delete[] causing heap corruption - c++

I'm well aware that there are countless problems like this, but I searched for hours and couldn't understand what I did wrong so I would really appreciate your help. (I'm new to programming)
I need to create a dictionary manager of sorts as part of my homework but I seem to have a problem with deleting words.
I get an error message "...triggered a breakpoint".
The usual answer people get to this problem is that this is heap corruption caused by going out of bounds but I can't see if and how I caused this.
I already made something similar with bus info management and it worked perfectly so that makes me even more confused... (Obviously, I did not make the mechanism exactly the same, but even after looking at my previous code I couldn't isolate the problem)
I added the functions I believe are of concern,
The adding function:
void Add_Word(char**& dictionary, int& dictionary_size, char word[])
{
char** temp = new char*[dictionary_size + 1]; // Create a new array of appropriate size.
int i;
for (i = 0; i < dictionary_size; i++)
{
temp[i] = dictionary[i]; // Copy head pointers addresses for all existing items.
}
temp[i] = new char[strlen(word)]; // Add the space for the new word,
temp[i][strlen(word)] = '\0'; // mark its end
strcpy_s(temp[i], strlen(word) + 1, word); // then copy it.
// I'm really not so sure about what I should put in the buffer length but
// strlen(word) + 1 seemed to work... I know... not good, but strlen(word) alone caused a problem.
if (dictionary_size > 0)
delete []dictionary; // Delete previous head pointers array if there are any and
dictionary = temp; // reset the main pointer to the address of the new one.
dictionary_size++; // Finally, increase dictionary_size.
}
The deleting function:
void Delete_Word(char**& dictionary, int& dictionary_size, char* word)
{
// !!! This is where the crash thingy happens.
delete[] Search_For_Word(dictionary, dictionary_size, word); // Delete the word from the dictionary.
// Search_For_Word returns a pointer to the word it receives, from the dictionary.
char** temp = new char*[dictionary_size - 1]; // Create a new array of appropriate size.
int i;
for (i = 0; i < dictionary_size; i++)
{
if (dictionary[i][0])
temp[i] = dictionary[i]; // Copy the head pointers of the existing
// items to the new array except for the deleted word.
}
delete[] dictionary; // Delete previous head pointers array and
dictionary = temp; // reset the main pointer to the address of the new one.
dictionary_size--; // Finally, decrease dictionary_size.
}
EDIT: Any parts that are excessively inefficient or obviously broken are likely a result of me messing with my code trying to figure this out on my own (such as the calling 3 times to strlen mentioned (thanks again for that, kfsone...), or forgetting to +1 it for the '\0' to mark the end of a string
--actually, no, if we go by obvious you won't tell me my mistakes #.#).
As for the reason I'm dealing with char instead of strings and vectors please allow me to quote myself: "...as part of my homework". I just barely started programming. That, and I want to grasp the basics before moving on to using the more comfortable higher-up tools.

Change:
temp[i] = new char[strlen(word)]
To:
temp[i] = new char[strlen(word)+1]

Your code has several problems.
First, if you want to allocate a C-style string on the heap using new[], then you must pay attention to the terminating NUL character.
So, if you want to do a deep copy from a string word, then you must calculate enough room, considering strlen(word) + 1: the +1 is for the terminating NUL character.
e.g.:
// Original code (wrong):
//
// temp[i] = new char[strlen(word)];
//
// New code:
temp[i] = new char[strlen(word) + 1]; // consider terminating NUL (+1)
Moreover, following your code with explicit new[]s and delete[]s is not easy.
In modern C++, you may want to use convenient robust container classes like std::vector and string classes like std::string, instead of raw C-style pointers and strings.
You can simply store a list of strings using a std::vector<std::string>, and vector::push_back() method to add new strings to the vector. No need to complicate code with new[], delete[], strcpy_s(), etc.
And if you want to deep-copy strings, you can just use the simple natural overload of operator= for std::string, and copy constructors; e.g. std::string temp = word; will work just fine.

This is C++, why are you not using std::string instead of char buffers?
If you must use char buffer strings and the secure forms of strcpy_s know that the buffer length must always be the size of the destination buffer, never a strlen function. In your case it is a bit understandable since you created the buffer with the strlen function. But what you should do is set the value into a variable and then use that any time you need the buffer size.
Also, and where I think your bug is, you are writing temp[i][strlen(word)] = '\0'; But the actual indexes of the buffer go from 0 to strlen(word)-1 so you're writing outside the allocated memory.

The code is now working.
It was wrong all over.
I messed up pretty much any part that I could regarding the dynamic memory while trying to fix it before.
I initially didn't care about calling 3 times to strlen becuase it's just homework and a very small program but I guess it's better to get used to do things the right way...
I also dropped the copy which I evidently don't understand very well in favour of a simple for loop.
// Add function. The rest is cut.
int word_length = strlen(word);
temp[i] = new char[word_length + 1]; // Added +1 here.
temp[i][word_length] = '\0'; /* This was correct after all.
the word_length index is the correct ending.*/
for (int j = 0; j < word_length; j++) // copy replaced by for loop.
temp[i][j] = word[j];
// cut
}
void Delete_Word(char**& dictionary, int& dictionary_size, char* word)
{
delete[] Search_For_Word(dictionary, dictionary_size, word);
// There was a -1 mistake here I made in order to try and fix the thing earlier.
// No need for more, it works perfectly now.

Related

ostringstream gets wrong int value after incrementation [c++]

This is part of my code:
for(int i=0; i<10; i++)
{
tab[i]=new char[80];
oss.str("");
oss<<"line number: <<i;
temp=oss.str();
tab[i]=(char*)temp.c_str();
}
and when I print the tab, the result is in turns 8 and 9. Am I doing something wrong?
Taking a bunch of comments and consolidating (Why are you commentators not answering instead?):
What you are actually doing there is changing the pointer value of tab[i] from your allocated memory to the internal string of temp. Which is a bad idea because temp will free that memory as soon as it is destructed.
You don't need temp at all, as far as I can see.
The C way to fix this is to use strcpy(tab[i], oss.str().c_str()) which will copy the characters one by one. Note that your current code does a new char[80] so if your string is longer than 79 characters it will overflow. If you must use a new'd buffer, do it with something like new char[oss.str().size() + 1]
The C++ way to fix this is to use an array of std::string instead of an array of char*. Then you could just assign, like tab[i] = oss.str() and it would copy it properly. It would also clean up the memory used when the array goes out of scope, as your code right now doesn't do.

100% of array correct in function, 75% of array correct in CALLING function - C

Note: i'm using the c++ compiler, hence why I can use pass by reference
i have a strange problem, and I don't really know what's going on.
Basically, I have a text file: http://pastebin.com/mCp6K3HB
and I'm reading the contents of the text file in to an array of atoms:
typedef struct{
char * name;
char * symbol;
int atomic_number;
double atomic_weight;
int electrons;
int neutrons;
int protons;
} atom;
this is my type definition of atom.
void set_up_temp(atom (&element_record)[DIM1])
{
char temp_array[826][20];
char temp2[128][20];
int i=0;
int j=0;
int ctr=0;
FILE *f=fopen("atoms.txt","r");
for (i = 0; f && !feof(f) && i < 827; i++ )
{
fgets(temp_array[i],sizeof(temp_array[0]),f);
}
for (j = 0; j < 128; j++)
{
element_record[j].name = temp_array[ctr];
element_record[j].symbol = temp_array[ctr+1];
element_record[j].atomic_number = atol(temp_array[ctr+2]);
element_record[j].atomic_weight = atol(temp_array[ctr+3]);
element_record[j].electrons = atol(temp_array[ctr+4]);
element_record[j].neutrons = atol(temp_array[ctr+5]);
element_record[j].protons = atol(temp_array[ctr+6]);
ctr = ctr + 7;
}
//Close the file to free up memory and prevent leaks
fclose(f);
} //AT THIS POINT THE DATA IS FINE
Here is the function I'm using to read the data. When i debug this function, and let it run right up to the end, I use the debugger to check it's contents, and the array has 100% correct data, that is, all elements are what they should be relative to the text file.
http://i.imgur.com/SEq9w7Q.png This image shows what I'm talking about. On the left, all the elements, 0, up to 127, are perfect.
Then, I go down to the function I'm calling it from.
atom myAtoms[118];
set_up_temp(myAtoms); //AT THIS POINT DATA IS FINE
region current_button_pressed; // NOW IT'S BROKEN
load_font_named("arial", "cour.ttf", 20);
panel p1 = load_panel("atomicpanel.txt");
panel p2 = load_panel("NumberPanel.txt");
As soon as ANYTHING is called, after i call set_up_temp, the elements 103 to 127 of my array turn in to jibberish. As more things get called, EVEN MORE of the array turns to jibberish. This is weird, I don't know what's happening... Does anyone have any idea? Thanks.
for (j = 0; j < 128; j++)
{
element_record[j].name = temp_array[ctr];
You are storing, and then returning, pointers into temp_array, which is on the stack. The moment you return from the function, all of temp_array becomes invalid -- it's undefined behavior to dereference any of those pointers after that point. "Undefined behavior" includes the possibility that you can still read elements 0 through 102 with no trouble, but 103 through 127 turn to gibberish, as you say. You need to allocate space for these strings that will live as long as the atom object. Since as you say you are using C++, the easiest fix is to change both char * members to std::string. (If you don't want to use std::string, the second easiest fix is to use strdup, but then you have to free that memory explicitly.)
This may not be the only bug in this code, but it's probably the one causing your immediate problem.
In case you're curious, the reason the high end of the data is getting corrupted is that on most (but not all) computers, including the one you're using, the stack grows downward, i.e. from high addresses to low. Arrays, however, always index from low addresses to high. So the high end of the memory area that used to be temp_array is the part that's closest to the stack pointer in the caller, and thus most likely to be overwritten by subsequent function calls.
Casual inspection yields this:
char temp_array[826][20];
...
for (i = 0; f && !feof(f) && i < 827; i++ )
Your code potentially allows i to become 826. Which means you're accessing the 827th element of temp_array. Which is one past the end. Oops.
Additionally, you are allocating an array of 118 atoms (atom myAtoms[118];) but you are setting 128 of them inside of set_up_temp in the for (j = 0; j < 128; j++) loop.
The moral of this story: Mind your indices and since you use C++ leverage things like std::vector and std::string and avoid playing with arrays directly.
Update
As Zack pointed out, you're returning pointers to stack-allocated variables which will go away when the set_up_temp function returns. Additionally, the fgets you use doesn't do what you think it does and it's HORRIBLE code to begin with. Please read the documentation for fgets and ask yourself what your code does.
You are allocating an array with space for 118 elements but the code sets 128 of them, thus overwriting whatever happens to live right after the array.
Also as other noted you're storing in the array pointers to data that is temporary to the function (a no-no).
My suggestion is to start by reading a good book about C++ before programming because otherwise you're making your life harder for no reason. C++ is not a language in which you can hope to make serious progress by experimentation.

Resizing dynamic string array

I am trying to resize a dynamically allocated string array; here's the code!
void resize_array() {
size_t newSize = hash_array_length + 100;
string* newArr = new string[newSize];
fill_n(hash_array,newSize,"0"); //fills arrays with zeros
memcpy( newArr, hash_array, hash_array_length * sizeof(string) );
hash_array_length = newSize;
delete [] hash_array;
hash_array = newArr;
}
unfortunately it isn't working and gives a segmentation fault. any idea why? this is basically a linear probing hash table where the element gets inserted wherever there is a 0 hence I use fill_n to fill the newly created array with 0's. any help please?
memcpy( newArr, hash_array, hash_array_length * sizeof(string) );
This line is extremely dangerous, std::string is not a plain old data type,
you can't make sure that memcpy could initialize it correctly, it may cause
undefined behavior, one of the most nasty behavior of c++(or programming).
Besides, there are a better and safer(in most of the times) solution to create
a dynamic string array in c++, just use vector
//create a dynamic string array with newSize and initialize them with "0"
//in your case, I don't think you need to initialize it with "0"
std::vector<std::string> newArr(newSize, "0");
if the hash_array has the same type as newArr(std::vector)
The way of copy it is very easy.
c++98
std::copy(hash_array.begin(), hash_array.end(), newArr.begin());
c++11
std::copy(std::begin(hash_array), std::end(hash_array), std::begin(newArr));
Better treat c++ as a new language, it has too many things are different from c.
Besides, there are a lot of decent free IDE, like code::blocks and QtCreator
devc++ is a almost dead project.
If you are new to c++, c++ primer 5 is a good book to start.
If string is actually an std::string (and probably even if it isn't) then this will crash. You are creating a new array of strings, copying the old string classes over the top, and then freeing the old strings. But if the string class contains internal pointers to allocated memory this will result in a double free because all you are doing is copying the internal pointers - not making new memory allocations.
Think about it like this; imagine you had the following class:
class foo
{
char* bar;
foo() { bar = malloc(100); }
~foo() { free(bar);
};
foo* ptr1 = new foo;
foo* ptr2 = new foo;
memcpy(ptr2, ptr1, sizeof(foo*));
delete ptr1;
At this point, ptr2->bar points to the same memory that ptr1->bar did, but ptr1 and the memory it held has been freed
The best solution would be to use a std::vector because this handles the resizing automatically and you don't need to worry about copying arrays at all. But if you want to persist with your current approach, you need to change the memcpy call to the following:
for (int i = 0; i < hash_array_length; ++i)
{
newArr[i] = hash_array[i];
}
Rather than just copying the memory this will call the class's copy constructor and make a proper copy of its contents.
I suspect the culprit is memcpy call. string is complicated type which manages the char array by pointers (just as you are doing right now). Normally copying string is done using assignment operator, which for string also copies its own array. But memcpy simply copies byte-per-byte the pointer, and delete[] also deletes the array managed by string. Now the other string uses deleted string array, which is BAAAD.
You can use std::copy instead of memcpy, or even better yet, use std::vector, which is remedy to most of your dynamic memory handling problems ever.

C++ Seg Fault at end of function. Reference line = closing brace

In conclusion: Thanks so much everyone! All the responses posted below were correct. The initial error was me forgetting to leave room for the NULL terminator. Strcpy() is a dangerous function because when I used it, it didn't know when the end of the 'string' was. Therefore, strcpy() grabbed to much data and overwrote the return address.
EDIT: added more code from the program
SOLVED: Honestly, my initial implementation was crap. I don't even know why I wrote swap that way if I wanted to swap out elements of the array. (At the time, each element only had the a char array in it. So I was able to get away with the old implementation). I have re-written it to:
void swap(ArrayElement list[], int index1, int index2) {
ArrayElement temp;
temp = list[index1];
list[index1] = list[index2];
list[index2] = temp;
}
I'm having problems with a segmentation fault at the end of the following function.
struct ArrayElement {
char data[SIZE_OF_ELEMENT];
// Implemented this way so that I can expand to multiple values later on
}
//In main:
ArrayElement* list = new ArrayElement[NUM_OF_ELEMENTS];
void swap(ArrayElement list[], int index1, int index2) {
char temp[SIZE_OF_ELEMENT];
strcpy(temp, list[index2].data);
strcpy(list[index2].data, list[index1].data);
strcpy(list[index1].data, temp);
}
The error is a segmentation fault at line 45, which is the ending curly brace of the function. This was compiled using g++. I used gbd to try and debug it and everything works correctly until it hits the curly brace.
I can give more code from the program if it is needed. I don't want to post the entire thing because this is for a class.
My best guess is, the string at list[index2].data is larger than temp[] and by copying, you overwrote the stack and the return address.
Try inserting a test for the length:
#include <iostream>
...
int n = strlen(list[index2].data);
std::cerr << "len=" << n << ", SIZE_OF_ELEMENT=" << SIZE_OF_ELEMENT << std::endl;
and see, if n (list[index2].data) is larger than SIZE_OF_ELEMENT
strcpy is a hazardous function. If the length of the input string is SIZE_OF_ELEMENT or more, you will write past the end of your temp array. If you must use a fixed size array as the output array in strcpy, you should test that the strcpy will work before using the function.
Even better is to switch from using char arrays to std::string.
Is data defined like this char data[SOME_CONSTANT]? If so then are you sure that SIZE_OF_ELEMENT is large enough? You are remembering the NULL terminator too, right?
If in ArrayElement data is defined like this char *data; and is allocated with malloc at a later time then are you sure that index1 has a buffer large enough for the data in index2 and vice versa? Again, you are remembering the NULL terminator too, right?

Pass character array by value and return a new character array from the function?

I apologise if I'm completely misunderstanding C++ at the moment, so my question might be quite simple to solve. I'm trying to pass a character array into a function by value, create a new array of the same size and fill it with certain elements, and return that array from the function. This is the code I have so far:
char *breedMutation(char genome []){
size_t genes = sizeof(genome);
char * mutation = new char[genes];
for (size_t a = 0 ;a < genes; a++) {
mutation[a] = 'b';
}
return mutation;
}
The for loop is what updates the new array; right now, it's just dummy code, but hopefully the idea of the function is clear. When I call this function in main, however, I get an error of initializer fails to determine size of ‘mutation’. This is the code I have in main:
int main()
{
char target [] = "Das weisse leid"; //dummy message
char mutation [] = breedMutation(target);
return 0;
}
I need to learn more about pointers and character arrays, which I realise, but I'm trying to learn by example as well.
EDIT: This code, which I'm trying to modify for character arrays, is the basis for breedMutation.
int *f(size_t s){
int *ret=new int[s];
for (size_t a=0;a<s;a++)
ret[a]=a;
return ret;
}
Your error is because you can't declare mutation as a char[] and assign it the value of the char* being returned by breedMutation. If you want to do that, mutation should be declared as a char* and then deleted once you're done with it to avoid memory leaks in a real application.
Your breedMutation function, apart from dynamically allocating an array and returning it, is nothing like f. f simply creates an array of size s and fills each index in the array incrementally starting at 0. breedMutation would just fill the array with 'b' if you didn't have a logic error.
That error is that sizeof(genome); will return the size of a char*, which is generally 4 or 8 bytes on a common machine. You'll need to pass the size in as f does since arrays are demoted to pointers when passed to a function. However, with that snippet I don't see why you'd need to pass a char genome[] at all.
Also, in C++ you're better off using a container such as an std::vector or even std::array as opposed to dynamically allocated arrays (ones where you use new to create them) so that you don't have to worry about freeing them or keeping track of their size. In this case, std::string would be a good idea since it looks like you're trying to work with strings.
If you explain what exactly you're trying to do it might help us tell you how to go about your problem.
The line:
size_t genes = sizeof(genome);
will return the sizeof(char*) and not the number of elements in the genome array. You will need to pass the number of elements to the breedMutation() function:
breedMutation(target, strlen(target));
or find some other way of providing that information to the function.
Hope that helps.
EDIT: assuming it is the number of the elements in genome that you actually want.
Array are very limited.
Prefer to use std::vector (or std::string)
std::string breedMutation(std::string const& genome)
{
std::string mutation;
return mutation;
}
int main()
{
std::string target = "Das weisse leid"; //dummy message
std::string mutation = breedMutation(target);
}
Try replacing the second line of main() with:
char* mutation = breedMutation(target);
Also, don't forget to delete your mutation variable at the end.