HardenedBSD/libexec/revnetgroup/hash.c

211 lines
5.8 KiB
C
Raw Normal View History

/*-
* SPDX-License-Identifier: BSD-4-Clause
*
* Copyright (c) 1995
* Bill Paul <wpaul@ctr.columbia.edu>. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by Bill Paul.
* 4. Neither the name of the author nor the names of any co-contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY Bill Paul AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL Bill Paul OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#ifndef lint
static const char rcsid[] =
1999-08-28 02:22:10 +02:00
"$FreeBSD$";
#endif /* not lint */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/types.h>
#include "hash.h"
/*
* This hash function is stolen directly from the
* Berkeley DB package. It already exists inside libc, but
* it's declared static which prevents us from calling it
* from here.
*/
/*
* OZ's original sdbm hash
*/
u_int32_t
hash(const void *keyarg, size_t len)
{
const u_char *key;
size_t loop;
u_int32_t h;
#define HASHC h = *key++ + 65599 * h
h = 0;
key = keyarg;
if (len > 0) {
loop = (len + 8 - 1) >> 3;
switch (len & (8 - 1)) {
case 0:
do {
HASHC;
/* FALLTHROUGH */
case 7:
HASHC;
/* FALLTHROUGH */
case 6:
HASHC;
/* FALLTHROUGH */
case 5:
HASHC;
/* FALLTHROUGH */
case 4:
HASHC;
/* FALLTHROUGH */
case 3:
HASHC;
/* FALLTHROUGH */
case 2:
HASHC;
/* FALLTHROUGH */
case 1:
HASHC;
} while (--loop);
}
}
return (h);
}
/*
* Generate a hash value for a given key (character string).
* We mask off all but the lower 8 bits since our table array
* can only hold 256 elements.
*/
u_int32_t
hashkey(char *key)
{
if (key == NULL)
return (-1);
return(hash((void *)key, strlen(key)) & HASH_MASK);
}
/* Find an entry in the hash table (may be hanging off a linked list). */
char *
lookup(struct group_entry *table[], char *key)
{
struct group_entry *cur;
cur = table[hashkey(key)];
while (cur) {
if (!strcmp(cur->key, key))
return(cur->data);
cur = cur->next;
}
return(NULL);
}
/*
* Store an entry in the main netgroup hash table. Here's how this
* works: the table can only be so big when we initialize it (TABLESIZE)
* but the number of netgroups in the /etc/netgroup file could easily be
* much larger than the table. Since our hash values are adjusted to
* never be greater than TABLESIZE too, this means it won't be long before
* we find ourselves with two keys that hash to the same value.
*
* One way to deal with this is to malloc(2) a second table and start
* doing indirection, but this is a pain in the butt and it's not worth
* going to all that trouble for a dinky little program like this. Instead,
* we turn each table entry into a linked list and simply link keys
* with the same hash value together at the same index location within
* the table.
*
* That's a lot of comment for such a small piece of code, isn't it.
*/
void
store(struct group_entry *table[], char *key, char *data)
{
struct group_entry *new;
u_int32_t i;
i = hashkey(key);
new = (struct group_entry *)malloc(sizeof(struct group_entry));
new->key = strdup(key);
new->data = strdup(data);
new->next = table[i];
table[i] = new;
return;
}
/*
* Store a group member entry and/or update its grouplist. This is
* a bit more complicated than the previous function since we have to
* maintain not only the hash table of group members, each group member
* structure also has a linked list of groups hung off it. If handed
* a member name that we haven't encountered before, we have to do
* two things: add that member to the table (possibly hanging them
* off the end of a linked list, as above), and add a group name to
* the member's grouplist list. If we're handed a name that already has
* an entry in the table, then we just have to do one thing, which is
* to update its grouplist.
*/
void
mstore(struct member_entry *table[], char *key, char *data, char *domain)
{
struct member_entry *cur, *new;
struct grouplist *tmp;
u_int32_t i;
i = hashkey(key);
cur = table[i];
tmp = (struct grouplist *)malloc(sizeof(struct grouplist));
tmp->groupname = strdup(data);
tmp->next = NULL;
/* Check if all we have to do is insert a new groupname. */
while (cur) {
if (!strcmp(cur->key, key)) {
tmp->next = cur->groups;
cur->groups = tmp;
return;
}
cur = cur->next;
}
/* Didn't find a match -- add the whole mess to the table. */
new = (struct member_entry *)malloc(sizeof(struct member_entry));
new->key = strdup(key);
new->domain = domain ? strdup(domain) : "*";
new->groups = tmp;
new->next = table[i];
table[i] = new;
return;
}