ADDR2ASCII(3) Library Functions Manual ADDR2ASCII(3)

addr2ascii, ascii2addr — Generic address formatting routines

Standard C Library (libc, -lc)

#include <arpa/inet.h>

char *
addr2ascii(int af, const void *addrp, int len, char *buf);

int
ascii2addr(int af, const char *ascii, void *result);

The routines () and ascii2addr() are used to convert network addresses between binary form and a printable form appropriate to the address family. Both functions take an af argument, specifying the address family to be used in the conversion process. (Currently, only the AF_INET and AF_LINK address families are supported.)

The () function is used to convert binary, network-format addresses into printable form. In addition to af, there are three other arguments. The addrp argument is a pointer to the network address to be converted. The len argument is the length of the address. The buf argument is an optional pointer to a caller-allocated buffer to hold the result; if a null pointer is passed, addr2ascii() uses a statically-allocated buffer.

The () function performs the inverse operation to addr2ascii(). In addition to af, it takes two arguments, ascii and result. The ascii argument is a pointer to the string which is to be converted into binary. The result argument is a pointer to an appropriate network address structure for the specified family.

The following gives the appropriate structure to use for binary addresses in the specified family:

(in <arpa/inet.h>)
(in <net/if_dl.h>)

AF_INET and AF_LINK constants are defined in <sys/socket.h>

The addr2ascii() function returns the address of the buffer it was passed, or a static buffer if the a null pointer was passed; on failure, it returns a null pointer. The ascii2addr() function returns the length of the binary address in bytes, or -1 on failure.

The inet(3) functions inet_ntoa() and inet_aton() could be implemented thusly:

#include <sys/socket.h>
#include <arpa/inet.h>

char *
inet_ntoa(struct in_addr addr)
{
	return addr2ascii(AF_INET, &addr, sizeof addr, 0);
}

int
inet_aton(const char *ascii, struct in_addr *addr)
{
	return (ascii2addr(AF_INET, ascii, addr)
	    == sizeof(*addr));
}

In actuality, this cannot be done because addr2ascii() and ascii2addr() are implemented in terms of the inet(3) functions, rather than the other way around.

When a failure is returned, errno is set to one of the following values:

[]
The addr2ascii() routine was passed a len argument which was inappropriate for the address family given by af.
[]
Either routine was passed an af argument other than AF_INET or AF_LINK.
[]
The string passed to ascii2addr() was improperly formatted for address family af.

inet(3), linkaddr(3), inet(4)

An interface close to this one was originally suggested by Craig Partridge. This particular interface originally appeared in the INRIA IPv6 implementation.

Code and documentation by Garrett A. Wollman, MIT Laboratory for Computer Science.

The original implementations supported IPv6. This support should eventually be resurrected. The NRL implementation also included support for the AF_ISO and AF_NS address families.

The genericity of this interface is somewhat questionable. A truly generic interface would provide a means for determining the length of the buffer to be used so that it could be dynamically allocated, and would always require a “struct sockaddr” to hold the binary address. Unfortunately, this is incompatible with existing practice. This limitation means that a routine for printing network addresses from arbitrary address families must still have internal knowledge of the maximum buffer length needed and the appropriate part of the address to use as the binary address.

June 13, 1996 Mac OS X 12