public Node doParse()
{
//% is epsilon.
Stack<string> tokens = new Stack<string>();
tokens.Push("$");
tokens.Push("<MiniRE-program>");
Stack<Node> nodes = new Stack<Node>();
Node miniRE = new Node();
while (tokens.Count != 0)
{
string topElem = tokens.Pop();
string nextToken = scanner.peekToken();
if (nextToken == "$")
{
if (topElem == "$")
break;
else
throw new RDParser.SyntaxError("All tokens not popped off stack");
}
if (topElem.Contains('<') && topElem.Contains('>'))
{
//This is a non-term, we need to look up and replace.
List<string> newrule;
try
{
if (isID(nextToken))
newrule = parsetable.getRuleMatchingToken(topElem, "ID");
else if(isREGEX(nextToken))
newrule = parsetable.getRuleMatchingToken(topElem, "REGEX");
else if(isASCIISTR(nextToken))
newrule = parsetable.getRuleMatchingToken(topElem, "ASCII-STR");
else
newrule = parsetable.getRuleMatchingToken(topElem, nextToken);
}
catch (KeyNotFoundException e)
{
throw new RDParser.SyntaxError("Invalid token in position.");
}
//Take the existing rule, create its node, add to top elem on stack, pop top elem if full.
Node newElem = GetNodeFromRuleName(topElem);
if (!(newrule.Count == 0 && newrule[0] == "%"))
{
if (nodes.Count != 0 && nodes.Peek().IsFull)
nodes.Pop();
if (nodes.Count != 0)
nodes.Peek().Nodes.Add(newElem);
if (nodes.Count != 0 && nodes.Peek().IsFull && nodes.Peek() is MiniRE)
{
miniRE = nodes.Pop();
}
else if (nodes.Count != 0 && nodes.Peek().IsFull)
nodes.Pop();
if (!newElem.IsFull && newrule[0] != "%")
{
nodes.Push(newElem);
}
newrule = new List<string>(newrule);
newrule.Reverse();
if (newrule.Count > 0 && newrule[0] != "%")
{
foreach (string s in newrule)
{
tokens.Push(s);
}
}
}
else
{
StringNode sn = new StringNode();
sn.Token = "%";
newElem.Nodes.Add(sn);
nodes.Peek().Nodes.Add(newElem);
if (nodes.Peek().IsFull && nodes.Peek() is MiniRE)
miniRE = nodes.Pop();
}
}
else
{
nextToken = scanner.getToken();
//Is a terminal; must match top of stack, possibly do stuff with AST depending on token.
if (topElem == nextToken)
{
switch (nextToken)
{
case "maxfreqstring":
AssignmentStatement xyz = (AssignmentStatement)nodes.Peek();
xyz.Type = AssignmentStatementType.MaxFreqString;
break;
case "#":
AssignmentStatement xyza = (AssignmentStatement)nodes.Peek();
xyza.Type = AssignmentStatementType.Length;
break;
case "replace":
OtherStatement abc = (OtherStatement)nodes.Peek();
abc.Mode = OtherStatementMode.Replace;
break;
case "print":
OtherStatement bcd = (OtherStatement)nodes.Peek();
bcd.Mode = OtherStatementMode.Print;
break;
case "recursivereplace":
OtherStatement cde = (OtherStatement)nodes.Peek();
cde.Mode = OtherStatementMode.RecursiveReplace;
break;
case "diff":
BinOp op1 = (BinOp)nodes.Peek();
StringNode n1 = new StringNode();
n1.Token = "diff";
op1.Nodes.Add(n1);
break;
case "union":
BinOp op2 = (BinOp)nodes.Peek();
StringNode n2 = new StringNode();
n2.Token = "union";
op2.Nodes.Add(n2);
break;
case "inters":
BinOp op3 = (BinOp)nodes.Peek();
StringNode n3 = new StringNode();
n3.Token = "inters";
op3.Nodes.Add(n3);
break;
default:
break;
}
}
else if (topElem == "REGEX" && nextToken[0] == '\'' && nextToken[nextToken.Length - 1] == '\'')
{
Regex re = new Regex();
re.Pattern = nextToken.Substring(1, nextToken.Length - 2);
nodes.Peek().Nodes.Add(re);
if (nodes.Peek().IsFull)
nodes.Pop();
}
else if (topElem == "ASCII-STR" && nextToken[0] == '"' && nextToken[nextToken.Length - 1] == '"')
{
Node top = nodes.Peek();
if (top is OtherStatement)
{
OtherStatement nd = (OtherStatement)top;
nd.ReplaceText = nextToken.Substring(1, nextToken.Length-2);
}
else if (top is Filename)
{
Filename fn = (Filename)top;
fn.Path = nextToken.Substring(1, nextToken.Length-2);
}
if (top.IsFull)
nodes.Pop();
}
else if (topElem == "ID")
{
StringNode sn = new StringNode();
sn.Token = nextToken;
nodes.Peek().Nodes.Add(sn);
if (nodes.Peek().IsFull)
nodes.Pop();
}
else
{
throw new RDParser.SyntaxError("Token did not match top of LL(1) Parse Stack.");
}
}
}
return miniRE;
}
public void Union()
{
//begin
//x = (find 'a(b|c)' in abc.txt) union (find 'bc in abc.txt)
//end
SymbolTable table = new SymbolTable();
MiniRE miniRe = new MiniRE();
StatementList sl = new StatementList();
miniRe.StatementList = sl;
Statement statement = new Statement();
statement.Id = new StringNode("x");
sl.Statement = statement;
AssignmentStatement as1 = new AssignmentStatement();
statement.AssignmentStatement = as1;
Exp exp = new Exp();
as1.Exp = exp;
#region First find
Term t1 = new Term();
exp.Term = t1;
Regex r1 = new Regex();
r1.Pattern = "([a-zA-Z])*ment([a-zA-Z])*";
t1.Regex = r1;
Filename file1 = new Filename();
file1.Path = "../../../TestFiles/file1.txt";
t1.Filename = file1;
#endregion
ExpTail expTail = new ExpTail();
exp.Tail = expTail;
BinOp bop = new BinOp();
expTail.Binop = bop;
StringNode operation = new StringNode("union");
bop.Operation = operation;
#region Second find
Term t2 = new Term();
expTail.Term = t2;
ExpTail expTail2 = new ExpTail();
expTail.Tail = expTail2;
Regex r2 = new Regex();
r2.Pattern = "([a-zA-Z])*(a|A)([a-zA-Z])*";
t2.Regex = r2;
Filename file2 = new Filename();
file2.Path = "../../../TestFiles/file2.txt";
t2.Filename = file2;
#endregion
miniRe.Execute(table);
StringMatchList matches = (StringMatchList) table["x"];
Assert.AreEqual(8, matches.Length);
String expected = "{\"mentor\"<'../../../TestFiles/file1.txt', 1>" +
"\"filament\"<'../../../TestFiles/file1.txt', 1>" +
"\"argument\"<'../../../TestFiles/file1.txt', 1>" +
"\"argumentative\"<'../../../TestFiles/file1.txt', 1><'../../../TestFiles/file2.txt', 1>" +
"\"Tournament\"<'../../../TestFiles/file1.txt', 1>" +
"\"Argument\"<'../../../TestFiles/file2.txt', 1>" +
"\"predicament\"<'../../../TestFiles/file2.txt', 1>" +
"\"apple\"<'../../../TestFiles/file2.txt', 1>" +
"}";
Assert.AreEqual(expected, matches.ToString());
}