Invalid Pointer/Seg Fault

[Swiftle]

Hello there,

I'm writing a dictionary that stores keywords and the pages which contain the keyword. This is a follow-up on a PageRank algorithm we're doing for an assignment. The first part is the read from a file containing all the web addresses and store the in a matrix, and then read the addresses themselves to add even more names in the matrix. Now this part works fine but I'm having trouble with the current assignment. In my dictionary I have and index that stores all the keywords and their pages, A keyword + page is an Entry class in my case. So in the dictionnary I have Entry **Index. I'm not sure if this is the right way to do it or should I have something like Entry *Index[100],

//YANCHEV Konstantin
//Projet 4
//Groupe 3

#include<iostream>
#include<fstream>
#include<cstring>
#include<math.h>
#include<iomanip>
#include <string>
#include"ListeSL.hpp"


using namespace std;

const int MAX_NB_LINK = 500;
const int MAX_SIZE_URL = 300;
const double eps = 10e-5;
const double c = 0.15;
int nmbr_liens = 0;

class Dictionnaire{
	class Entry{
	public:
		string motClef;
		Liste pages;
	};
public:
	Dictionnaire();
	Dictionnaire(const Dictionnaire& orig);
	Dictionnaire& operator=(const Dictionnaire& orig);
	void inserer(char*);
	void inserer(int m, int key);
	int chercher(char*);
	bool estPresent(char*);
	bool contientMot(string str, char ligne[MAX_SIZE_URL]);
	void generateDictionnary(char L[MAX_NB_LINK][MAX_SIZE_URL], char *words, char *urls);
	void afficherIndex() const;
	~Dictionnaire();
	Liste returnPages(char *v);
private:
	int indiceCourant;
	
	Entry **Index;
	
	void effacer();
	void copier(const Dictionnaire& orig);		
};		

Dictionnaire::Dictionnaire(){
	Index = NULL;
	indiceCourant = 0;
}

Dictionnaire::Dictionnaire(const Dictionnaire& orig){
	copier(orig);
}

Dictionnaire& Dictionnaire::operator=(const Dictionnaire& orig){
	if(&orig != this){	
		for(int i = 0; i < indiceCourant; i++){
			Index[i]->pages = orig.Index[i]->pages;
			Index[i]->motClef = orig.Index[i]->motClef;
		}
	}
	return *this;
}

void Dictionnaire::inserer(char* name){
	Index[indiceCourant] = new Entry;
	Index[indiceCourant]->motClef = name;
	indiceCourant++;
}

void Dictionnaire::afficherIndex() const{
	for(int i = 0; i < indiceCourant; i++){
		cout<<"Mot: "<<Index[i]->motClef<<endl;
		cout<<"Pages: ";
		Index[i]->pages.afficherListe();
	}
}

void Dictionnaire::inserer(int m, int key){
	Index[m]->pages.insererEnTete(key);
}

int Dictionnaire::chercher(char* name){
	int ligne;
	bool found = false;
	string str1 = name;
	for(ligne = 0; ligne<indiceCourant and !found; ligne++){
		if(str1.compare(Index[ligne]->motClef) == 0)
			found = true;			
	}
	if(!found){
		cerr<<"Le mots n'est pas dans l'index"<<endl;
		exit(1);
	}
	return ligne-1;
}

bool Dictionnaire::estPresent(char* name){
	string str1 = name;
	bool present = false;
	for(int i = 0; i < indiceCourant and !present; i++){
		if(str1.compare(Index[i]->motClef) == 0)
			present = true;
	}
	return present;
}	

bool Dictionnaire::contientMot(string str, char ligne[MAX_SIZE_URL]){
	bool contient = false;
	//conversion du string en char
	char mot[50];
	int i, longueur = 0;
	for(i = 0; str[i] !='\0';i++){
		mot[i] = str[i];
	}
	mot[i] = '\0';
	//recherche sur la ligne;
	for(int j = 0; ligne[j] != '\0' and !contient; j++){
		for(int k = 0; k < i;k++){
			if(mot[k] == ligne[j+k])
				longueur++;
		}
		if(longueur == i)
			contient = true;
		longueur = 0;
	}
	return contient;
}
	

void Dictionnaire::generateDictionnary(char L[MAX_NB_LINK][MAX_SIZE_URL], char *words, char *urls){
	char URL[MAX_SIZE_URL];
	//1)Rempli Index[] avec les mots
	ifstream file;
	file.open(words);
	while(!file.eof()){
		file.getline(URL, MAX_SIZE_URL);
		if(!estPresent(URL))
			inserer(URL);
	}
	//Cherche Pour le mots dans Index[i] tout les pages webs qui contient ce MC
	for(int mot = 0; mot < indiceCourant; mot++){
		//Ouvre le fichier qui se situe a la Ieme ligne de L pour le mot X
		for(int ligne = 0; ligne < nmbr_liens; ligne++){
			ifstream file2;
			file2.open(L[ligne]);
			bool contient = false;
			//Ouvre la Jeme ligne de la page I
			for(int line_file = 0; !file.eof() and !contient; line_file++){
				file2.getline(URL, MAX_SIZE_URL);
				if(contientMot(Index[mot]->motClef, URL)){
					inserer(mot, ligne);
					contient = true;
				}		
			}
		}
	}
}


Dictionnaire::~Dictionnaire(){
	for(int i = 0; i < indiceCourant; i++){
		delete Index[i];
	}
	delete [] Index;
}

Liste Dictionnaire::returnPages(char *v){
	int index;
	index = chercher(v);
	
	return Index[index]->pages;
}

void Dictionnaire::effacer(){
	for(int i = 0; i < indiceCourant; i++)
		delete Index[i];
	delete [] Index;
}

void Dictionnaire::copier(const Dictionnaire& orig){
}

void calculMu1(double A[MAX_NB_LINK][MAX_NB_LINK], double Mu[MAX_NB_LINK], double eps, int n);

void calculMuGoogle(double A[MAX_NB_LINK][MAX_NB_LINK], double Mu[MAX_NB_LINK], double c, double eps, int n);

int totalLiens(int i,double M[MAX_NB_LINK][MAX_NB_LINK]){
	int Liens = 0;
	for(int indx = 0; indx < nmbr_liens; indx++){
		Liens += M[i][indx];
	}
	return Liens;
}

void MversA(double M[MAX_NB_LINK][MAX_NB_LINK], int n){
	
	double Mu[n], save;
	int LiensParLigne[n];
	for(int indx = 0; indx < n; indx++){
		LiensParLigne[indx] = totalLiens(indx,M) ;
	}
	
	for(int i = 0; i < n; i++){
		for(int j = 0; j <= i; j++){
			if(M[i][j] >=1){
					save = M[i][j];
					M[i][j] = M[j][i]/LiensParLigne[j];					
					M[j][i] = save/LiensParLigne[i];
			}
			else if(M[j][i] >=1){
					save = M[j][i];
					M[j][i] = 0;					
					M[i][j] = save/LiensParLigne[j];
			}
			else{
				M[i][j] = 0;
				M[j][i] = 0;
				}
			}
		}

	cout<<"???METHODE GOOGLE???"<<endl;
	calculMuGoogle(M, Mu, c, eps, n);
	for(int page = 0; page < n; page++){
		cout<<"Page "<<page<<": "<<Mu[page]<<endl;
	}
	cout<<"???METHODE ULB???"<<endl;
	calculMu1(M, Mu, eps, n);

	for(int page = 0; page < n; page++){
		cout<<"Page "<<page<<": "<<setprecision(3)<<Mu[page]<<endl;
	}
	
	/*Affichage de la matrice A */
	/*for(int ligne =0;ligne < nmbr_liens; ligne++){
		cout<<"Lien"<<ligne<<":";
		for(int colonne = 0; colonne < nmbr_liens; colonne++){
			cout<<M[ligne][colonne];
		}
	cout<<endl;
	}*/
}

void calculMu1(double A[MAX_NB_LINK][MAX_NB_LINK], double Mu[MAX_NB_LINK], double eps, int n){


	int iterrations = 0;
	double vectTemp[n]; 
	for(int i=0; i < n; i++){
			vectTemp[i] = 1.0/n;
	}
	bool test;
	do{			
		if(iterrations>0){
			for(int k = 0; k < n; k++){
				vectTemp[k] = Mu[k];
			}
		}
		for(int ligneA = 0; ligneA < n;ligneA ++){
			double Somme = 0.0;
			for(int colA = 0; colA < n; colA++){
				Somme += (A[ligneA][colA] * vectTemp[colA]);
			}
			Mu[ligneA] = Somme;
		}
	iterrations++;		
	test = true;
		for(int j = 0; j < n; j++){
			if(fabs(Mu[j]-vectTemp[j]) >= eps)
				test = false;
		}
	}while(!test);	
	cout<<"Iterrations Mu1: "<<iterrations<<endl;
}

void calculMuGoogle(double A[MAX_NB_LINK][MAX_NB_LINK], double Mu[MAX_NB_LINK], double c, double eps, int n){
	bool test;
	int iterrations = 0;
	double E[n], vectTemp[n];
	/*initialisation e*/
	for(int i=0; i < n; i++){
		E[i] = 1.0/n;
	}
	for(int i=0; i < n; i++){
		vectTemp[i] = 1.0/n;
	}			
	for(int i=0; i < n; i++){
		Mu[i] = 0;
	}
	do{			
		if(iterrations>0){
			for(int k = 0; k < n; k++){
				vectTemp[k] = Mu[k];
			}
		}
		for(int ligneA = 0; ligneA < n;ligneA ++){
			double Somme = 0.0;
			for(int colA = 0; colA < n; colA++){
				Somme += (A[ligneA][colA] * vectTemp[colA]);
			}
			Mu[ligneA] = E[ligneA]*c+(1-c)*Somme;
		}				
		iterrations++;
	test = true;
	for(int j = 0; j < n; j++){
		if(fabs(Mu[j]-vectTemp[j]) >= eps)
			test = false;
		}
	}
	while(!test);
	cout<<"Iterrations MuGoogle: "<<iterrations<<endl;	
}

void cptLiens(char L[MAX_NB_LINK][MAX_SIZE_URL], double M[MAX_NB_LINK][MAX_NB_LINK]){
	char line[MAX_SIZE_URL];
	char lien_tmp[MAX_SIZE_URL];
	int resultat, lien_traitee;
	
	ifstream file;
	file.open("listepages.txt");
	//Rempli la matrice L
	for(; nmbr_liens < MAX_NB_LINK and !file.eof(); nmbr_liens++){
		resultat = 1;
		file.getline(line,MAX_SIZE_URL);
		if(nmbr_liens==0){
			int long_lien;
			for(long_lien=0;long_lien<MAX_SIZE_URL and line[long_lien] !='\0'; 
					long_lien++){
				L[nmbr_liens][long_lien] = line[long_lien];
			}
			L[nmbr_liens][long_lien] = '\0';
		}
		for(int k = 0; k<nmbr_liens and resultat !=0;k++){
			resultat = strcmp(L[k], line);
			if(resultat == 0){
				nmbr_liens--;
			}
			else if(k+1 == nmbr_liens){
				int long_lien;
				for(long_lien=0;long_lien<MAX_SIZE_URL 
						and line[long_lien] !='\0'; long_lien++){
					L[nmbr_liens][long_lien] = line[long_lien];
				}
				L[nmbr_liens][long_lien]='\0';
			}
		}
	}
 	/* A la sortie du cycle le dernier index de L reste vide. Il faut 
 	 * donc enlever l'indice non-rempli*/
 	nmbr_liens--; 
	
	/* Le cycle suivant ce deroule ainsi:
			1)Lis L ligne par ligne.
			2)Cherche <a href..> dans la ligne actuelle.
			3)Lis le lien trouvee.
			4)Compare le lien trouvee aux autres
			5)Si le lien est deja dans L on recuper son indice sinon on le 
			  rajoute a la fin de L.
			6)M[lien_actu][lien_traitee] ++.
	*/
	
	for(int g=0;g < MAX_NB_LINK; g++){
		for(int t=0; t <MAX_NB_LINK; t++){
			M[g][t] = 0.0;
		}
	}
	//ouvre un site qui se trouve sur L[i];
	for(int lien_actu = 0; lien_actu < nmbr_liens; lien_actu++){ 
		ifstream file2;
		file2.open(L[lien_actu]);
		//ouvre la I-eme ligne du site sur L[i];
		for(int lignes_fich = 0; lignes_fich < MAX_NB_LINK and !file2.eof(); 
						lignes_fich++){
			file2.getline(line,MAX_SIZE_URL);
			//parcours la ligne a la recherche de <a href>
			for(int char_line = 0;line[char_line] != '\0'; char_line++){ 
			if(line[char_line] == '<' 
					and (line[char_line+1] =='a' or line[char_line+1]=='A')
					and line[char_line+2] == ' ' 
					and (line[char_line+3] == 'h' or line[char_line+3]=='H')
					and (line[char_line+4] == 'r' or line[char_line+4]=='R')){
				//recopie le lien trouvee dans un vectuer temporaire
				int p=0;
				for(int debut_line = char_line + 9; line[debut_line] != '#' 
						and line[debut_line] !='"';debut_line ++, p++){ 
					lien_tmp[p] = line[debut_line];
				}
				lien_tmp[p] = '\0';
				resultat = 1;
				//compare le lien trouvee aux liens dans L				
				int k;
				for(k = 0; k < nmbr_liens and resultat !=0; k++){	
					resultat = strcmp(lien_tmp,L[k]);
					if(resultat == 0){
						lien_traitee = k;											
					}
				}
				//si c'est un lien nouveau on le met dans L
				if(k == nmbr_liens){
					int long_lien;
					for(long_lien = 0;long_lien < MAX_SIZE_URL 
							and line[long_lien] != '\0';long_lien++){ 
						L[nmbr_liens][long_lien]= lien_tmp[long_lien];
					}
					L[nmbr_liens][long_lien]='\0';
					lien_traitee = nmbr_liens;
					nmbr_liens++;
				}								
				M[lien_actu][lien_traitee]++;
					}
				}
			}	
	}
} 

void affiche(char L[MAX_NB_LINK][MAX_SIZE_URL], double M[MAX_NB_LINK][MAX_NB_LINK]){
	/*Affichage de la matrice L */
	for(int ligne = 0; ligne<nmbr_liens; ligne++){
		cout<<ligne+1<<":"<<L[ligne]<<endl;
	}
	/*Affichage de la matrice M */
	for(int ligne =0;ligne < nmbr_liens; ligne++){
		cout<<"Lien"<<ligne+1<<":";
		for(int colonne = 0; colonne < nmbr_liens; colonne++){
			cout<<M[ligne][colonne];
		}
	cout<<endl;
	}
}

int main(){
	Dictionnaire Dico;
	char words[] = {"motscle.txt"};
	char urls[] = {"listepages.txt"};
	char L[MAX_NB_LINK][MAX_SIZE_URL];
	double M[MAX_NB_LINK][MAX_NB_LINK];
	cptLiens(L, M);
	affiche(L, M);
	
	Dico.generateDictionnary(L, words, urls);
	//Dico.afficherIndex();
	
	
	MversA(M, (nmbr_liens));

}

This is the whole thing. If I use Entry **Index by the time I get to generateDictionnary() I get a Segmentation Fault. If I use Entry *Index[100] I get nothing after affiche(); or I get free() invalid pointer followed by abort. I tried to check it with gdb but since I just picked it up it didn't help me very much, I used valgrind but it showed no memory leaks. I really have no idea what to do. I suspect the problem is with the Entry **Index and maybe the initializing of a Dictionnaire object. Btw the important code is upto line 188, below that it's tested and working.

[Ancient Dragon]

line 59: What will that function do if the parameter Dictionary object gets destroyed?? A: All those pointers will become invalid. It is better if the this object creates its own copy of the Index array so that it will own all the memory instead of relying on the memory allocated by some other dictionary object.

[Swiftle]

line 59: What will that function do if the parameter Dictionary object gets destroyed?? A: All those pointers will become invalid. It is better if the this object creates its own copy of the Index array so that it will own all the memory instead of relying on the memory allocated by some other dictionary object.

Thanks for the advice, I modified my code (I will update it in my original post) so now I can create entries in the Index. However I have a problem with file.eof().
I'm using

while(!file.eof()){
         //do stuff
}

The file I'm using has 18, A minor annoyance that it will do one more loop and store an empty keyword. For example if the the 18th line in my file reads vector it will then create a 19th entry which has an empty keyword. Now the bigger problem is that I get a Segmentation Fault after the eof(). The program tries to exit the while{} but it gets a seg fault. I have no idea what's causing this. Meh can't edit the original code so here is the code:

#include<iostream>
#include<fstream>
#include<cstring>
#include<math.h>
#include<iomanip>
#include <string>
#include"ListeSL.hpp"


using namespace std;

const int MAX_NB_LINK = 500;
const int MAX_SIZE_URL = 300;
const double eps = 10e-5;
const double c = 0.15;
int nmbr_liens = 0;

class Dictionnaire{
	class Entry{
	public:
		string motClef;
		Liste pages;
	};
public:
	Dictionnaire();
	Dictionnaire(const Dictionnaire& orig);
	Dictionnaire& operator=(const Dictionnaire& orig);
	void inserer(char*);
	void inserer(int m, int key);
	int chercher(char*);
	bool estPresent(char*);
	bool contientMot(string str, char ligne[MAX_SIZE_URL]);
	void generateDictionnary(char L[MAX_NB_LINK][MAX_SIZE_URL], char *words, char *urls);
	void afficherIndex() const;
	~Dictionnaire();
	Liste returnPages(char *v);
private:
	int indiceCourant;
	
	Entry *Index[100];
	
	void effacer();
	void copier(const Dictionnaire& orig);		
};		

Dictionnaire::Dictionnaire(){
	//this->Index = NULL;
	this->indiceCourant = 0;
}

Dictionnaire::Dictionnaire(const Dictionnaire& orig){
	copier(orig);
}

Dictionnaire& Dictionnaire::operator=(const Dictionnaire& orig){
	if(&orig != this){	
		for(int i = 0; i < indiceCourant; i++){
			Index[i]->pages = orig.Index[i]->pages;
			Index[i]->motClef = orig.Index[i]->motClef;
		}
	}
	return *this;
}

void Dictionnaire::inserer(char* name){
	Index[indiceCourant] = new Entry;
	Index[indiceCourant]->motClef = name;
	indiceCourant++;
}

void Dictionnaire::afficherIndex() const{
	for(int i = 0; i < indiceCourant; i++){
		cout<<"Mot: "<<Index[i]->motClef<<endl;
		cout<<"Pages: ";
		Index[i]->pages.afficherListe();
	}
}

void Dictionnaire::inserer(int m, int key){
	Index[m]->pages.insererEnTete(key);
}

int Dictionnaire::chercher(char* name){
	int ligne;
	bool found = false;
	string str1 = name;
	for(ligne = 0; ligne<indiceCourant and !found; ligne++){
		if(str1.compare(Index[ligne]->motClef) == 0)
			found = true;			
	}
	if(!found){
		cerr<<"Le mots n'est pas dans l'index"<<endl;
		exit(1);
	}
	return ligne-1;
}

bool Dictionnaire::estPresent(char* name){
	string str1 = name;
	bool present = false;
	for(int i = 0; i < indiceCourant and !present; i++){
		if(str1.compare(Index[i]->motClef) == 0)
			present = true;
	}
	return present;
}	

bool Dictionnaire::contientMot(string str, char ligne[MAX_SIZE_URL]){
	bool contient = false;
	//conversion du string en char
	char mot[50];
	int i, longueur = 0;
	for(i = 0; str[i] !='\0';i++){
		mot[i] = str[i];
	}
	mot[i] = '\0';
	//recherche sur la ligne;
	for(int j = 0; ligne[j] != '\0' and !contient; j++){
		for(int k = 0; k < i;k++){
			if(mot[k] == ligne[j+k])
				longueur++;
		}
		if(longueur == i)
			contient = true;
		longueur = 0;
	}
	return contient;
}
	

void Dictionnaire::generateDictionnary(char L[MAX_NB_LINK][MAX_SIZE_URL], char *words, char *urls){
	char URL[MAX_SIZE_URL];
	//1)Rempli Index[] avec les mots
	ifstream file;
	file.open("motscles.txt");
	
	while(!file.eof()){
		file.getline(URL, MAX_SIZE_URL);
		if(!estPresent(URL)){
			inserer(URL);
			cout<<"OKSORTIE"<<indiceCourant-1<<endl;
		}
	cout<<"test"<<endl;

	}
	cout<<"SORTI";
		/*
	//Cherche Pour le mots dans Index[i] tout les pages webs qui contient ce MC
	for(int mot = 0; mot < indiceCourant; mot++){
		//Ouvre le fichier qui se situe a la Ieme ligne de L pour le mot X
		for(int ligne = 0; ligne < nmbr_liens; ligne++){
			ifstream file2;
			file2.open(L[ligne]);
			bool contient = false;
			//Ouvre la Jeme ligne de la page I
			for(int line_file = 0; !file.eof() and !contient; line_file++){
				file2.getline(URL, MAX_SIZE_URL);
				if(contientMot(Index[mot]->motClef, URL)){
					inserer(mot, ligne);
					contient = true;
				}		
			}
		}
	}*/
}


Dictionnaire::~Dictionnaire(){
	for(int i = 0; i < indiceCourant; i++){
		delete Index[i];
	}
	delete [] Index;
}

Liste Dictionnaire::returnPages(char *v){
	int index;
	index = chercher(v);
	
	return Index[index]->pages;
}

void Dictionnaire::effacer(){
	for(int i = 0; i < indiceCourant; i++)
		delete Index[i];
	delete [] Index;
}

void Dictionnaire::copier(const Dictionnaire& orig){
}

int main(){
	Dictionnaire Dico;
	char words[] = {"motscles.txt"};
	char urls[] = {"listepages.txt"};
	char L[MAX_NB_LINK][MAX_SIZE_URL];
	double M[MAX_NB_LINK][MAX_NB_LINK];
	cptLiens(L, M);
	//affiche(L, M);
	
	Dico.generateDictionnary(L, words, urls);
	//Dico.afficherIndex();
	
	
	//MversA(M, (nmbr_liens));

}

Deleted the unrelated part. Anyway, in generateDicionnary() after the end of while I placed a cout<<"Sortie"; but it never displays as it gets a Seg Fault before exiting.

[Dave Sinkula]

The file I'm using has 18, A minor annoyance that it will do one more loop and store an empty keyword. For example if the the 18th line in my file reads vector it will then create a 19th entry which has an empty keyword.

That's pretty much expected behavior.

Avoid loop control using eof()

[Swiftle]

Thanks Dave! I'll try that later, About the seg fault: I'm getting this in gdb:

0x080492dd in Liste::supprimer ()
(gdb) where
#0  0x080492dd in Liste::supprimer ()
#1  0x08049357 in Liste::effacer ()
#2  0x0804937d in Liste::~Liste ()
#3  0x0804b3b5 in Dictionnaire::Entry::~Entry ()
#4  0x0804aa12 in Dictionnaire::~Dictionnaire ()
#5  0x0804b2bb in main ()

Aparently it tries to call some destructors at the end of the while. But why does this happens Oo Why does the function calls them automaticaly at the end of the while()? Also is there another way to store an Entry other than using
Entry *EntryPtr[]?

[Dave Sinkula]

It's hard for me to tell. The posted code isn't complete (or did I miss the header? or was it all in the first post? but that still seems to #include a header), so I can't build it, which prevents me from even thinking about running it. I don't care to give it a static type of analysis (well, Lint has a lot to say, but I know a lot of that has to do with missing header data).

[Swiftle]

Here are the two files you need:

HEADER

class Liste
{ 
  class Elem 
  {
  public:
    Elem(int, Elem *);
    int info;
    Elem *suiv;
  };
  
public:

  class Place 
  {
  public:
    friend class Liste;
    Place suivant() const;
    int info() const;
    bool valide() const;
    Place();
  private:
    Place(Elem *);
    Elem *elem;
  };
  
  Place tete() const;
  int longueur() const;
  bool vide() const;
  Place insererEnTete(int);
  Place insererApres(Place, int);
  void supprimer(Place);
  void afficherListe() const;
  Place trouver(int) const;
  Liste& operator=(const Liste &);
  Liste();
  Liste(const Liste &);
  ~Liste();
private:
  void effacer();
  void copier(Place);
  Elem *tetePtr;
};

Source

#include <iostream> 
#include "ListeSL.hpp"

using namespace std;

Liste::Elem::Elem(int info, Elem *suiv) 
{
  this->info = info;
  this->suiv = suiv;
}

Liste::Place Liste::Place::suivant() const
{
  return Place(elem->suiv);
}

int Liste::Place::info() const
{
  return elem->info;
}

bool Liste::Place::valide() const
{
  return (elem != NULL);
}

Liste::Place::Place()
{
  this->elem = NULL;
}

Liste::Place::Place(Elem *elem)
{
  this->elem = elem;
}

Liste::Place Liste::tete() const
{
  return Place(tetePtr);
}

int Liste::longueur() const
{
  int res = 0;
  for (Place p = tete(); p.valide(); p = p.suivant())
    ++res;
  return res;
}

bool Liste::vide() const
{
  return !tete().valide();
}

Liste::Place Liste::insererEnTete(int i)
{
  tetePtr = new Elem(i, tetePtr);
  return tete();
}

Liste::Place Liste::insererApres(Place p, int i)
{
  p.elem->suiv = new Elem(i, p.elem->suiv);
  return p.suivant();
}

void Liste::supprimer(Place p)
{
  Elem *i = p.elem;
  if (i == tetePtr)
    tetePtr = tetePtr->suiv;
  else
  {
    Elem *j = tetePtr, *k = NULL;
    while (j != i)
    {
      k = j;
      j = j->suiv;
    }
    k->suiv = i->suiv;
  }
  delete i;
}

Liste::Place Liste::trouver(int i) const
{
  Place p;
  for (p = tete(); p.valide() && p.info() != i; p = p.suivant())
    ;
  return p;
}

void Liste::afficherListe() const
{
 
  for (Place p = tete(); p.valide(); p = p.suivant())
    cout<<p.info()<<"  ";
  cout<<endl;
  
}

Liste& Liste::operator=(const Liste &l)
{
  if (&l != this) 
  {
    effacer();
    copier(l.tete());
  }
  return *this;
}

Liste::Liste()
{
  tetePtr = NULL;
}

Liste::Liste(const Liste &l)
{
  copier(l.tete());
}

Liste::~Liste()
{
  effacer();
}

void Liste::effacer()
{
  while (!vide()) 
    supprimer(tete());
}

void Liste::copier(Place p)
{
  if (p.valide()) 
  {
    insererEnTete(p.info());
    p = p.suivant();
  }
  Place q = tete();
  while(p.valide())
  {
    q = insererApres(q,p.info());
    p = p.suivant();
  }
}

[thomas_naveen]

Are you allocating memory to Index?

I can see "Index[Counter] = new"

But I don't see "Index = new" anywhere or it's equivalent.

[Swiftle]

You mean that if I have Entry *Index[100] I have to do Index[0] = new Entry in the default constructor?

If I use Entry **Index I do Index = new *Entry[100];

[thomas_naveen]

My apologies. I was looking at the original post. I believe you have got it right in the post after AD's post.

[Swiftle]

If I use Entry *Index[100];

I get

#0  0xb7f99410 in __kernel_vsyscall ()
#1  0xb7d40085 in raise () from /lib/tls/i686/cmov/libc.so.6
#2  0xb7d41a01 in abort () from /lib/tls/i686/cmov/libc.so.6
#3  0xb7d78b7c in ?? () from /lib/tls/i686/cmov/libc.so.6
#4  0xb7d80a85 in ?? () from /lib/tls/i686/cmov/libc.so.6
#5  0xb7d844f0 in free () from /lib/tls/i686/cmov/libc.so.6
#6  0xb7f4cb11 in operator delete () from /usr/lib/libstdc++.so.6
#7  0xb7f4cb6d in operator delete[] () from /usr/lib/libstdc++.so.6
#8  0x080498d3 in Dictionnaire::effacer ()
#9  0x080498e7 in Dictionnaire::~Dictionnaire ()
#10 0x0804a0d5 in main ()

If I use Entry **Index I get

#0  0x0804a6b9 in Liste::supprimer ()
#1  0x0804a733 in Liste::effacer ()
#2  0x0804a759 in Liste::~Liste ()
#3  0x0804a287 in Dictionnaire::Entry::~Entry ()
#4  0x08049917 in Dictionnaire::effacer ()
#5  0x08049973 in Dictionnaire::~Dictionnaire ()
#6  0x0804a1c0 in main ()

It's a problem withthe destructors because If I remove them everything works fine.

[Swiftle]

Sorry, double post =/

[mitrmkar]

>> If I use Entry *Index[100];
Assuming the destructor still looks like

Dictionnaire::~Dictionnaire(){
	for(int i = 0; i < indiceCourant; i++){
		delete Index[i];
	}
	delete [] Index;
}

delete [] is guaranteed not to work, so rather use Entry ** Index or otherwise remove delete [] Index altogether.

Then there is this Dictionnaire::effacer() which duplicates what the destructor is doing. This is a somewhat suspicious construct, because if both of these methods are in effect, you'll be double-freeing memory, which you must not do.

Perhaps post the latest code (that compiles) and the input files that you are using.

[Swiftle]

~Dictionnaire() is supposed to call void effacer in order to work. I probably copy-pasted too much but in my current source

~Dictionnaire()

calls

void effacer()

. About the delete: what's the problem with d

elete [] Index

? Index is a pointer to a pointer. When I initialize the dict I'm doing

Index = new Entry*[100]

. With

delete [] Index[i]

I'm deleting each pointer in the Entry array,

delete [] Index

deletes the pointer to the array. Atleast that's what we're being told at school :P Anyway after playing with gdb it seems that all the problems are linked with the usage of the class Liste. It fails to destroy the list properly If I create my own destructor, it fails to add an entry to the list, it fails to display it. Also the STRANGEST thing so far, if I modify Entry->pages it somehow manages to screw up Entry->Mots. By changing it aswell Oo which I have no idea how it happens.

EDIT: To answer your question I'm using Entry **Index thus the delete [] Index.

[mitrmkar]

>> About the delete: what's the problem with delete [] Index;
At one point, you had declared Index as

Entry * Index[100];

i.e. not dynamically allocated and hence impossible to be delete [] d.

>> To answer your question I'm using Entry **Index thus the delete [] Index
That's OK then.

But delete [] Index[i] looks bad. If Index[i] is a pointer to a single instance of Entry, you must do delete Index[i] instead.

All in all, you seem to be having serious dynamic memory management issues (that's when the strangest things start to happen), I hope you get those solved. And if you get tired and weary solving these problems, maybe post back (with code that compiles + the input files).