cc [ flag... ] file... -lnsl [ library...]
#include <netdir.h>

int netdir_getbyname(struct netconfig *config,
 struct nd_hostserv *service, struct nd_addrlist **addrs);

int netdir_getbyaddr(struct netconfig *config,
 struct nd_hostservlist **service, struct netbuf *netaddr);

void netdir_free(void *ptr, int struct_type);

int netdir_options(struct netconfig *config, int option, int fildes,
 char *pointer_to_args);

char *taddr2uaddr(struct netconfig *config, struct netbuf *addr);

struct netbuf *uaddr2taddr(struct netconfig *config, char *uaddr);

void netdir_perror(char *s);

char *netdir_sperror(void);

The netdir functions provide a generic interface for name-to-address mapping that will work with all transport protocols. This interface provides a generic way for programs to convert transport specific addresses into common structures and back again. The netconfig structure, described on the netconfig(4) manual page, identifies the transport.

The netdir_getbyname() function maps the machine name and service name in the nd_hostserv structure to a collection of addresses of the type understood by the transport identified in the netconfig structure. This function returns all addresses that are valid for that transport in the nd_addrlist structure. The nd_hostserv structure contains the following members:

char *h_host; /* host name */
char *h_serv; /* service name */

The nd_addrlist structure contains the following members:

int n_cnt; /* number of addresses */
struct netbuf *n_addrs;

The netdir_getbyname() function accepts some special-case host names. The host names are defined in <netdir.h>. The currently defined host names are: HOST_SELF

All fields of the nd_hostserv structure must be initialized.

To find the address of a given host and service on all available transports, call the netdir_getbyname() function with each struct netconfig structure returned by getnetconfig(3NSL).

The netdir_getbyaddr() function maps addresses to service names. The function returns service, a list of host and service pairs that yield these addresses. If more than one tuple of host and service name is returned, the first tuple contains the preferred host and service names:

struct nd_hostservlist {
 int *h_cnt; /* number of hostservs found */
 struct hostserv *h_hostservs;
}

The netdir_free() structure is used to free the structures allocated by the name to address translation functions. The ptr parameter points to the structure that has to be freed. The parameter struct_type identifies the structure:

struct netbuf ND_ADDR
struct nd_addrlist ND_ADDRLIST
struct hostserv ND_HOSTSERV
struct nd_hostservlist ND_HOSTSERVLIST

The netdir_options() function is used to do all transport-specific setups and option management. fildes is the associated file descriptor. option, fildes, and pointer_to_args are passed to the netdir_options() function for the transport specified in config. Currently four values are defined for option:

ND_SET_BROADCAST

ND_SET_RESERVEDPORT

ND_CHECK_RESERVEDPORT

ND_MERGEADDR

The specific actions of each option follow. ND_SET_BROADCAST

char s_uaddr; /* server's universal address */
char c_uaddr; /* client's universal address */
char m_uaddr; /* the result */

If a transport provider does not support an option, netdir_options returns -1 and the error message can be printed through netdir_perror() or netdir_sperror().

The taddr2uaddr() and uaddr2taddr() functions support translation between universal addresses and TLI type netbufs. The taddr2uaddr() function takes a struct netbuf data structure and returns a pointer to a string that contains the universal address. It returns NULL if the conversion is not possible. This is not a fatal condition as some transports do not support a universal address form.

The free() function is used to free the universal address returned by the taddr2uaddr() function.

The uaddr2taddr() function is the reverse of the taddr2uaddr() function. It returns the struct netbuf data structure for the given universal address and NULL on failure.

The netdir_perror() function prints an error message in standard output that states the cause of a name-to-address mapping failure. The error message is preceded by the string given as an argument.

The netdir_sperror() function returns a string with an error message that states the cause of a name-to-address mapping failure.

The netdir_sperror() function returns a pointer to a buffer which contains the error message string. The buffer is overwritten on each call. In multithreaded applications, this buffer is implemented as thread-specific data.

The netdir_getbyname(), netdir_getbyaddr(), and netdir_options() functions return 0 on success and a non-zero value on failure.

See attributes(5) for descriptions of the following attributes:

gethostname(3C), getnetconfig(3NSL), getnetpath(3NSL), netconfig(4), attributes(5)