Non null-terminated strings

I'm used to strings being terminated with a NULL (or NUL - take your pick) character. I've been told that the <type> string defined in string.h is not NULL terminated. How is string length determined without a NULL character? Is a <type> int prefix attached to the string containing a value of the string length? Pascal uses that structure with a <type> char prefix (hence, Pascal strings can only be 255 character long).

Thank you.

Casey

However the library implementors want to do it. My understanding is that a length member variable is commonly used.

Originally Posted by Casey G

I'm used to strings being terminated with a NULL (or NUL - take your pick) character. I've been told that the <type> string defined in string.h is not NULL terminated. How is string length determined without a NULL character? Is a <type> int prefix attached to the string containing a value of the string length? Pascal uses that structure with a <type> char prefix (hence, Pascal strings can only be 255 character long).

Thank you.

Casey

Typically it's done with a struct...

Code:

typedef struct tString
  { char *string;
     int size;           // used for realloc()
     int length;        // MyStr.len
     int refcount; }  // used by garbage collector
   String;

As pointed out it is implementation specific but that's the general idea.

Thanks - I'm getting the general drift. I'm new to C++ (No! Really?) and when I was told strings were not null terminated, I was having trouble wrapping my head around the concept. I featured there must be some kind of typedef struct invloved where there is a variable defining the string length, a value for allocating memory for the string, and a pointer to the string. When I was told that a <type> string is not null terminated, I ask how the elements in an array of <type> string are defined (what seperates the array elements?). The person shook his head and said, "I dunno." The blind leading the blind. LMAO.

I'd like to see the _CDecl for <type> string, but I couldn't find it in string.h. Is it defined in a std???.h file?

Originally Posted by Casey G

Thanks - I'm getting the general drift. I'm new to C++ (No! Really?) and when I was told strings were not null terminated, I was having trouble wrapping my head around the concept. I featured there must be some kind of typedef struct invloved where there is a variable defining the string length, a value for allocating memory for the string, and a pointer to the string. When I was told that a <type> string is not null terminated, I ask how the elements in an array of <type> string are defined (what seperates the array elements?). The person shook his head and said, "I dunno." The blind leading the blind. LMAO.
I'd like to see the _CDecl for <type> string, but I couldn't find it in string.h. Is it defined in a std???.h file?

_cdecl is a calling convention... I'm guessing you mean the declaration...

It's likely in the string library and thanks to C++ code hiding, we probably have no access to it. Really... I'm not 100% sure that's how it works but that's how most compilers do it. It is, afterall, implementation defined, which means different libraries can do it in different ways so long as it's syntactically transparent to us as users. GCC might use a totally different approach than VC++ or MinGW... but it really doesn't matter, so long as it works.

Originally Posted by Casey G

I've been told that the <type> string defined in string.h is not NULL terminated.

Sorry, but what is a "<type> string"? Are you talking about std::string, or more generally about std::basic_string? Furthermore, when you say string.h, do you mean the <string.h> header inherited from C, or the <cstring> C++ version thereof, or the C++ specific <string> header?

In the case of std::basic_string, or more specifically std::string, the obvious way to implement is to keep track of size and capacity explicitly, rather than keep track of size through a null terminator. In fact, it is possible to embed null characters in a std::string.