I'm trying to do a implementation of a deque with dynamic allocation, but I'm having some trouble since the values of the variables in the struct are different out of the function initialize, and I dont know why.
By some reason, the program always end up with a segmentation fault in the push_front/back part.
// C program for vetor implementation of d
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
typedef int data;
// A structure to represent a d
typedef struct deque
{
int front, rear, size;
unsigned int capacidade;
data* vetor;
}deque;
// function to create a d of given capacidade.
// It initializes size of d as 0
void initialize(deque *d, unsigned int capacidade){
d = (deque*) malloc(sizeof(deque));
d->capacidade = capacidade;
d->front = 0;
d->size = 0;
d->rear = capacidade-1; // This is important, see the enqueue
d->vetor = (data*) malloc(d->capacidade * sizeof(data));
}
// deque is full when size becomes equal to the capacidade
int full(deque* d){
if(d->size == d->capacidade)
return 1;
else return 0;
}
// deque is empty when size is 0
int empty(deque* d){
return (d->size == 0);
}
// Function to add an item to the d.
// It changes rear and size
int push_back(deque* d, int item){
if (full(&d))
return 0;
d->rear = (d->rear + 1)%d->capacidade;
d->vetor[d->rear] = item;
d->size = d->size + 1;
}
int push_front(deque* d, int item){
if (full(&d))
return 0;
d->front = (d->front - 1+d->capacidade)%d->capacidade;
d->vetor[d->front] = item;
d->size = d->size + 1;
}
// Function to remove an item from d.
// It changes front and size
int pop_front(deque* d){
if (empty(&d))
return 0;
int item = d->vetor[d->front];
d->front = (d->front + 1)%d->capacidade;
d->size = d->size - 1;
return item;
}
int pop_back(deque* d){
if (empty(&d))
return 0;
int item = d->vetor[d->rear];
d->rear = (d->rear - 1+d->capacidade)%d->capacidade;
d->size = d->size - 1;
return item;
}
// Function to get front of d
int front(deque* d)
{
if (empty(d))
return 0;
return d->vetor[d->front];
}
// Function to get rear of d
int rear(deque* d)
{
if (empty(d))
return 0;
return d->vetor[d->rear];
}
// Driver program to test above functions./
int main()
{
deque* d;
int operacoes=0, tamdeque=0, i=0;
char opcao[100];
scanf("%d %d", &operacoes, &tamdeque);
initialize(&d, tamdeque);
while(i<=operacoes){
scanf("%s", opcao);
if(!strcmp(opcao, "insereI")){
if(full(&d)){
printf("cheia\n");
}
else{
data item;
scanf("%d", &item);
printf("%u\n", &d->capacidade);
push_front(&d, item);
}
}
else if(!strcmp(opcao,"insereF")){
if(full(&d)){
printf("cheia\n");
}
else{
data item;
scanf("%d", &item);
push_back(&d, item);
}
}
else if(!strcmp(opcao, "removeI")){
if(empty(&d)){
printf("vazia\n");
}
else{
pop_front(&d);
}
}
else if(!strcmp(opcao, "removeF")){
if (empty(&d)){
printf("vazia\n");
}
else{
pop_back(&d);
}
}
i++;
}
return 0;
}
Mis-match arguments
In many places, the following type of error. Readily findable will all warnings enabled.
warning: passing argument 1 of 'pop_back' from incompatible pointer type [-Wincompatible-pointer-types]
int push_back(deque* d, int item){
// if (full(&d))
if (full(d))
main()
// deque* d;
deque d;
initialize(&d, tamdeque);
mixing int/unsigned math.
Recommend a design change to use just one.
Missing return value
Minor: int push_back(), push_front()
Other problems exist. (About 30 total warnings)
Related
I'm making simple patient managing program using circular queue but q.rear always have "0" value while executing exit_hos()
I thought that addq() makes variable "rear" different, but It doesn't work.
is_empty() always return front and rear is same.
I think I'm misunderstanding some codes and memory concepts.
how can I fix these functions?
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#define MAX_SIZE 50
#define MAX_QUEUE_SIZE 6
typedef struct {
char** value;
int front;
int rear;
} Queue;
void init_queue(Queue* q) {
q->value = (char**)malloc(sizeof(char*) * MAX_QUEUE_SIZE);
q->front = 0;
q->rear = 0;
}
int is_full(Queue* q) {
if (((q->rear +1) % MAX_QUEUE_SIZE) == q->front)
return 1;
else
return 0;
}
int is_empty(Queue* q) {
if (q->front == q->rear)
return 1;
else
return 0;
}
void addq(Queue* q, char* value) {
q->rear = (q->rear+1) % MAX_QUEUE_SIZE;
q->value[q->rear] = value;
printf("addq: %s", value);
return;
}
char* deleteq(Queue* q) {
q->front = (q->front + 1) % MAX_QUEUE_SIZE;
return q->value[q->front];
}
void arrive(Queue q) {
int input;
char name[MAX_SIZE];
printf("\n");
printf("1. submit\n");
printf("2. cancel\n");
scanf("%d", &input);
if (input == 1) {
if (is_full(&q) == 1) {
printf("Service is not available\n");
}
else {
printf("name: ");
scanf("%s", name);
addq(&q, name);
}
}
else if (input == 2) {
return;
}
else {
printf("input error\n");
return;
}
return;
}
void exit_hos(Queue q) {
char patient[MAX_SIZE];
if (is_empty(&q) == 1)
{
printf("There is no patient waiting\n");
}
else {
strcpy(patient, deleteq(&q));
printf("patient: %s", patient);
}
return;
}
int main() {
int input;
Queue q;
init_queue(&q);
while (1)
{
printf("\nINPUT\n");
printf("1. Arrive hostpital\n");
printf("2. Exit hospital\n");
printf("3. service exit\n");
scanf("%d", &input);
if (input == 1)
arrive(q);
else if (input == 2) {
exit_hos(q);
}
else if (input == 3) {
printf("exit\n");
return 0;
}
else {
printf("input error\n");
}
}
free(q.value);
return 0;
}
I think that this line is wrong:
q->value = (char**)malloc(sizeof(char*) * MAX_QUEUE_SIZE);
I think that it should be:
char * _value = (char*)malloc(sizeof(char*) * MAX_QUEUE_SIZE);
q->value = &_value;
malloc is going to return a pointer to a char array. q->value is a pointer to a pointer to a char array. So you want to set it to the address of the char array that malloc is created for you.
Change you init_queue code to this and it will work:
void init_queue(Queue* q) {
char * _value = (char*)malloc(sizeof(char*) * MAX_QUEUE_SIZE);
q->value = &_value;
q->front = 0;
q->rear = 0;
}
Output:
Chris#DESKTOP-BCMC1RF ~
$ ./main.exe
INPUT
1. Arrive hostpital
2. Exit hospital
3. service exit
1
1. submit
2. cancel
1
name: fred
addq: fred
INPUT
1. Arrive hostpital
2. Exit hospital
3. service exit
2
If you already have a max queue size and a max size, you are better off pre-allocating the whole thing as an array, reducing memory headaches. As a general rule, avoid headaches unless they provide a feature you want.
Note: This method of keeping track of and re-using memory is called a circular buffer (not to be confused with the linked list types that are more commonly called queues).
#define MAX_SIZE 50
#define MAX_QUEUE_SIZE 6
typedef struct {
char value [MAX_QUEUE_SIZE][MAX_SIZE + 1]; //+1 to hold extra null termination
unsigned int front;
unsigned int size; //size is a clearer than rear, which could have meant end item or end+1 and needed special empty queue handling
} Queue;
void init_queue(Queue* q) {
memset(q,0,sizeof(Queue)); //just zero it all
//more info on this and some situation-dependent alternatives https://stackoverflow.com/questions/11152160/initializing-a-struct-to-0
}
int is_full(const Queue* q) {
return q->size >= MAX_QUEUE_SIZE;
}
int is_empty(const Queue* q) {
return q->size == 0;
}
//sometimes called a push operation
//return 0 if failed
int addq(Queue* q, const char* value) {
//error check, abort, error handling section:
//full queue -> abort
if(is_full(q)) return 0;
//long value -> truncate handled via strncpy
//actual operation
const unsigned int destination = (q->front + q->size) % MAX_QUEUE_SIZE;
strncpy(q->value[destination],value,MAX_SIZE);
q->size = q->size + 1;
printf("addq: %s", q->value[destination]);
return q->size;
}
//sometimes called a pop operation
//return value may not persist if addq is called, but fine for your use of copying on call
const char* deleteq(Queue* q) {
if(is_empty(q)) return 0;
const char * retval = q->value[q->front];
q->front = (q->front + 1) % MAX_QUEUE_SIZE;
q->size = q->size - 1;
return retval;
}
also remember to use either MAX_SIZE + 1 or strncpy with MAX_SIZE - 1 since "No null-character is implicitly appended at the end of destination if source is longer than num."
(and strcpy and scanf as you sling them onto arrays is unsafe)
I am trying to find a formula for insert front in a dynamic circular array. One problem I have is when try to run display in which case the inx after the first two insertion will be off because there is an unoccupied index. This is assuming that I am following the correct way of choosing the start index.
void insertCDAfront(CDA *items,void *val){//insert in the slot prior t
assert(items->array!=0);
if(sizeCDA(items)==0){
items->array[items->start]=val;
items->size++;
}
else{
if(items->size==items->capacity){
items->capacity=items->capacity*items->factor;
void **arr=calloc(items->capacity,sizeof(void *));
assert(arr!=0);
for(int i=0;i<items->size;i++){
arr[i]=getCDA(items,i);
}
items->array=arr;
items->start=0;
}
items->start=(items->start-1+items->capacity)%items->capacity;
items->array[items->start]=val;
items->start=0;
items->size++;
}
}
void *getCDA(CDA *items,int index){//
assert (index<items->size&&index>=0);
int spot=(items->start+index+items->capacity)%items->capacity;
return items->array[spot];
}
void displayCDA(FILE *f,CDA *items){
int i=0;
if (items->size==0){
fprintf(f,"("")");
}
else if (items->size==1){
fprintf(f,"(");
items->display(f,items->array[i]);
fprintf(f,")");
}
else{
fprintf(f,"%s","(");
while (i!=items->size-1){ //i!=end
items->display(f,getCDA(items,i));
fprintf(f,"%s",",");
i++;
}
items->display(f,getCDA(items,i));
fprintf(f,"%s",")");
}
}
sounds like you are trying to make a ring buffer... Here's an academic starting point:
#define PREDETERMINED_ERROR_VALUE (0)
#define MAX_SIZE (100)
int ring_buffer[MAX_SIZE] = {0};
unsigned int head = 0;
unsigned int tail = 0;
void add_item(int new_data)
//!\todo check for overrun
{ring_buffer[head++] = new_data;}
unsigned int get_size_of_ring_buffer(void)
{
unsigned int ret;
if tail <= head
ret = head - tail;
else
ret = head + MAX_SIZE - tail;
return (ret);
}
int get_item(void)
{
//!\todo consider returning error code and modifying a param with the data
int ret;
if (tail != head)
ret = ring_buffer[tail++];
else
ret = PREDETERMINED_ERROR_VALUE;
return (ret);
}
I am trying to implement a generic hash structure that can support any type of data and any hash function.
A wrote the code and try to run it, it dosn't work, it breaks. I try to debug it and there it works well. I don't know where the problem is?
Here is the code that I used for implementing the structure:
The "hash.h" file:
typedef struct tip_hash_nod
{
void *info;
struct tip_hash_nod *urm;
}NOD_LISTA_HASH;
typedef struct
{
NOD_LISTA_HASH *Table;
int size;
int sizeMemory;
int (*hash)(const void *obiect,const int m);
void (*distruge)(void *obiect);
}*HASH;
void initializare_hash(HASH *h,int size,int (*hash_dat)(const void *obiect,const int m),void (*distruge)(void *obiect));
int hash_insert(HASH *h,void *obiect,int sizeOfObiect);
int hash_search(HASH h,void *obiect,int (*compara)(const void *a,const void *b));
void hash_delete(HASH *h);
And the "hash.c" file:
void initializare_hash(HASH *h,int size,int (*hash_dat)(const void *obiect,const int m),void (*distruge)(void *obiect))
{
int i;
(*h) = (HASH)malloc(sizeof(HASH));
(*h)->sizeMemory = size;
if(size != 0)
{
(*h)->Table = (NOD_LISTA_HASH *)malloc((*h)->sizeMemory * sizeof(NOD_LISTA_HASH));
for(i=0;i<(*h)->sizeMemory;i++)
{
(*h)->Table[i].info = NULL;
(*h)->Table[0].urm = NULL;
}
}
else
{
(*h)->Table = (NOD_LISTA_HASH *)malloc(sizeof(NOD_LISTA_HASH));
(*h)->Table[0].info = NULL;
(*h)->Table[0].urm = NULL;
(*h)->sizeMemory = 1;
}
(*h)->size = 0;
(*h)->hash = hash_dat;
(*h)->distruge = distruge;
}
int hash_insert(HASH *h,void *obiect,int sizeOfObiect)
{
int i,poz;
NOD_LISTA_HASH *p;
if((*h)->size == (*h)->sizeMemory)
{
HASH h1;
initializare_hash(&h1,2*(*h)->sizeMemory,(*h)->hash,(*h)->distruge);
for(i=0;i<(*h)->sizeMemory;i++)
{
if((*h)->Table[i].info != NULL)
hash_insert(&h1,(*h)->Table[i].info,sizeOfObiect);
p=(*h)->Table[i].urm;
while(p!=NULL)
{
hash_insert(&h1,p->info,sizeOfObiect);
p = p->urm;
}
}
hash_delete(h);
*h=h1;
return hash_insert(h,obiect,sizeOfObiect);
}
else
{
poz = (*h)->hash(obiect,(*h)->sizeMemory);
if((*h)->Table[poz].info == NULL)
{
(*h)->Table[poz].info = malloc(sizeOfObiect);
memcpy((*h)->Table[poz].info,obiect,sizeOfObiect);
(*h)->Table[poz].urm = NULL;
(*h)->size++;
}
else
{
p = &((*h)->Table[poz]);
while(p->urm!=NULL)
p = p->urm;
p->urm = (NOD_LISTA_HASH *)malloc(sizeof(NOD_LISTA_HASH));
p = p->urm;
p->info = malloc(sizeOfObiect);
memcpy(p->info,obiect,sizeOfObiect);
p->urm = NULL;
}
return poz;
}
}
int hash_search(HASH h,void *obiect,int (*compara)(const void *a,const void *b))
{
int poz;
NOD_LISTA_HASH *p;
poz = h->hash(obiect,h->sizeMemory);
if(h->Table[poz].info == NULL)
return -1;
else
if(compara(h->Table[poz].info,obiect)==0)
return poz;
else
{
p=h->Table[poz].urm;
while(p != NULL)
{
if(compara(p->info,obiect)==0)
return poz;
p = p->urm;
}
return -1;
}
}
static void distruge_lista(NOD_LISTA_HASH *p,void (*distruge_obiect)(void *obiect))
{
if(p->urm != NULL)
distruge_lista(p->urm,distruge_obiect);
else
{
if(p->info != NULL)
distruge_obiect(p->info);
free(p);
}
}
void hash_delete(HASH *h)
{
int i;
for(i=0;i<(*h)->sizeMemory;i++)
{
if((*h)->Table[i].info != NULL && (*h)->Table[i].urm != NULL)
{
distruge_lista((*h)->Table[i].urm,(*h)->distruge);
}
}
free((*h)->Table);
*h = NULL;
}
And this is my "main.c" file:
#include <stdio.h>
#include <stdlib.h>
#include <conio.h>
#include "hash.h"
int comparare(const void *a,const void *b)
{
return (*(int *)a - *(int *)b);
}
int hash(const void *obiect,int m)
{
return (*(int *)obiect) % m;
}
void distruge_obiect(void *obiect)
{
free((int *)obiect);
}
int main()
{
HASH h;
int val,error;
initializare_hash(&h,0,hash,distruge_obiect);
val = 20;
hash_insert(&h,&val,sizeof(int));
val = 800;
hash_insert(&h,&val,sizeof(int));
val = 2000;
hash_insert(&h,&val,sizeof(int));
val = 765;
hash_insert(&h,&val,sizeof(int));
val = 800;
error = hash_search(h,&val,comparare);
if(error == -1)
printf("Elementul %d nu se afla in hash.\n",val);
else
printf("Elementul %d se afla pe pozitia: %d.\n",val,error);
hash_delete(&h);
getch();
return 0;
}
How I already sad if I try to debug it works with no problem, but when I run it, it crashes. I can onely make an assumption that it can not dealocate the memory or something. My call stack loocks like this:
You've dropped a pretty big pile of code on us, without much to go on. I had a quick look anyway, and noticed this incorrect allocation:
(*h) = (HASH)malloc(sizeof(HASH));
HASH is a pointer type, so you are allocating only enough memory for one pointer. You want to allocate memory for the thing to which it points:
*h = malloc(sizeof(**h));
(The cast is not required in C, and some folks around here will be strident about not using one.)
That error would be entirely enough to cause all manner of bad behavior. In particular, the erroneous code might seem to work until you dynamically allocate more memory and write to that, so perhaps that explains why your tests crash on the second insertion.
Hello world (hi people),
First, I'd say this is my first post, so please be clement.
As the title says, I've a heap corruption when I wants to free my object(s). I passed a couple of hours trying to fix it but I just can't see what's wrong even though I'm sure it's obvious!
So that's why I come to you.
My goal is to create some functions to (poorly) mimic some std::vector function in C. All objects are dynamically created. The implementation may be tricky due to the massive use of pointers.
I got the heap corruption when I free the vector object but I don't know if it comes from the pushBack or the destroy function.
Feel free to ask more information. Constructive comments are welcome!
Here is some piece of code:
Header content:
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
typedef struct _array
{
unsigned int size;
int *pData;
} tArray, *ptArray;
typedef struct _vector
{
unsigned int size;
ptArray *pData;
} tVector, *ptVector;
ptArray createArray(unsigned int size);
void destroyArray(ptArray *pArray);
ptVector createVector();
int pushBackArray(ptVector pVector, ptArray pArrayToAppend);
int popBackArray(ptVector pVector);
void destroyVector(ptVector *pVector);
Main content:
int main(char argc, char *argv[])
{
unsigned int size = 5, i;
time_t seed = NULL;
ptArray pArr = createArray(size);
ptVector pVec = createVector();
srand(seed);
for(i = 0; i < pArr->size; i++)
{
pArr->pData[i] = rand() % 9 + 1;
}
//destroyVector(&pVec); // Works at this point
pushBackArray(pVec, pArr);
destroyArray(&pArr);
destroyVector(&pVec); // Heap corruption on free pVector->pData
getchar();
return 0;
}
Body content:
ptArray createArray(unsigned int size)
{
ptArray pArray = (ptArray)calloc(1, sizeof(tArray));
if(pArray)
{
pArray->pData = (int*)malloc(size * sizeof(int));
if(pArray->pData)
pArray->size = size;
}
return pArray;
}
void destroyArray(ptArray *pArray)
{
if(pArray)
{
free((*pArray)->pData);
free((*pArray));
(*pArray) = NULL;
}
}
ptVector createVector()
{
ptVector pVector = (ptVector)calloc(1, sizeof(*pVector));
return pVector;
}
int pushBackArray(ptVector pVector, ptArray pArrayToAppend)
{
int res = 0;
if(pVector && pArrayToAppend)
{
pVector->pData = (ptArray*) realloc(pVector->pData
, sizeof(ptArray) * pVector->size + 1);
if(pVector->pData)
{
pVector->pData[pVector->size] = createArray(pArrayToAppend->size);
if(pVector->pData[pVector->size])
{
memcpy(pVector->pData[pVector->size]->pData
, pArrayToAppend->pData
, pVector->pData[pVector->size]->size * sizeof(int));
pVector->size++;
}
else
{
pVector->pData = (ptArray*) realloc(pVector->pData
, sizeof(ptArray) * pVector->size);
res = 3;
}
}
else
res = 2;
}
else
res = 1;
return res;
}
void destroyVector(ptVector *pVector)
{
if(pVector)
{
unsigned int i;
for(i = 0; i < (*pVector)->size; i++)
destroyArray( &((*pVector)->pData[i]) );
free((*pVector)->pData); // Heap Corruption throw
free(*pVector);
(*pVector) = NULL;
}
}
specifying the size is incorrect at function pushBackArray
sizeof(ptArray) * pVector->size + 1
to
sizeof(ptArray) * (pVector->size + 1)
Edit:
Hash.c is updated with revisions from the comments, I am still getting a Seg fault. I must be missing something here that you guys are saying
I have created a hash table ADT using C but I am encountering a segmentation fault when I try to call a function (find_hash) in the ADT.
I have posted all 3 files that I created parse.c, hash.c, and hash.h, so you can see all of the variables. We are reading from the file gettysburg.txt which is also attached
The seg fault is occuring in parse.c when I call find_hash. I cannot figure out for the life of me what is going on here. If you need anymore information I can surely provide it.
sorry for the long amount of code I have just been completely stumped for a week now on this. Thanks in advance
The way I run the program is first:
gcc -o parse parse.c hash.c
then: cat gettysburg.txt | parse
Parse.c
#include <stdio.h>
#include <ctype.h>
#include <string.h>
#include "hash.h"
#define WORD_SIZE 40
#define DICTIONARY_SIZE 1000
#define TRUE 1
#define FALSE 0
void lower_case_word(char *);
void dump_dictionary(Phash_table );
/*Hash and compare functions*/
int hash_func(char *);
int cmp_func(void *, void *);
typedef struct user_data_ {
char word[WORD_SIZE];
int freq_counter;
} user_data, *Puser_data;
int main(void)
{
char c, word1[WORD_SIZE];
int char_index = 0, dictionary_size = 0, num_words = 0, i;
int total=0, largest=0;
float average = 0.0;
Phash_table t; //Pointer to main hash_table
int (*Phash_func)(char *)=NULL; //Function Pointers
int (*Pcmp_func)(void *, void *)=NULL;
Puser_data data_node; //pointer to hash table above
user_data * find;
printf("Parsing input ...\n");
Phash_func = hash_func; //Assigning Function pointers
Pcmp_func = cmp_func;
t = new_hash(1000,Phash_func,Pcmp_func);
// Read in characters until end is reached
while ((c = getchar()) != EOF) {
if ((c == ' ') || (c == ',') || (c == '.') || (c == '!') || (c == '"') ||
(c == ':') || (c == '\n')) {
// End of a word
if (char_index) {
// Word is not empty
word1[char_index] = '\0';
lower_case_word(word1);
data_node = (Puser_data)malloc(sizeof(user_data));
strcpy(data_node->word,word1);
printf("%s\n", data_node->word);
//!!!!!!SEG FAULT HERE!!!!!!
if (!((user_data *)find_hash(t, data_node->word))){ //SEG FAULT!!!!
insert_hash(t,word1,(void *)data_node);
}
char_index = 0;
num_words++;
}
} else {
// Continue assembling word
word1[char_index++] = c;
}
}
printf("There were %d words; %d unique words.\n", num_words,
dictionary_size);
dump_dictionary(t); //???
}
void lower_case_word(char *w){
int i = 0;
while (w[i] != '\0') {
w[i] = tolower(w[i]);
i++;
}
}
void dump_dictionary(Phash_table t){ //???
int i;
user_data *cur, *cur2;
stat_hash(t, &(t->total), &(t->largest), &(t->average)); //Call to stat hash
printf("Number of unique words: %d\n", t->total);
printf("Largest Bucket: %d\n", t->largest);
printf("Average Bucket: %f\n", t->average);
cur = start_hash_walk(t);
printf("%s: %d\n", cur->word, cur->freq_counter);
for (i = 0; i < t->total; i++)
cur2 = next_hash_walk(t);
printf("%s: %d\n", cur2->word, cur2->freq_counter);
}
int hash_func(char *string){
int i, sum=0, temp, index;
for(i=0; i < strlen(string);i++){
sum += (int)string[i];
}
index = sum % 1000;
return (index);
}
/*array1 and array2 point to the user defined data struct defined above*/
int cmp_func(void *array1, void *array2){
user_data *cur1= array1;
user_data *cur2= array2;//(user_data *)array2;
if(cur1->freq_counter < cur2->freq_counter){
return(-1);}
else{ if(cur1->freq_counter > cur2->freq_counter){
return(1);}
else return(0);}
}
hash.c
#include "hash.h"
Phash_table new_hash (int size, int(*hash_func)(char*), int(*cmp_func)(void*, void*)){
int i;
Phash_table t;
t = (Phash_table)malloc(sizeof(hash_table)); //creates the main hash table
t->buckets = (hash_entry **)malloc(sizeof(hash_entry *)*size); //creates the hash table of "size" buckets
t->size = size; //Holds the number of buckets
t->hash_func = hash_func; //assigning the pointer to the function in the user's program
t->cmp_func = cmp_func; // " "
t->total=0;
t->largest=0;
t->average=0;
t->sorted_array = NULL;
t->index=0;
t->sort_num=0;
for(i=0;i<size;i++){ //Sets all buckets in hash table to NULL
t->buckets[i] = NULL;}
return(t);
}
void free_hash(Phash_table table){
int i;
hash_entry *cur;
for(i = 0; i<(table->size);i++){
if(table->buckets[i] != NULL){
for(cur=table->buckets[i]; cur->next != NULL; cur=cur->next){
free(cur->key); //Freeing memory for key and data
free(cur->data);
}
free(table->buckets[i]); //free the whole bucket
}}
free(table->sorted_array);
free(table);
}
void insert_hash(Phash_table table, char *key, void *data){
Phash_entry new_node; //pointer to a new node of type hash_entry
int index;
new_node = (Phash_entry)malloc(sizeof(hash_entry));
new_node->key = (char *)malloc(sizeof(char)*(strlen(key)+1)); //creates the key array based on the length of the string-based key
new_node->data = data; //stores the user's data into the node
strcpy(new_node->key,key); //copies the key into the node
//calling the hash function in the user's program
index = table->hash_func(key); //index will hold the hash table value for where the new node will be placed
table->buckets[index] = new_node; //Assigns the pointer at the index value to the new node
table->total++; //increment the total (total # of buckets)
}
void *find_hash(Phash_table table, char *key){
int i;
hash_entry *cur;
printf("Inside find_hash\n"); //REMOVE
for(i = 0;i<table->size;i++){
if(table->buckets[i]!=NULL){
for(cur = table->buckets[i]; cur->next != NULL; cur = cur->next){
if(strcmp(table->buckets[i]->key, key) == 0)
return((table->buckets[i]->data));} //returns the data to the user if the key values match
} //otherwise return NULL, if no match was found.
}
return NULL;
}
void stat_hash(Phash_table table, int *total, int *largest, float *average){
int node_num[table->size]; //creates an array, same size as table->size(# of buckets)
int i,j, count = 0;
int largest_buck = 0;
hash_entry *cur;
for(i = 0; i < table->size; i ++){
if(table->buckets[i] != NULL){
for(cur=table->buckets[i]; cur->next!=NULL; cur = cur->next){
count ++;}
node_num[i] = count;
count = 0;}
}
for(j = 0; j < table->size; j ++){
if(node_num[j] > largest_buck)
largest_buck = node_num[j];}
*total = table->total;
*largest = largest_buck;
*average = (table->total) / (table->size);
}
void *start_hash_walk(Phash_table table){
Phash_table temp = table;
int i, j, k;
hash_entry *cur; //CHANGE IF NEEDED to HASH_TABLE *
if(table->sorted_array != NULL) free(table->sorted_array);
table->sorted_array = (void**)malloc(sizeof(void*)*(table->total));
for(i = 0; i < table->total; i++){
if(table->buckets[i]!=NULL){
for(cur=table->buckets[i]; cur->next != NULL; cur=cur->next){
table->sorted_array[i] = table->buckets[i]->data;
}}
}
for(j = (table->total) - 1; j > 0; j --) {
for(k = 1; k <= j; k ++){
if(table->cmp_func(table->sorted_array[k-1], table->sorted_array[k]) == 1){
temp -> buckets[0]-> data = table->sorted_array[k-1];
table->sorted_array[k-1] = table->sorted_array[k];
table->sorted_array[k] = temp->buckets[0] -> data;
}
}
}
return table->sorted_array[table->sort_num];
}
void *next_hash_walk(Phash_table table){
table->sort_num ++;
return table->sorted_array[table->sort_num];
}
hash.h
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
typedef struct hash_entry_ { //Linked List
void *data; //Generic pointer
char *key; //String-based key value
struct hash_entry_ *next; //Self-Referencing pointer
} hash_entry, *Phash_entry;
typedef struct hash_table_ {
hash_entry **buckets; //Pointer to a pointer to a Linked List of type hash_entry
int (*hash_func)(char *);
int (*cmp_func)(void *, void *);
int size;
void **sorted_array; //Array used to sort each hash entry
int index;
int total;
int largest;
float average;
int sort_num;
} hash_table, *Phash_table;
Phash_table new_hash(int size, int (*hash_func)(char *), int (*cmp_func)(void *, void *));
void free_hash(Phash_table table);
void insert_hash(Phash_table table, char *key, void *data);
void *find_hash(Phash_table table, char *key);
void stat_hash(Phash_table table, int *total, int *largest, float *average);
void *start_hash_walk(Phash_table table);
void *next_hash_walk(Phash_table table);
Gettysburg.txt
Four score and seven years ago, our fathers brought forth upon this continent a new nation: conceived in liberty, and dedicated to the proposition that all men are created equal.
Now we are engaged in a great civil war. . .testing whether that nation, or any nation so conceived and so dedicated. . . can long endure. We are met on a great battlefield of that war.
We have come to dedicate a portion of that field as a final resting place for those who here gave their lives that that nation might live. It is altogether fitting and proper that we should do this.
But, in a larger sense, we cannot dedicate. . .we cannot consecrate. . . we cannot hallow this ground. The brave men, living and dead, who struggled here have consecrated it, far above our poor power to add or detract. The world will little note, nor long remember, what we say here, but it can never forget what they did here.
It is for us the living, rather, to be dedicated here to the unfinished work which they who fought here have thus far so nobly advanced. It is rather for us to be here dedicated to the great task remaining before us. . .that from these honored dead we take increased devotion to that cause for which they gave the last full measure of devotion. . . that we here highly resolve that these dead shall not have died in vain. . . that this nation, under God, shall have a new birth of freedom. . . and that government of the people. . .by the people. . .for the people. . . shall not perish from the earth.
It's possible that one of several problems with this code are loops like:
for(table->buckets[i];
table->buckets[i]->next != NULL;
table->buckets[i] = table->buckets[i]->next)
...
The initializing part of the for loop (table->buckets[i]) has no effect. If i is 0 and table->buckets[0] == NULL, then the condition on this loop (table->buckets[i]->next != NULL) will dereference a null pointer and crash.
That's where your code seemed to be crashing for on my box, at least. When I changed several of your loops to:
if (table->buckets[i] != NULL) {
for(;
table->buckets[i]->next != NULL;
table->buckets[i] = table->buckets[i]->next)
...
}
...it kept crashing, but in a different place. Maybe that will help get you unstuck?
Edit: another potential problem is that those for loops are destructive. When you call find_hash, do you really want all of those buckets to be modified?
I'd suggest using something like:
hash_entry *cur;
// ...
if (table->buckets[i] != NULL) {
for (cur = table->buckets[i]; cur->next != NULL; cur = cur->next) {
// ...
}
}
When I do that and comment out your dump_dictionary function, your code runs without crashing.
Hmm,
here's hash.c
#include "hash.h"
Phash_table new_hash (int size, int(*hash_func)(char*), int(*cmp_func)(void*, void*)){
int i;
Phash_table t;
t = (Phash_table)calloc(1, sizeof(hash_table)); //creates the main hash table
t->buckets = (hash_entry **)calloc(size, sizeof(hash_entry *)); //creates the hash table of "size" buckets
t->size = size; //Holds the number of buckets
t->hash_func = hash_func; //assigning the pointer to the function in the user's program
t->cmp_func = cmp_func; // " "
t->total=0;
t->largest=0;
t->average=0;
for(i=0;t->buckets[i] != NULL;i++){ //Sets all buckets in hash table to NULL
t->buckets[i] = NULL;}
return(t);
}
void free_hash(Phash_table table){
int i;
for(i = 0; i<(table->size);i++){
if(table->buckets[i]!=NULL)
for(table->buckets[i]; table->buckets[i]->next != NULL; table->buckets[i] = table->buckets[i]->next){
free(table->buckets[i]->key); //Freeing memory for key and data
free(table->buckets[i]->data);
}
free(table->buckets[i]); //free the whole bucket
}
free(table->sorted_array);
free(table);
}
void insert_hash(Phash_table table, char *key, void *data){
Phash_entry new_node; //pointer to a new node of type hash_entry
int index;
new_node = (Phash_entry)calloc(1,sizeof(hash_entry));
new_node->key = (char *)malloc(sizeof(char)*(strlen(key)+1)); //creates the key array based on the length of the string-based key
new_node->data = data; //stores the user's data into the node
strcpy(new_node->key,key); //copies the key into the node
//calling the hash function in the user's program
index = table->hash_func(key); //index will hold the hash table value for where the new node will be placed
table->buckets[index] = new_node; //Assigns the pointer at the index value to the new node
table->total++; //increment the total (total # of buckets)
}
void *find_hash(Phash_table table, char *key){
int i;
hash_entry *cur;
printf("Inside find_hash\n"); //REMOVE
for(i = 0;i<table->size;i++){
if(table->buckets[i]!=NULL){
for (cur = table->buckets[i]; cur != NULL; cur = cur->next){
//for(table->buckets[i]; table->buckets[i]->next != NULL; table->buckets[i] = table->buckets[i]->next){
if(strcmp(cur->key, key) == 0)
return((cur->data));} //returns the data to the user if the key values match
} //otherwise return NULL, if no match was found.
}
return NULL;
}
void stat_hash(Phash_table table, int *total, int *largest, float *average){
int node_num[table->size];
int i,j, count = 0;
int largest_buck = 0;
hash_entry *cur;
for(i = 0; i < table->size; i ++)
{
if(table->buckets[i]!=NULL)
for (cur = table->buckets[i]; cur != NULL; cur = cur->next){
//for(table->buckets[i]; table->buckets[i]->next != NULL; table->buckets[i] = table->buckets[i]->next){
count ++;}
node_num[i] = count;
count = 0;
}
for(j = 0; j < table->size; j ++){
if(node_num[j] > largest_buck)
largest_buck = node_num[j];}
*total = table->total;
*largest = largest_buck;
*average = (table->total) /(float) (table->size); //oook: i think you want a fp average
}
void *start_hash_walk(Phash_table table){
void* temp = 0; //oook: this was another way of overwriting your input table
int i, j, k;
int l=0; //oook: new counter for elements in your sorted_array
hash_entry *cur;
if(table->sorted_array !=NULL) free(table->sorted_array);
table->sorted_array = (void**)calloc((table->total), sizeof(void*));
for(i = 0; i < table->size; i ++){
//for(i = 0; i < table->total; i++){ //oook: i don't think you meant total ;)
if(table->buckets[i]!=NULL)
for (cur = table->buckets[i]; cur != NULL; cur = cur->next){
//for(table->buckets[i]; table->buckets[i]->next != NULL; table->buckets[i] = table->buckets[i]->next){
table->sorted_array[l++] = cur->data;
}
}
//oook: sanity check/assert on expected values
if (l != table->total)
{
printf("oook: l[%d] != table->total[%d]\n",l,table->total);
}
for(j = (l) - 1; j > 0; j --) {
for(k = 1; k <= j; k ++){
if (table->sorted_array[k-1] && table->sorted_array[k])
{
if(table->cmp_func(table->sorted_array[k-1], table->sorted_array[k]) == 1){
temp = table->sorted_array[k-1]; //ook. changed temp to void* see assignment
table->sorted_array[k-1] = table->sorted_array[k];
table->sorted_array[k] = temp;
}
}
else
printf("if (table->sorted_array[k-1] && table->sorted_array[k])\n");
}
}
return table->sorted_array[table->sort_num];
}
void *next_hash_walk(Phash_table table){
/*oook: this was blowing up since you were incrementing past the size of sorted_array..
NB: *you **need** to implement some bounds checking here or you will endup with more seg-faults!!*/
//table->sort_num++
return table->sorted_array[table->sort_num++];
}
here's parse.c
#include <stdio.h>
#include <ctype.h>
#include <string.h>
#include <assert.h> //oook: added so you can assert ;)
#include "hash.h"
#define WORD_SIZE 40
#define DICTIONARY_SIZE 1000
#define TRUE 1
#define FALSE 0
void lower_case_word(char *);
void dump_dictionary(Phash_table );
/*Hash and compare functions*/
int hash_func(char *);
int cmp_func(void *, void *);
typedef struct user_data_ {
char word[WORD_SIZE];
int freq_counter;
} user_data, *Puser_data;
int main(void)
{
char c, word1[WORD_SIZE];
int char_index = 0, dictionary_size = 0, num_words = 0, i;
int total=0, largest=0;
float average = 0.0;
Phash_table t; //Pointer to main hash_table
int (*Phash_func)(char *)=NULL; //Function Pointers
int (*Pcmp_func)(void *, void *)=NULL;
Puser_data data_node; //pointer to hash table above
user_data * find;
printf("Parsing input ...\n");
Phash_func = hash_func; //Assigning Function pointers
Pcmp_func = cmp_func;
t = new_hash(1000,Phash_func,Pcmp_func);
// Read in characters until end is reached
while ((c = getchar()) != EOF) {
if ((c == ' ') || (c == ',') || (c == '.') || (c == '!') || (c == '"') ||
(c == ':') || (c == '\n')) {
// End of a word
if (char_index) {
// Word is not empty
word1[char_index] = '\0';
lower_case_word(word1);
data_node = (Puser_data)calloc(1,sizeof(user_data));
strcpy(data_node->word,word1);
printf("%s\n", data_node->word);
//!!!!!!SEG FAULT HERE!!!!!!
if (!((user_data *)find_hash(t, data_node->word))){ //SEG FAULT!!!!
dictionary_size++;
insert_hash(t,word1,(void *)data_node);
}
char_index = 0;
num_words++;
}
} else {
// Continue assembling word
word1[char_index++] = c;
}
}
printf("There were %d words; %d unique words.\n", num_words,
dictionary_size);
dump_dictionary(t); //???
}
void lower_case_word(char *w){
int i = 0;
while (w[i] != '\0') {
w[i] = tolower(w[i]);
i++;
}
}
void dump_dictionary(Phash_table t){ //???
int i;
user_data *cur, *cur2;
stat_hash(t, &(t->total), &(t->largest), &(t->average)); //Call to stat hash
printf("Number of unique words: %d\n", t->total);
printf("Largest Bucket: %d\n", t->largest);
printf("Average Bucket: %f\n", t->average);
cur = start_hash_walk(t);
if (!cur) //ook: do test or assert for null values
{
printf("oook: null== (cur = start_hash_walk)\n");
exit(-1);
}
printf("%s: %d\n", cur->word, cur->freq_counter);
for (i = 0; i < t->total; i++)
{//oook: i think you needed these braces
cur2 = next_hash_walk(t);
if (!cur2) //ook: do test or assert for null values
{
printf("oook: null== (cur2 = next_hash_walk(t) at i[%d])\n",i);
}
else
printf("%s: %d\n", cur2->word, cur2->freq_counter);
}//oook: i think you needed these braces
}
int hash_func(char *string){
int i, sum=0, temp, index;
for(i=0; i < strlen(string);i++){
sum += (int)string[i];
}
index = sum % 1000;
return (index);
}
/*array1 and array2 point to the user defined data struct defined above*/
int cmp_func(void *array1, void *array2){
user_data *cur1= array1;
user_data *cur2= array2;//(user_data *)array2;
/* ooook: do assert on programmatic errors.
this function *requires non-null inputs. */
assert(cur1 && cur2);
if(cur1->freq_counter < cur2->freq_counter){
return(-1);}
else{ if(cur1->freq_counter > cur2->freq_counter){
return(1);}
else return(0);}
}
follow the //ooks
Explanation:
There were one or two places this was going to blow up in.
The quick fix and answer to your question was in parse.c, circa L100:
cur = start_hash_walk(t);
printf("%s: %d\n", cur->word, cur->freq_counter);
..checking that cur is not null before calling printf fixes your immediate seg-fault.
But why would cur be null ? ~because of this bad-boy:
void *start_hash_walk(Phash_table table)
Your hash_func(char *string) can (& does) return non-unique values. This is of course ok except that you have not yet implemented your linked list chains. Hence you end up with table->sorted_array containing less than table->total elements ~or you would if you were iterating over all table->size buckets ;)
There are one or two other issues.
For now i hacked Nate Kohl's for(cur=table->buckets[i]; cur->next != NULL; cur=cur->next) further, to be for(cur=table->buckets[i]; cur != NULL; cur=cur->next) since you have no chains. But this is *your TODO so enough said about that.
Finally. note that in next_hash_walk(Phash_table table) you have:
table->sort_num++
return table->sorted_array[table->sort_num];
Ouch! Do check those array bounds!
Notes
1) If you're function isn't designed to change input, then make the input const. That way the compiler may well tell you when you're inadvertently trashing something.
2) Do bound checking on your array indices.
3) Do test/assert for Null pointers before attempting to use them.
4) Do unit test each of your functions; never write too much code before compiling & testing.
5) Use minimal test-data; craft it such that it limit-tests your code & attempts to break it in cunning ways.
6) Do initialise you data structures!
7)Never use egyptian braces ! {
only joking ;)
}
PS Good job so far ~> pointers are tricky little things! & a well asked question with all the necessary details so +1 and gl ;)
(//oook: maybe add a homework tag)