fixed bug that happens when the tmpImage was overflowed. Now tmpImage (in DFA.cpp) is a wstring.

Author: d00mas

examples/simple/simple.grm

@@ -13,7 +13,7 @@ StringLiteral = ''{String Ch}*''
 NumberLiteral = {Digit}+('.'{Digit}+)?
 
 <Statements> ::= <Statement> <Statements>
-               | <Statement>
+               |
 
 <Statement>  ::= display <Expression>
                | display <Expression> read ID
@@ -38,6 +38,7 @@ NumberLiteral = {Digit}+('.'{Digit}+)?
 <Mult Exp>    ::= <Mult Exp> '*' <Negate Exp> 
                 | <Mult Exp> '/' <Negate Exp> 
                 | <Negate Exp> 
+              
 
 <Negate Exp>  ::= '-' <Value> 
                 | <Value> 

include/CharacterSetTable.h

@@ -21,14 +21,13 @@
  #include "wchar.h"
  #include "misc.h"
 
- 
 
  class CharacterSetTable {
    public:
    integer nbrEntries;
    wchar_t **characters;
 
-   CharacterSetTable (integer nbrEntries);
+   CharacterSetTable  (integer nbrEntries);
    ~CharacterSetTable ();
 
  };

include/DFA.h

@@ -43,16 +43,16 @@
    ~DFA ();
 
    bool scan (wchar_t *text);
-   bool scan (char *text);
+   bool scan (char	  *text);
 
    vector <Token*> &getTokens ();
 
    ErrorTable *getErrors();
 
  private:
-   const DFAStateTable *stateTable;
-   const SymbolTable *symbolTable;
-   const CharacterSetTable *characterSetTable;
+   const DFAStateTable		*stateTable;
+   const SymbolTable		*symbolTable;
+   const CharacterSetTable	*characterSetTable;
    integer startState;
    bool caseSensitive;
 

include/DFAStateTable.h

@@ -40,7 +40,6 @@
 
    DFAStateTable (integer nbrEntries);
    ~DFAStateTable ();
-
  };
 
  #endif

include/LALR.h

@@ -30,13 +30,6 @@
  #include "Terminal.h"
  #include "NonTerminal.h" 
 
-/*
- #define REDUCTION_OK 0
- #define REDUCTION_ERROR 1
- #define REDUCTION_TEXT_ACCEPTED 2
- #define REDUCTION_SIMPLYFIED 3
-*/
-
  #if defined (WIN32) && defined (_USRDLL)
     class __declspec(dllexport) LALR;
  #endif
@@ -55,15 +48,15 @@ using namespace std;
    const RuleTable *ruleTable, integer startState);
    ~LALR ();
 
-    /*!
+   /*!
      setup the parsing engine.
    */
    void init (const vector <Token*> &tokens);
 
    /*
-     parse just until ONE reduction is performed
+     parses just until ONE reduction is performed
    */
-   Symbol *nextReduction (bool trimReduction, bool reportOnlyOneError);
+   Symbol *nextReduction (bool trimReductions, bool reportOnlyOneError);
 
    /*!
      Gets the result constant for the last reduction attempt.
@@ -78,9 +71,7 @@ using namespace std;
    */
    Symbol *parse (const vector <Token*> &tokens, bool trimReductions, bool reportOnlyOneError);
 
- //  void symplifyParseTree (Reduction *reduction);
-
-   void printRule (integer rule);
+   void printRule          (integer rule);
    void printReductionTree (Symbol *reduction, int deep);
 
    ErrorTable *getErrors ();

include/SymbolTable.h

@@ -20,10 +20,8 @@
  #include "misc.h"
  #include <wchar.h>
 
-
  enum SymbolType {TERMINAL = 1, NON_TERMINAL = 0};
 
-
  typedef struct SymbolStruct {
    wchar_t *name;
    SymbolType kind;
@@ -37,7 +35,6 @@
 
    SymbolTable (integer nbrEntries);
    ~SymbolTable ();
-
  };
 
 

src/ASTCreator.cpp

@@ -59,6 +59,34 @@
 		rdcChildren = ((NonTerminal*) reduction)->children;
 	 } 
 
+     /*
+        If the symbol constants are included it is possible to use a switch-case 
+        instead.
+        switch (reduction->symbolIndex) {
+
+            // <If Statement> ::= if <Expression> then <StatementList> end
+            case RULE_IF_THEN_END_STATEMENT:
+                CREATE_NODE (IfStatement, ifStmt);
+                    
+                ifStmt->addChild (getASTNode(rdcChildren[1], ifStmt));
+                ifStmt->addChild (getASTNode(rdcChildren[3], ifStmt));
+	
+                return ifStmt;
+
+            // <If Statement> ::= if <Expression> then <StatementList> else <StatementList> end
+            case RULE_IF_THEN_END_ELSE_STATEMENT:
+                CREATE_NODE (IfStatement, ifStatement);
+
+                ifStmt->addChild (getASTNode(rdcChildren[1], ifStmt));
+                ifStmt->addChild (getASTNode(rdcChildren[3], ifStmt));
+	            ifStmt->addChild (getASTNode(rdcChildren[5], ifStmt));
+                return ifStmt;
+            default: return searchEquivNode( rdcChildren, parent);
+        }
+
+     */
+
+
 	 /*
 	  Next is a colection of if statements that checks for the
 	  equivalency between nodes in the reduction tree and in the

src/CGTFile.cpp

@@ -62,7 +62,7 @@
 
     cgtStream.open (filename, ifstream::in | ifstream::binary);
 
-    if (cgtStream.bad()) {
+    if (((void*) cgtStream) == NULL) {
         return false;
     }
 

src/DFA.cpp

@@ -58,7 +58,8 @@
    integer currentLine = 1, currentCol = 1, tokenBeginCol = 1;
 
    Token *t = NULL;
-   wchar_t tmpImage[64];
+  
+   wstring tmpImage;
    wchar_t *tokenSymbol, *tokenImage;
    wchar_t currentChar;
    integer currentState = startState;
@@ -85,8 +86,9 @@
             errorTab->addError (ERROR_SCAN, END_COMMENT_NOT_FOUND, 
                                 NULL, currentLine, tokenBeginCol);
             currentState = startState;
-            imgIndex = 0;
-            return false;
+           
+            tmpImage.clear();
+			return false;
          } else {
             run = false;
          }
@@ -96,7 +98,7 @@
        currentCol = 1;
        if (commentLine) {
          commentLine = false;
-       //  i++;
+       
          if (text[i] == 0) {
            break;
          }
@@ -111,8 +113,7 @@
 
      if (!commentLine) {
 
-       tmpImage[imgIndex] = text[i];
-       imgIndex++;
+	   tmpImage.append (1,text[i]);
 
        // Check which is the next state
        integer e;
@@ -159,12 +160,12 @@
              // Terminal symbol
              case 1 :
              if (!commentBlock) {
-               tmpImage[imgIndex-1] = 0;
+              // tmpImage[imgIndex-1] = 0;
                tokenSymbol = new wchar_t [wcslen(symbolTable->symbols[index2].name)+1];
-               tokenImage = new wchar_t [wcslen (tmpImage)+1];
+               tokenImage = new wchar_t [tmpImage.length()+1];
 
                wcscpy (tokenSymbol, symbolTable->symbols[index2].name);
-               wcscpy (tokenImage, tmpImage);
+               wcscpy (tokenImage, tmpImage.c_str());
 
                t = new Token();
                t->symbol = tokenSymbol;
@@ -206,14 +207,16 @@
              break;
            }
            currentState = startState;
-           imgIndex = 0;
-         } else {
+       
+		   tmpImage.clear();
+		 } else {
            if (!commentBlock) {
              // Set error and return
              errorTab->addError (ERROR_SCAN, UNKNOWN_TOKEN, NULL, currentLine, tokenBeginCol);
              currentState = startState;
-             imgIndex = 0;
-             return false;
+     
+			 tmpImage.clear();            
+			 return false;
            }
            tokenBeginCol = currentCol;
          }

src/LALR.cpp

@@ -87,10 +87,10 @@
 
   /return next reduction or NULL if error or test accepted
 
-  /sa getResult (), buildParseTree ()
+  /sa getResult (), parse ()
  */
 
- Symbol *LALR::nextReduction (bool trimReduction, bool reportOnlyOneError) {
+ Symbol *LALR::nextReduction (bool trimReductions, bool reportOnlyOneError) {
    Action *actObj;
    Token *tok;
    NonTerminal  *newNonTerminal;
@@ -202,7 +202,7 @@
          // User can decide to simplify this by enabling the trimming
          if ((ruleTable->rules[target].symbols.size() == 1) &&
             (symbolTable->symbols[ruleTable->rules[target].symbols[0]].kind ==
-             NON_TERMINAL) && trimReduction) {
+             NON_TERMINAL) && trimReductions) {
                 trim = true; 
 				newNonTerminal->trimmed = true;
          } else {
@@ -264,16 +264,6 @@
            return NULL;
          }
 
-		 /*
-         if (trim) {
-            reductionResult = REDUCTION_SIMPLIFIED;
-         //   return NULL;
-            return newNonTerminal;
-         } else {
-            reductionResult = REDUCTION_OK;
-            return newNonTerminal;
-         }
-		 */
 		 return newNonTerminal;
          break;