freebsd-dev/lib/libc/stdlib/hcreate.3

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

326 lines
7.0 KiB
Groff
Raw Normal View History

.\"-
.\" Copyright (c) 1999 The NetBSD Foundation, Inc.
.\" All rights reserved.
.\"
.\" This code is derived from software contributed to The NetBSD Foundation
.\" by Klaus Klein.
.\"
.\" 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.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. 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 THE FOUNDATION 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.
.\"
.\" $FreeBSD$
.\"
.Dd February 6, 2017
.Dt HCREATE 3
.Os
.Sh NAME
.Nm hcreate ,
.Nm hcreate_r ,
.Nm hdestroy ,
.Nm hdestroy_r ,
.Nm hsearch ,
.Nm hsearch_r
.Nd manage hash search table
.Sh LIBRARY
.Lb libc
.Sh SYNOPSIS
.In search.h
.Ft int
.Fn hcreate "size_t nel"
.Ft int
.Fn hcreate_r "size_t nel" "struct hsearch_data *table"
.Ft void
.Fn hdestroy "void"
.Ft void
.Fn hdestroy_r "struct hsearch_data *table"
.Ft ENTRY *
.Fn hsearch "ENTRY item" "ACTION action"
.Ft int
.Fn hsearch_r "ENTRY item" "ACTION action" "ENTRY ** itemp" "struct hsearch_data *table"
.Sh DESCRIPTION
The
.Fn hcreate ,
.Fn hcreate_r ,
.Fn hdestroy ,
.Fn hdestroy_r
.Fn hsearch ,
and
.Fn hsearch_r
functions manage hash search tables.
.Pp
The
.Fn hcreate
function allocates sufficient space for the table, and the application should
ensure it is called before
.Fn hsearch
is used.
The
.Fa nel
argument is an estimate of the maximum
number of entries that the table should contain.
Replace implementation of hsearch() by one that scales. Traditionally the hcreate() function creates a hash table that uses chaining, using a fixed user-provided size. The problem with this approach is that this often either wastes memory (table too big) or yields bad performance (table too small). For applications it may not always be easy to estimate the right hash table size. A fixed number only increases performance compared to a linked list by a constant factor. This problem can be solved easily by dynamically resizing the hash table. If the size of the hash table is at least doubled, this has no negative on the running time complexity. If a dynamically sized hash table is used, we can also switch to using open addressing instead of chaining, which has the advantage of just using a single allocation for the entire table, instead of allocating many small objects. Finally, a problem with the existing implementation is that its deterministic algorithm for hashing makes it possible to come up with fixed patterns to trigger an excessive number of collisions. We can easily solve this by using FNV-1a as a hashing algorithm in combination with a randomly generated offset basis. Measurements have shown that this implementation is about 20-25% faster than the existing implementation (even if the existing implementation is given an excessive number of buckets). Though it allocates more memory through malloc() than the old implementation (between 4-8 pointers per used entry instead of 3), process memory use is similar to the old implementation as if the estimated size was underestimated by a factor 10. This is due to the fact that malloc() needs to perform less bookkeeping. Reviewed by: jilles, pfg Obtained from: https://github.com/NuxiNL/cloudlibc Differential Revision: https://reviews.freebsd.org/D4644
2015-12-27 07:50:11 +00:00
As this implementation resizes the hash table dynamically,
this argument is ignored.
.Pp
The
.Fn hdestroy
function disposes of the search table, and may be followed by another call to
.Fn hcreate .
After the call to
.Fn hdestroy ,
the data can no longer be considered accessible.
The
.Fn hdestroy
function calls
.Xr free 3
for each comparison key in the search table
but not the data item associated with the key.
.Pp
The
.Fn hsearch
function is a hash-table search routine.
It returns a pointer into a hash table
indicating the location at which an entry can be found.
The
.Fa item
argument is a structure of type
.Vt ENTRY
(defined in the
.In search.h
header) that contains two pointers:
.Fa item.key
points to the comparison key (a
.Vt "char *" ) ,
and
.Fa item.data
(a
.Vt "void *" )
points to any other data to be associated with
that key.
The comparison function used by
.Fn hsearch
is
.Xr strcmp 3 .
The
.Fa action
argument is a
member of an enumeration type
.Vt ACTION
indicating the disposition of the entry if it cannot be
found in the table.
.Dv ENTER
indicates that the
.Fa item
should be inserted in the table at an
appropriate point.
.Dv FIND
indicates that no entry should be made.
Unsuccessful resolution is
indicated by the return of a
.Dv NULL
pointer.
.Pp
The comparison key (passed to
.Fn hsearch
as
.Fa item.key )
must be allocated using
.Xr malloc 3
if
.Fa action
is
.Dv ENTER
and
.Fn hdestroy
is called.
.Pp
The
.Fn hcreate_r ,
.Fn hdestroy_r ,
and
.Fn hsearch_r
functions are re-entrant versions of the above functions that can
operate on a table supplied by the user.
The
.Fn hsearch_r
function returns
.Dv 0
if the action is
.Dv ENTER
and the element cannot be created,
.Dv 1
otherwise.
If the element exists or can be created, it will be placed in
.Fa itemp ,
otherwise
.Fa itemp
will be set to
.Dv NULL .
.Sh RETURN VALUES
The
.Fn hcreate
and
.Fn hcreate_r
functions return 0 if the table creation failed and the global variable
.Va errno
is set to indicate the error;
otherwise, a non-zero value is returned.
.Pp
The
.Fn hdestroy
and
.Fn hdestroy_r
functions return no value.
.Pp
The
.Fn hsearch
and
.Fn hsearch_r
functions return a
.Dv NULL
pointer if either the
.Fa action
is
.Dv FIND
and the
.Fa item
could not be found or the
.Fa action
is
.Dv ENTER
and the table is full.
.Sh EXAMPLES
The following example reads in strings followed by two numbers
and stores them in a hash table, discarding duplicates.
It then reads in strings and finds the matching entry in the hash
table and prints it out.
.Bd -literal
#include <stdio.h>
#include <search.h>
#include <string.h>
#include <stdlib.h>
struct info { /* This is the info stored in the table */
int age, room; /* other than the key. */
};
#define NUM_EMPL 5000 /* # of elements in search table. */
int
main(void)
{
char str[BUFSIZ]; /* Space to read string */
struct info info_space[NUM_EMPL]; /* Space to store employee info. */
struct info *info_ptr = info_space; /* Next space in info_space. */
ENTRY item;
ENTRY *found_item; /* Name to look for in table. */
char name_to_find[30];
int i = 0;
/* Create table; no error checking is performed. */
(void) hcreate(NUM_EMPL);
while (scanf("%s%d%d", str, &info_ptr->age,
&info_ptr->room) != EOF && i++ < NUM_EMPL) {
/* Put information in structure, and structure in item. */
item.key = strdup(str);
item.data = info_ptr;
info_ptr++;
/* Put item into table. */
(void) hsearch(item, ENTER);
}
/* Access table. */
item.key = name_to_find;
while (scanf("%s", item.key) != EOF) {
if ((found_item = hsearch(item, FIND)) != NULL) {
/* If item is in the table. */
(void)printf("found %s, age = %d, room = %d\en",
found_item->key,
((struct info *)found_item->data)->age,
((struct info *)found_item->data)->room);
} else
(void)printf("no such employee %s\en", name_to_find);
}
hdestroy();
return 0;
}
.Ed
2005-01-20 09:17:07 +00:00
.Sh ERRORS
The
.Fn hcreate ,
.Fn hcreate_r ,
.Fn hsearch ,
and
.Fn hsearch_r
functions will fail if:
2005-01-20 09:17:07 +00:00
.Bl -tag -width Er
.It Bq Er ENOMEM
Insufficient memory is available.
2005-01-20 09:17:07 +00:00
.El
.Pp
The
.Fn hsearch
and
.Fn hsearch_r
functions will also fail if the action is
.Dv FIND
and the element is not found:
.Bl -tag -width Er
.It Bq Er ESRCH
The
.Fa item
given is not found.
.El
.Sh SEE ALSO
.Xr bsearch 3 ,
.Xr lsearch 3 ,
.Xr malloc 3 ,
.Xr strcmp 3 ,
.Xr tsearch 3
.Sh STANDARDS
The
.Fn hcreate ,
.Fn hdestroy ,
and
.Fn hsearch
functions conform to
.St -xpg4.2 .
.Sh HISTORY
The
.Fn hcreate ,
.Fn hdestroy ,
and
.Fn hsearch
functions first appeared in
.At V .
The
.Fn hcreate_r ,
.Fn hdestroy_r
and
.Fn hsearch_r
functions are
.Tn GNU
extensions.
.Sh BUGS
The original,
.Pf non- Tn GNU
interface permits the use of only one hash table at a time.