Memo on strings in C and C++

0. char*

The 2 following statements are equivalent:

char s[50] = {'H', 'e', 'l', 'l', 'o', '\0'}; // 50 bytes reserved, but only the first 6 initialized char s[50] = "Hello"; // the same.

1. null terminated char*-strings

Most operations on char* strings assume that they are terminated by a character zero:

char s[20] = {'1', '2', '\0', '4', '5', '\0'}; // length is 2

strlen
strcpy
strncpy
sprintf
snprintf
strcat
strncat
strdup
etc.

To add the terminating zero, do this:

char myString[MAX_LENGTH+1]; myString[MAX_LENGTH] = 0; // or myString[MAX_LENGTH] = '\0';

Don't do that:

char myString[MAX_LENGTH+1]; myString[strlen(myString)] = 0; // wrong! because strlen needs the zero! strcat(myString, "0"); // wrong! because strcat needs the zero!

/!\ Using memcpy for a string is error prone. Don't do that:

char myString[100]; memcpy(myString, "12345", 5); // wrong! missing terminating zero!

2. char*-string operations

char s[100] = "123"; strcpy(s, "1234"); // 5 characters copied, including zero strncpy(s, "1234", 4); // wrong! 4 characters copied, missing zero strncpy(s, "1234", 5); // 5 characters copied, including zero strncpy(s, "1234", 10); // 10 characters copied, including 6 zeros strcat(s, "-4"); // { '1', '2', '3', '-', '4', '\0' } strncat(s, "-4", 1); // { '1', '2', '3', '-', '\0' } With strcat and strncat the resulting string is always null terminated. sprintf(s, "%02X", 255); // { 'F', 'F', '\0' } sprintf(s, "%02X", 1048575); // { 'F', 'F', 'F', 'F', 'F', '\0' } // (not limited to 2 characters!) snprintf(s, 4, "%02X", 1048575); // { 'F', 'F', 'F', '\0' } // 4 characters copied snprintf() writes at most N bytes (including the trailing null byte ('\0')).

3. sizeof

char name[100]; sizeof(name); // returns 100 char *city; city = (char*)malloc(100); sizeof(city); // returns the size of an integer. 4 on a 32-bit CPU.

4. Other string types

4.1 std::string (STL library)

In this paragraph, we use [ ] to represent the contents of string objects.

std::string s; s = "abc"; // [ 'a', 'b', 'c' ] size = 3 s += "-12"; // [ 'a', 'b', 'c', '-', '1', '2' ] size = 6 s += '\0'; // [ 'a', 'b', 'c', '-', '1', '2', '\0' ] size = 7 s.c_str(); // returns a null-terminated char*, suitable for use with strcpy, strlen, etc. s.size(); // returns the size of the string

4.1 std::ostringstream (STL library)

std::ostringstream oss; oss << hex << 1048575 << "-" << dec << 123; // contents: "fffff-123" std::string s = oss.str(); // converts to a std::string object oss.str(""); // contents: ""

4.3 CString (MFC library)

CString s; s.GetBuffer(0); // returns a null-terminated char*, suitable for use with strcpy, etc.