%/* % * Sun RPC is a product of Sun Microsystems, Inc. and is provided for % * unrestricted use provided that this legend is included on all tape % * media and as a part of the software program in whole or part. Users % * may copy or modify Sun RPC without charge, but are not authorized % * to license or distribute it to anyone else except as part of a product or % * program developed by the user or with the express written consent of % * Sun Microsystems, Inc. % * % * SUN RPC IS PROVIDED AS IS WITH NO WARRANTIES OF ANY KIND INCLUDING THE % * WARRANTIES OF DESIGN, MERCHANTIBILITY AND FITNESS FOR A PARTICULAR % * PURPOSE, OR ARISING FROM A COURSE OF DEALING, USAGE OR TRADE PRACTICE. % * % * Sun RPC is provided with no support and without any obligation on the % * part of Sun Microsystems, Inc. to assist in its use, correction, % * modification or enhancement. % * % * SUN MICROSYSTEMS, INC. SHALL HAVE NO LIABILITY WITH RESPECT TO THE % * INFRINGEMENT OF COPYRIGHTS, TRADE SECRETS OR ANY PATENTS BY SUN RPC % * OR ANY PART THEREOF. % * % * In no event will Sun Microsystems, Inc. be liable for any lost revenue % * or profits or other special, indirect and consequential damages, even if % * Sun has been advised of the possibility of such damages. % * % * Sun Microsystems, Inc. % * 2550 Garcia Avenue % * Mountain View, California 94043 % */ /* * nis_object.x * * Copyright (c) 1988-1992 Sun Microsystems Inc * All Rights Reserved. */ /* From: %#pragma ident "@(#)nis_object.x 1.10 94/05/03 SMI" */ #if RPC_HDR % %#ifndef __nis_object_h %#define __nis_object_h % #endif /* * This file defines the format for a NIS object in RPC language. * It is included by the main .x file and the database access protocol * file. It is common because both of them need to deal with the same * type of object. Generating the actual code though is a bit messy because * the nis.x file and the nis_dba.x file will generate xdr routines to * encode/decode objects when only one set is needed. Such is life when * one is using rpcgen. * * Note, the protocol doesn't specify any limits on such things as * maximum name length, number of attributes, etc. These are enforced * by the database backend. When you hit them you will no. Also see * the db_getlimits() function for fetching the limit values. * */ /* Some manifest constants, chosen to maximize flexibility without * plugging the wire full of data. */ const NIS_MAXSTRINGLEN = 255; const NIS_MAXNAMELEN = 1024; const NIS_MAXATTRNAME = 32; const NIS_MAXATTRVAL = 2048; const NIS_MAXCOLUMNS = 64; const NIS_MAXATTR = 16; const NIS_MAXPATH = 1024; const NIS_MAXREPLICAS = 128; const NIS_MAXLINKS = 16; const NIS_PK_NONE = 0; /* no public key (unix/sys auth) */ const NIS_PK_DH = 1; /* Public key is Diffie-Hellman type */ const NIS_PK_RSA = 2; /* Public key if RSA type */ const NIS_PK_KERB = 3; /* Use kerberos style authentication */ /* * The fundamental name type of NIS. The name may consist of two parts, * the first being the fully qualified name, and the second being an * optional set of attribute/value pairs. */ struct nis_attr { string zattr_ndx<>; /* name of the index */ opaque zattr_val<>; /* Value for the attribute. */ }; typedef string nis_name<>; /* The NIS name itself. */ /* NIS object types are defined by the following enumeration. The numbers * they use are based on the following scheme : * 0 - 1023 are reserved for Sun, * 1024 - 2047 are defined to be private to a particular tree. * 2048 - 4095 are defined to be user defined. * 4096 - ... are reserved for future use. */ enum zotypes { BOGUS_OBJ = 0, /* Uninitialized object structure */ NO_OBJ = 1, /* NULL object (no data) */ DIRECTORY_OBJ = 2, /* Directory object describing domain */ GROUP_OBJ = 3, /* Group object (a list of names) */ TABLE_OBJ = 4, /* Table object (a database schema) */ ENTRY_OBJ = 5, /* Entry object (a database record) */ LINK_OBJ = 6, /* A name link. */ PRIVATE_OBJ = 7 /* Private object (all opaque data) */ }; /* * The types of Name services NIS knows about. They are enumerated * here. The Binder code will use this type to determine if it has * a set of library routines that will access the indicated name service. */ enum nstype { UNKNOWN = 0, NIS = 1, /* Nis Plus Service */ SUNYP = 2, /* Old NIS Service */ IVY = 3, /* Nis Plus Plus Service */ DNS = 4, /* Domain Name Service */ X500 = 5, /* ISO/CCCIT X.500 Service */ DNANS = 6, /* Digital DECNet Name Service */ XCHS = 7, /* Xerox ClearingHouse Service */ CDS= 8 }; /* * DIRECTORY - The name service object. These objects identify other name * servers that are serving some portion of the name space. Each has a * type associated with it. The resolver library will note whether or not * is has the needed routines to access that type of service. * The oarmask structure defines an access rights mask on a per object * type basis for the name spaces. The only bits currently used are * create and destroy. By enabling or disabling these access rights for * a specific object type for a one of the accessor entities (owner, * group, world) the administrator can control what types of objects * may be freely added to the name space and which require the * administrator's approval. */ struct oar_mask { u_long oa_rights; /* Access rights mask */ zotypes oa_otype; /* Object type */ }; struct endpoint { string uaddr<>; string family<>; /* Transport family (INET, OSI, etc) */ string proto<>; /* Protocol (TCP, UDP, CLNP, etc) */ }; /* * Note: pkey is a netobj which is limited to 1024 bytes which limits the * keysize to 8192 bits. This is consider to be a reasonable limit for * the expected lifetime of this service. */ struct nis_server { nis_name name; /* Principal name of the server */ endpoint ep<>; /* Universal addr(s) for server */ u_long key_type; /* Public key type */ netobj pkey; /* server's public key */ }; struct directory_obj { nis_name do_name; /* Name of the directory being served */ nstype do_type; /* one of NIS, DNS, IVY, YP, or X.500 */ nis_server do_servers<>; /* <0> == Primary name server */ u_long do_ttl; /* Time To Live (for caches) */ oar_mask do_armask<>; /* Create/Destroy rights by object type */ }; /* * ENTRY - This is one row of data from an information base. * The type value is used by the client library to convert the entry to * it's internal structure representation. The Table name is a back pointer * to the table where the entry is stored. This allows the client library * to determine where to send a request if the client wishes to change this * entry but got to it through a LINK rather than directly. * If the entry is a "standalone" entry then this field is void. */ const EN_BINARY = 1; /* Indicates value is binary data */ const EN_CRYPT = 2; /* Indicates the value is encrypted */ const EN_XDR = 4; /* Indicates the value is XDR encoded */ const EN_MODIFIED = 8; /* Indicates entry is modified. */ const EN_ASN1 = 64; /* Means contents use ASN.1 encoding */ struct entry_col { u_long ec_flags; /* Flags for this value */ opaque ec_value<>; /* It's textual value */ }; struct entry_obj { string en_type<>; /* Type of entry such as "passwd" */ entry_col en_cols<>; /* Value for the entry */ }; /* * GROUP - The group object contains a list of NIS principal names. Groups * are used to authorize principals. Each object has a set of access rights * for members of its group. Principal names in groups are in the form * name.directory and recursive groups are expressed as @groupname.directory */ struct group_obj { u_long gr_flags; /* Flags controlling group */ nis_name gr_members<>; /* List of names in group */ }; /* * LINK - This is the LINK object. It is quite similar to a symbolic link * in the UNIX filesystem. The attributes in the main object structure are * relative to the LINK data and not what it points to (like the file system) * "modify" privleges here indicate the right to modify what the link points * at and not to modify that actual object pointed to by the link. */ struct link_obj { zotypes li_rtype; /* Real type of the object */ nis_attr li_attrs<>; /* Attribute/Values for tables */ nis_name li_name; /* The object's real NIS name */ }; /* * TABLE - This is the table object. It implements a simple * data base that applications and use for configuration or * administration purposes. The role of the table is to group together * a set of related entries. Tables are the simple database component * of NIS. Like many databases, tables are logically divided into columns * and rows. The columns are labeled with indexes and each ENTRY makes * up a row. Rows may be addressed within the table by selecting one * or more indexes, and values for those indexes. Each row which has * a value for the given index that matches the desired value is returned. * Within the definition of each column there is a flags variable, this * variable contains flags which determine whether or not the column is * searchable, contains binary data, and access rights for the entry objects * column value. */ const TA_BINARY = 1; /* Means table data is binary */ const TA_CRYPT = 2; /* Means value should be encrypted */ const TA_XDR = 4; /* Means value is XDR encoded */ const TA_SEARCHABLE = 8; /* Means this column is searchable */ const TA_CASE = 16; /* Means this column is Case Sensitive */ const TA_MODIFIED = 32; /* Means this columns attrs are modified*/ const TA_ASN1 = 64; /* Means contents use ASN.1 encoding */ struct table_col { string tc_name<64>; /* Column Name */ u_long tc_flags; /* control flags */ u_long tc_rights; /* Access rights mask */ }; struct table_obj { string ta_type<64>; /* Table type such as "passwd" */ int ta_maxcol; /* Total number of columns */ u_char ta_sep; /* Separator character */ table_col ta_cols<>; /* The number of table indexes */ string ta_path<>; /* A search path for this table */ }; /* * This union joins together all of the currently known objects. */ union objdata switch (zotypes zo_type) { case DIRECTORY_OBJ : struct directory_obj di_data; case GROUP_OBJ : struct group_obj gr_data; case TABLE_OBJ : struct table_obj ta_data; case ENTRY_OBJ: struct entry_obj en_data; case LINK_OBJ : struct link_obj li_data; case PRIVATE_OBJ : opaque po_data<>; case NO_OBJ : void; case BOGUS_OBJ : void; default : void; }; /* * This is the basic NIS object data type. It consists of a generic part * which all objects contain, and a specialized part which varies depending * on the type of the object. All of the specialized sections have been * described above. You might have wondered why they all start with an * integer size, followed by the useful data. The answer is, when the * server doesn't recognize the type returned it treats it as opaque data. * And the definition for opaque data is {int size; char *data;}. In this * way, servers and utility routines that do not understand a given type * may still pass it around. One has to be careful in setting * this variable accurately, it must take into account such things as * XDR padding of structures etc. The best way to set it is to note one's * position in the XDR encoding stream, encode the structure, look at the * new position and calculate the size. */ struct nis_oid { u_long ctime; /* Time of objects creation */ u_long mtime; /* Time of objects modification */ }; struct nis_object { nis_oid zo_oid; /* object identity verifier. */ nis_name zo_name; /* The NIS name for this object */ nis_name zo_owner; /* NIS name of object owner. */ nis_name zo_group; /* NIS name of access group. */ nis_name zo_domain; /* The administrator for the object */ u_long zo_access; /* Access rights (owner, group, world) */ u_long zo_ttl; /* Object's time to live in seconds. */ objdata zo_data; /* Data structure for this type */ }; #if RPC_HDR % %#endif /* if __nis_object_h */ % #endif