = sibling.Next;
next2.Next = null;
next2.Sibling = Alloc.NewNfa(spec);
next2.Sibling.Edge = '\n';
next2.Sibling.Next = nfaPair.end;
return(nfaPair);
}
/*
* Function: cat_expr
* Description: Recursive descent regular expression parser.
*/
private static void cat_expr(NfaPair pair)
{
NfaPair e2_pair;
#if DESCENT_DEBUG
Utility.enter("cat_expr", spec.lexeme, spec.current_token);
#endif
#if DEBUG
Utility.assert(null != pair);
#endif
e2_pair = Alloc.NewNfaPair();
if (first_in_cat(spec.current_token))
{
factor(pair);
}
while (first_in_cat(spec.current_token))
{
factor(e2_pair);
/* Destroy */
pair.end.mimic(e2_pair.start);
discardNfa(e2_pair.start);
pair.end = e2_pair.end;
}
#if DESCENT_DEBUG
Utility.leave("cat_expr", spec.lexeme, spec.current_token);
#endif
}
private static int add_to_dstates(Spec s, Bunch bunch)
{
Dfa dfa = Alloc.NewDfa(s);
dfa.SetNFASet(new List <Nfa>(bunch.GetNFASet()));
dfa.SetNFABit(new BitSet(bunch.GetNFABit()));
dfa.SetAccept(bunch.GetAccept());
dfa.SetAnchor(bunch.GetAnchor());
dfa.ClearMarked();
s.dfa_sets[dfa.GetNFABit()] = dfa;
return(dfa.Label);
}
private static void cat_expr(NfaPair pair)
{
NfaPair nfaPair = Alloc.NewNfaPair();
if (MakeNfa.first_in_cat(MakeNfa.spec.current_token))
{
MakeNfa.factor(pair);
}
while (MakeNfa.first_in_cat(MakeNfa.spec.current_token))
{
MakeNfa.factor(nfaPair);
pair.end.mimic(nfaPair.start);
MakeNfa.discardNfa(nfaPair.start);
pair.end = nfaPair.end;
}
}
private static Nfa machine()
{
Nfa nfa = Alloc.NewNfa(MakeNfa.spec);
Nfa nfa2 = nfa;
BitSet states = MakeNfa.gen.GetStates();
MakeNfa.spec.current_token = Tokens.EOS;
MakeNfa.gen.Advance();
if (Tokens.END_OF_INPUT != MakeNfa.spec.current_token)
{
nfa2.Next = MakeNfa.rule();
MakeNfa.ProcessStates(states, nfa2.Next);
}
while (Tokens.END_OF_INPUT != MakeNfa.spec.current_token)
{
states = MakeNfa.gen.GetStates();
MakeNfa.gen.Advance();
if (Tokens.END_OF_INPUT == MakeNfa.spec.current_token)
{
break;
}
nfa2.Sibling = Alloc.NewNfa(MakeNfa.spec);
nfa2 = nfa2.Sibling;
nfa2.Next = MakeNfa.rule();
MakeNfa.ProcessStates(states, nfa2.Next);
}
nfa2.Sibling = Alloc.NewNfa(MakeNfa.spec);
nfa2 = nfa2.Sibling;
nfa2.Next = Alloc.NewNfa(MakeNfa.spec);
Nfa next = nfa2.Next;
next.Edge = '';
next.Next = Alloc.NewNfa(MakeNfa.spec);
next.SetCharSet(new CharSet());
next.GetCharSet().add((int)MakeNfa.spec.BOL);
next.GetCharSet().add((int)MakeNfa.spec.EOF);
next.Next.SetAccept(new Accept(null));
for (int i = 0; i < MakeNfa.spec.States.Count; i++)
{
List <Nfa> list = MakeNfa.spec.state_rules[i];
list.Add(next);
}
return(nfa);
}
private static Nfa rule()
{
int num = 0;
NfaPair nfaPair = Alloc.NewNfaPair();
Nfa nfa;
Nfa end;
if (MakeNfa.spec.current_token == Tokens.AT_BOL)
{
num |= 1;
MakeNfa.gen.Advance();
MakeNfa.expr(nfaPair);
nfa = Alloc.NewNfa(MakeNfa.spec);
nfa.Edge = MakeNfa.spec.BOL;
nfa.Next = nfaPair.start;
end = nfaPair.end;
}
else
{
MakeNfa.expr(nfaPair);
nfa = nfaPair.start;
end = nfaPair.end;
}
if (Tokens.AT_EOL == MakeNfa.spec.current_token)
{
MakeNfa.gen.Advance();
NfaPair nfaPair2 = Alloc.NewNLPair(MakeNfa.spec);
end.Next = Alloc.NewNfa(MakeNfa.spec);
Nfa next = end.Next;
next.Next = nfaPair2.start;
next.Sibling = Alloc.NewNfa(MakeNfa.spec);
next.Sibling.Edge = MakeNfa.spec.EOF;
next.Sibling.Next = nfaPair2.end;
end = nfaPair2.end;
num |= 2;
}
if (end == null)
{
Error.ParseError(Errors.ZERO, MakeNfa.gen.InputFilePath, MakeNfa.input.line_number);
}
end.SetAccept(MakeNfa.gen.packAccept());
end.SetAnchor(num);
return(nfa);
}
private static void expr(NfaPair pair)
{
NfaPair nfaPair = Alloc.NewNfaPair();
MakeNfa.cat_expr(pair);
while (Tokens.OR == MakeNfa.spec.current_token)
{
MakeNfa.gen.Advance();
MakeNfa.cat_expr(nfaPair);
Nfa nfa = Alloc.NewNfa(MakeNfa.spec);
nfa.Sibling = nfaPair.start;
nfa.Next = pair.start;
pair.start = nfa;
nfa = Alloc.NewNfa(MakeNfa.spec);
pair.end.Next = nfa;
nfaPair.end.Next = nfa;
pair.end = nfa;
}
}
/*
* Function: factor
* Description: Recursive descent regular expression parser.
*/
private static void factor(NfaPair pair)
{
Nfa start = null;
Nfa end = null;
#if DESCENT_DEBUG
Utility.enter("factor", spec.lexeme, spec.current_token);
#endif
term(pair);
if (Gen.CLOSURE == spec.current_token ||
Gen.PLUS_CLOSE == spec.current_token ||
Gen.OPTIONAL == spec.current_token)
{
start = Alloc.NewNfa(spec);
end = Alloc.NewNfa(spec);
start.SetNext(pair.start);
pair.end.SetNext(end);
if (Gen.CLOSURE == spec.current_token ||
Gen.OPTIONAL == spec.current_token)
{
start.SetSib(end);
}
if (Gen.CLOSURE == spec.current_token ||
Gen.PLUS_CLOSE == spec.current_token)
{
pair.end.SetSib(pair.start);
}
pair.start = start;
pair.end = end;
gen.Advance();
}
#if DESCENT_DEBUG
Utility.leave("factor", spec.lexeme, spec.current_token);
#endif
}
private static void factor(NfaPair pair)
{
MakeNfa.term(pair);
if (Tokens.CLOSURE == MakeNfa.spec.current_token || Tokens.PLUS_CLOSE == MakeNfa.spec.current_token || Tokens.OPTIONAL == MakeNfa.spec.current_token)
{
Nfa nfa = Alloc.NewNfa(MakeNfa.spec);
Nfa nfa2 = Alloc.NewNfa(MakeNfa.spec);
nfa.Next = pair.start;
pair.end.Next = nfa2;
if (MakeNfa.spec.current_token == Tokens.CLOSURE || MakeNfa.spec.current_token == Tokens.OPTIONAL)
{
nfa.Sibling = nfa2;
}
if (MakeNfa.spec.current_token == Tokens.CLOSURE || MakeNfa.spec.current_token == Tokens.PLUS_CLOSE)
{
pair.end.Sibling = pair.start;
}
pair.start = nfa;
pair.end = nfa2;
MakeNfa.gen.Advance();
}
}
/*
* Function: expr
* Description: Recursive descent regular expression parser.
*/
private static void expr(NfaPair pair)
{
NfaPair e2_pair;
Nfa p;
#if DESCENT_DEBUG
Utility.enter("expr", spec.lexeme, spec.current_token);
#endif
#if DEBUG
Utility.assert(null != pair);
#endif
e2_pair = Alloc.NewNfaPair();
cat_expr(pair);
while (Gen.OR == spec.current_token)
{
gen.Advance();
cat_expr(e2_pair);
p = Alloc.NewNfa(spec);
p.SetSib(e2_pair.start);
p.SetNext(pair.start);
pair.start = p;
p = Alloc.NewNfa(spec);
pair.end.SetNext(p);
e2_pair.end.SetNext(p);
pair.end = p;
}
#if DESCENT_DEBUG
Utility.leave("expr", spec.lexeme, spec.current_token);
#endif
}
/*
* function: add_to_dstates
* Description: Takes as input a CBunch with details of
* a dfa state that needs to be created.
* 1) Allocates a new dfa state and saves it in the appropriate Spec list
* 2) Initializes the fields of the dfa state with the information in the CBunch.
* 3) Returns index of new dfa.
*/
private static int add_to_dstates(Spec s, Bunch bunch)
{
Dfa dfa;
#if DEBUG
Utility.assert(null != bunch.GetNFASet());
Utility.assert(null != bunch.GetNFABit());
Utility.assert(null != bunch.GetAccept() || Spec.NONE == bunch.GetAnchor());
#endif
/* Allocate, passing Spec so dfa label can be set. */
dfa = Alloc.NewDfa(s);
/* Initialize fields, including the mark field. */
dfa.SetNFASet(new ArrayList(bunch.GetNFASet()));
dfa.SetNFABit(new BitSet(bunch.GetNFABit()));
dfa.SetAccept(bunch.GetAccept());
dfa.SetAnchor(bunch.GetAnchor());
dfa.ClearMarked();
#if OLD_DUMP_DEBUG
Console.WriteLine("[Created new dfa_state #" + Int32.ToString(dfa.GetLabel()) + "]");
dfa.dump();
#endif
/* Register dfa state using BitSet in spec Hashtable. */
s.dfa_sets[dfa.GetNFABit()] = dfa;
#if OLD_DUMP_DEBUG
Console.Write("Registering set : ");
Print_Set(dfa.GetNFASet());
Console.WriteLine("");
#endif
return(dfa.GetLabel());
}
/*
* Function: term
* Description: Recursive descent regular expression parser.
*/
private static void term(NfaPair pair)
{
Nfa start;
bool isAlphaL;
#if DESCENT_DEBUG
Utility.enter("term", spec.lexeme, spec.current_token);
#endif
if (Gen.OPEN_PAREN == spec.current_token)
{
gen.Advance();
expr(pair);
if (Gen.CLOSE_PAREN == spec.current_token)
{
gen.Advance();
}
else
{
Error.parse_error(Error.E_SYNTAX, input.line_number);
}
}
else
{
start = Alloc.NewNfa(spec);
pair.start = start;
start.SetNext(Alloc.NewNfa(spec));
pair.end = start.GetNext();
if (Gen.L == spec.current_token && Char.IsLetter(spec.lexeme))
{
isAlphaL = true;
}
else
{
isAlphaL = false;
}
if (false == (Gen.ANY == spec.current_token ||
Gen.CCL_START == spec.current_token ||
(spec.ignorecase && isAlphaL)))
{
start.SetEdge(spec.lexeme);
gen.Advance();
}
else
{
start.SetEdge(Nfa.CCL);
start.SetCharSet(new CharSet());
CharSet cset = start.GetCharSet();
/* Match case-insensitive letters using character class. */
if (spec.ignorecase && isAlphaL)
{
cset.addncase(spec.lexeme);
}
/* Match dot (.) using character class. */
else if (Gen.ANY == spec.current_token)
{
cset.add('\n');
cset.add('\r');
/* exclude BOL and EOF from character classes */
cset.add(spec.BOL);
cset.add(spec.EOF);
cset.complement();
}
else
{
gen.Advance();
if (Gen.AT_BOL == spec.current_token)
{
gen.Advance();
/* exclude BOL and EOF from character classes */
cset.add(spec.BOL);
cset.add(spec.EOF);
cset.complement();
}
if (!(Gen.CCL_END == spec.current_token))
{
dodash(cset);
}
}
gen.Advance();
}
}
#if DESCENT_DEBUG
Utility.leave("term", spec.lexeme, spec.current_token);
#endif
}
/*
* Function: rule
* Description: Recursive descent regular expression parser.
*/
private static Nfa rule()
{
NfaPair pair;
Nfa start = null;
Nfa end = null;
int anchor = Spec.NONE;
#if DESCENT_DEBUG
Utility.enter("rule", spec.lexeme, spec.current_token);
#endif
pair = Alloc.NewNfaPair();
if (Gen.AT_BOL == spec.current_token)
{
anchor = anchor | Spec.START;
gen.Advance();
expr(pair);
start = Alloc.NewNfa(spec);
start.SetEdge(spec.BOL);
start.SetNext(pair.start);
end = pair.end;
}
else
{
expr(pair);
start = pair.start;
end = pair.end;
}
if (Gen.AT_EOL == spec.current_token)
{
gen.Advance();
NfaPair nlpair = Alloc.NewNLPair(spec);
end.SetNext(Alloc.NewNfa(spec));
Nfa enext = end.GetNext();
enext.SetNext(nlpair.start);
enext.SetSib(Alloc.NewNfa(spec));
enext.GetSib().SetEdge(spec.EOF);
enext.GetSib().SetNext(nlpair.end);
end = nlpair.end;
anchor = anchor | Spec.END;
}
/* check for null rules */
if (end == null)
{
Error.parse_error(Error.E_ZERO, input.line_number);
}
/* Handle end of regular expression */
end.SetAccept(gen.packAccept());
end.SetAnchor(anchor);
#if DESCENT_DEBUG
Utility.leave("rule", spec.lexeme, spec.current_token);
#endif
return(start);
}
/*
* Function: machine
* Description: Recursive descent regular expression parser.
*/
private static Nfa machine()
{
Nfa start;
Nfa p;
BitSet states;
#if DESCENT_DEBUG
Utility.enter("machine", spec.lexeme, spec.current_token);
#endif
start = Alloc.NewNfa(spec);
p = start;
states = gen.GetStates();
/* Begin: Added for states. */
spec.current_token = Gen.EOS;
gen.Advance();
/* End: Added for states. */
if (Gen.END_OF_INPUT != spec.current_token)
{
p.SetNext(rule());
ProcessStates(states, p.GetNext());
}
while (Gen.END_OF_INPUT != spec.current_token)
{
/* Make state changes HERE. */
states = gen.GetStates();
/* Begin: Added for states. */
gen.Advance();
if (Gen.END_OF_INPUT == spec.current_token)
{
break;
}
/* End: Added for states. */
p.SetSib(Alloc.NewNfa(spec));
p = p.GetSib();
p.SetNext(rule());
ProcessStates(states, p.GetNext());
}
/*
* add pseudo-rules for BOL and EOF
*/
p.SetSib(Alloc.NewNfa(spec));
p = p.GetSib();
p.SetNext(Alloc.NewNfa(spec));
Nfa pnext = p.GetNext();
pnext.SetEdge(Nfa.CCL);
pnext.SetNext(Alloc.NewNfa(spec));
pnext.SetCharSet(new CharSet());
pnext.GetCharSet().add(spec.BOL);
pnext.GetCharSet().add(spec.EOF);
// do-nothing accept rule
pnext.GetNext().SetAccept(new Accept(null, input.line_number + 1));
/* add the pseudo rules */
for (int i = 0; i < spec.states.Count; i++)
{
ArrayList srule = spec.state_rules[i];
srule.Add(pnext);
}
#if DESCENT_DEBUG
Utility.leave("machine", spec.lexeme, spec.current_token);
#endif
return(start);
}
private static void term(NfaPair pair)
{
if (Tokens.OPEN_PAREN == MakeNfa.spec.current_token)
{
MakeNfa.gen.Advance();
MakeNfa.expr(pair);
if (Tokens.CLOSE_PAREN == MakeNfa.spec.current_token)
{
MakeNfa.gen.Advance();
return;
}
Error.ParseError(Errors.SYNTAX, MakeNfa.gen.InputFilePath, MakeNfa.input.line_number);
return;
}
else
{
Nfa nfa = Alloc.NewNfa(MakeNfa.spec);
pair.start = nfa;
nfa.Next = Alloc.NewNfa(MakeNfa.spec);
pair.end = nfa.Next;
bool flag = MakeNfa.spec.current_token == Tokens.LETTER && char.IsLetter(MakeNfa.spec.current_token_value);
if (MakeNfa.spec.current_token != Tokens.ANY && MakeNfa.spec.current_token != Tokens.CCL_START && (!MakeNfa.spec.IgnoreCase || !flag))
{
nfa.Edge = MakeNfa.spec.current_token_value;
MakeNfa.gen.Advance();
return;
}
nfa.Edge = '';
nfa.SetCharSet(new CharSet());
CharSet charSet = nfa.GetCharSet();
if (MakeNfa.spec.IgnoreCase && flag)
{
charSet.addncase(MakeNfa.spec.current_token_value);
}
else
{
if (MakeNfa.spec.current_token == Tokens.ANY)
{
charSet.add(10);
charSet.add(13);
charSet.add((int)MakeNfa.spec.BOL);
charSet.add((int)MakeNfa.spec.EOF);
charSet.complement();
}
else
{
MakeNfa.gen.Advance();
if (MakeNfa.spec.current_token == Tokens.CHAR_CLASS)
{
MakeNfa.gen.Advance();
if (!charSet.AddClass(MakeNfa.spec.class_name.ToLower()))
{
Error.ParseError(Errors.InvalidCharClass, MakeNfa.gen.InputFilePath, MakeNfa.input.line_number);
}
}
else
{
if (MakeNfa.spec.current_token == Tokens.AT_BOL)
{
MakeNfa.gen.Advance();
charSet.add((int)MakeNfa.spec.BOL);
charSet.add((int)MakeNfa.spec.EOF);
charSet.complement();
}
}
if (MakeNfa.spec.current_token != Tokens.CCL_END)
{
MakeNfa.dodash(charSet);
}
}
}
MakeNfa.gen.Advance();
return;
}
}
