How to find length of non null terminated sequence of character's length in C and C++?

Question

Hi,

How to find length of non null terminated sequence of character's length in C and C++?

For null terminated sequence of characters we can use strlen but for non null terminated sequence of characters how to find length?

In below code array b is not null terminated and if I use strlen over array b, it gives incorrect value. How to find length of array b.

char b[3]={'a', 'b', 'c'};
int n = strlen(b);
char b1[3]={'a', 'b'};
int n1= strlen(b1);
char* p="ab";
int n2 = strlen(p);

Answer 1

In these cases you can determine the size of the sequences at compile time -

    char x[3] = { 'a', 'b', 'c' };
    wchar_t y[] = { L'd', L'e', L'f', L'g' };

    auto lenX = sizeof x; // size in bytes
    auto lenX2 = _countof(x); // number of array elements
    auto lenX3 = ARRAYSIZE(x); // number of array elements

    auto lenY = sizeof y; // size in bytes
    auto lenY2 = _countof(y); // number of array elements
    auto lenY3 = ARRAYSIZE(y); // number of array elements

Answer 2

In C a string is defined as a nul-terminated sequence of
characters. In your example, neither b nor b1 constitute
a string.

Array b does not have room for a nul at the end of the
three characters.

Array b1 does have room for a nul at the end after the
two chars but as you have not specified a value for the
third element in the array it may not contain a nul.

If these lines of code are at global scope - outside of
main() or any other function - then the 3rd element in
b1 will be default-initialized to binary zero which is
the nul-terminator character.

If these lines of code are at local scope - inside of
main() or any other function - then the 3rd element in
b1 will be uninitialized.

Some compilers - such as VS/VC++ - will store a special
character in uninitialized variables during Debug builds.
In Release builds the contents will be random.

You cannot use any C functions which operate on strings
with ill-formed character arrays. Attempting to do so
is a programming error and will yield unpredictable
results.

The above comments also apply when working with a C style
string (a nul-terminated sequence of characters) in C++.

C++ also has its own string class std::string which is
the preferred object to use when working with strings
in C++.

Note also that in C++ you will get an error from this line:

char* p = "ab";

as C++ will not allow assigning a pointer to a constant
object to a pointer to a non-const object. The string
literal "ab" is a constant string which cannot be altered.

In C++ a cast would be required, either a C cast or a C++
cast, to override/discard the const-ness of the string:

char* p = (char*)"ab";

char* p = const_cast<char*>("ab");

However, doing this creates the possibility of errors
at run time so it is better to define the pointer as
being to a const string:

char const *p = "ab";

This will allow the compiler to catch attempts to alter
the constant string, rather than wait for an exception
at run time.

• Wayne