new char issues

Hi
I am trying to use strtok. The problem is that when I dynamically allocate an array of two characters as below, it would seem a bigger array is created.

Code:

---

input = new char[2]; /*get the length of the string*/

---

This is what I get from the debugger:
input = 0x00362aa8 "ÍÍýýýý««««««««îþ"

I tried with different sizes and it would seem there are always a certain number of I which correspond to the size with the extra stuff at the end.
The issue is that strtok gives the last token with these garbage values and I can't figure out what to do. Thanks
Amish

You have to set the null terminator for the debugger to show the string properly. In your code example, input can hold a string of length one because it needs the second spot for the null terminator. As always, the C++ string class is generally preferred for string processing.

Quote:

---

This is what I get from the debugger:
input = 0x00362aa8 "ÍÍýýýý««««««««îþ"

I tried with different sizes and it would seem there are always a certain number of I which correspond to the size with the extra stuff at the end.

---

This:

0x00362aa8

is the address in memory that is stored in input. This:

"ÍÍýýýý««««««««îþ"

is the value at that address in memory.

Quote:

---

The issue is that strtok gives the last token with these garbage values and I can't figure out what to do.

---

strtok() as well as the rest of the functions in <cstring> require null terminated char arrays--otherwise known as c-strings. All those funtions have loops like:

Code:

---

while(str[i] != '\0')
{
    //do something
    ++i;
}

---

Technically, it is the sequence of characters starting at that address until a null character is found (or until the debug window gets tired of showing weird characters and stops).

Since the string is uninitialized, unknown data is in that memory. In debug builds, often the same "unknown" data is used to aid in debugging, which is why you usually see the I.

"dynamically allocate"

I hate this phrase, sure it was dynamic back then, but now it just seems silly to call it dynamic.

I created the array with one more charater at the end and filled it with 0. Thanks. Incidentally I was wondering if there was any easy way to convert from const char * to char *. or from std::string to char *. Thanks,
Amish

Quote:

---

Originally Posted by loopshot

"dynamically allocate"

I hate this phrase, sure it was dynamic back then, but now it just seems silly to call it dynamic.

---

How so? And what else would you call it?

Quote:

---

Incidentally I was wondering if there was any easy way to convert from const char * to char *. or from std::string to char *

---

Code:

---

/*
const char* str = "hello";
char* ptr = const_cast<char*>(str);
ptr[0] = 'a';

No compiler error, however program will crash because
string literals are constants by definition, and you can't change them.*/

const char str[] = "hello"; 
char* ptr = const_cast<char*>(&str[0]);
ptr[0] = 'a';
cout<<ptr<<endl;
cout<<str<<endl<<endl;
//The string literal is copied char by char into str by operator=
//so str can be changed without violating any C++ rules, i.e. str isn't
//a pointer to a string literal; it's just an array with some chars in it

string myStr = "hello";
char* strPtr = const_cast<char*>(myStr.c_str());
strPtr[0] = 'a';
cout<<strPtr<<endl;
cout<<myStr<<endl;

---

You shouldn't need to be casting away const.

To convert from const char* or std::string to non-const char*, use strcpy to copy the string. If you need to modify the actual string, then you should find another method that doesn't involve converting to char*.