Creating a Dictionary in C - c

Question

I am currently working on creating a dictionary using a binary search tree-like structure we designed in class.
#include <ctype.h>
#include <stdio.h>
#include <stdlib.h>
#include <strings.h>
struct entry
{
char* word;
unsigned int n; /* n is the number of times the word appears in the source. */
struct entry *left;
struct entry *right;
};
/*input_from_args: if no additional argument is given, return stdin. Else, open the text file and read it.*/
FILE*
input_from_args(int argc, const char *argv[]){
if(argc==1){
return stdin;
}else{
return fopen(argv[1],"r");
}
}
Below is the insert function that we also wrote in my class. Given the new word we are looking at, if it is
struct entry*
insert(struct entry *table, char* str)
{
if(table == NULL){
table = (struct entry *)malloc(sizeof(struct entry));
strcpy(table->word,str);
table -> n = 1;
table -> left = NULL;
table -> right = NULL;
}else if(strcmp(str, table->word) == 0){
table -> n = (table ->n)+1;
}else if(strcmp(str, table->word) <0){
table->left = insert(table->left, str);
}else if(strcmp(str, table->word) >0){
table ->right = insert(table->right, str);
}
return table;
}
Below is a print function which I wrote myself which is to print every word in table and N, the number of times it occurs.
void
print_table(struct entry *table){
if(table!=NULL){
print_table(table->left);
printf("%s ", table->word);
printf("%d \n", table->n);
print_table(table->right);
}
}
And finally, below is the main function.
int
main(int argc, const char *argv[])
{
FILE *src = input_from_args(argc, argv);
if(src == NULL){
fprintf(stderr, "%s: unable to open %s\n", argv[0], argv[1]);
exit(EXIT_FAILURE);
}
char str[1024];
struct entry *table;
int c;
while((fscanf(src, "%s", str))!= EOF){
table = insert(table, str);
}
print_table(table);
return 0;
}
I'm having some very odd behavior when I run this function. It seems to only be happening when I run it with longer input.
When I run it with this input(in a .txt file):
This is a test.
This is a test.
This is a test.
I get the following output:
This 3
a 3
is 3
test 3
This is what I should be getting. However, when I give it slightly longer input, such as:
Apple Apple
Blue Blue
Cat Cat
Dog Dog
Elder Elder
Funions Funions
Gosh Gosh
Hairy Hairy
I get the following output:
Appme 2
Blue 2
Cat 2
Dog 2
Elder 2
Funions 2
Gosi 2
Hairy 2
Which is clearly correct as far as the numbers go, but why is it changing some of the letters in my words? I gave it Apple, it returned Appme. I gave it Gosh, it gave me Gosi. What's going on with my code that I am missing?

Answer 1

This line in the insert function is very problematic:
strcpy(table->word,str);
It's problematic because you don't actually allocate memory for the string. That means that table->word is uninitialized and its value will be indeterminate, so the strcpy call will lead to undefined behavior.
The simple solution? Use strdup to duplicate the string:
table->word = strdup(str);
The strdup function is not actually in standard C, but just about all platforms have it.

Answer 2

In your insert function, you do not allocate/malloc() space for the word pointer you are trying to strcpy() to:
if(table == NULL){
table = (struct entry *)malloc(sizeof(struct entry));
strcpy(table->word,str);
table -> n = 1;
table -> left = NULL;
table -> right = NULL;
}
Usually this code would exit with a segmentation fault, because you are copying data to memory you don't own, but this is easy to fix:
table->word = malloc(strlen(str) + 1);
strcpy(table->word, str);
You'll want to allocate one extra byte above the string length, to allow for the null terminator.
You do not need or want to cast the result of malloc(). In other words, this is fine:
table = malloc(sizeof(struct entry));
Get into the habit of using free() on any pointers you have malloc()-ed, when you are done with them. Otherwise, you end up with a memory leak.
Also, compile with -Wall -Weverything flags to enable all warnings.
Note: If one absolutely must use strdup(), it is easy to write a custom function to do so:
char* my_very_own_strdup(const char* src)
{
char* dest = NULL;
if (!src)
return dest;
size_t src_len = strlen(src) + 1;
dest = malloc(src_len);
if (!dest) {
perror("Error: Could not allocate space for string copy\n");
exit(EXIT_FAILURE);
}
memcpy(dest, src, src_len);
return dest;
}

Answer 3

On the line strcpy(table->word,str); where is table->word allocated?
So It copies only 4 bytes to table->word because pointer size is 4-bytes in your machine. So Be careful, you must allocate table->word there,
I would use this one instead of that table->word = strdup(str);