]>
Commit | Line | Data |
---|---|---|
5b2abdfb A |
1 | .\" Copyright (c) 1990, 1991, 1993 |
2 | .\" The Regents of the University of California. All rights reserved. | |
3 | .\" | |
4 | .\" This code is derived from software contributed to Berkeley by | |
5 | .\" the American National Standards Committee X3, on Information | |
6 | .\" Processing Systems. | |
7 | .\" | |
8 | .\" Redistribution and use in source and binary forms, with or without | |
9 | .\" modification, are permitted provided that the following conditions | |
10 | .\" are met: | |
11 | .\" 1. Redistributions of source code must retain the above copyright | |
12 | .\" notice, this list of conditions and the following disclaimer. | |
13 | .\" 2. Redistributions in binary form must reproduce the above copyright | |
14 | .\" notice, this list of conditions and the following disclaimer in the | |
15 | .\" documentation and/or other materials provided with the distribution. | |
5b2abdfb A |
16 | .\" 4. Neither the name of the University nor the names of its contributors |
17 | .\" may be used to endorse or promote products derived from this software | |
18 | .\" without specific prior written permission. | |
19 | .\" | |
20 | .\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND | |
21 | .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE | |
22 | .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE | |
23 | .\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE | |
24 | .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL | |
25 | .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS | |
26 | .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) | |
27 | .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT | |
28 | .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY | |
29 | .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF | |
30 | .\" SUCH DAMAGE. | |
31 | .\" | |
32 | .\" @(#)qsort.3 8.1 (Berkeley) 6/4/93 | |
1f2f436a | 33 | .\" $FreeBSD: src/lib/libc/stdlib/qsort.3,v 1.17 2007/01/09 00:28:10 imp Exp $ |
5b2abdfb | 34 | .\" |
3d9156a7 | 35 | .Dd September 30, 2003 |
5b2abdfb A |
36 | .Dt QSORT 3 |
37 | .Os | |
38 | .Sh NAME | |
ad3c9f2a A |
39 | .Nm heapsort , |
40 | #ifdef UNIFDEF_BLOCKS | |
41 | .Nm heapsort_b , | |
42 | #endif | |
43 | .Nm mergesort , | |
44 | #ifdef UNIFDEF_BLOCKS | |
45 | .Nm mergesort_b , | |
46 | #endif | |
47 | .Nm qsort , | |
48 | #ifdef UNIFDEF_BLOCKS | |
49 | .Nm qsort_b , | |
50 | #endif | |
51 | .Nm qsort_r | |
5b2abdfb | 52 | .Nd sort functions |
5b2abdfb A |
53 | .Sh SYNOPSIS |
54 | .In stdlib.h | |
ad3c9f2a A |
55 | .Ft int |
56 | .Fo heapsort | |
9385eb3d | 57 | .Fa "void *base" |
ad3c9f2a A |
58 | .Fa "size_t nel" |
59 | .Fa "size_t width" | |
9385eb3d A |
60 | .Fa "int \*[lp]*compar\*[rp]\*[lp]const void *, const void *\*[rp]" |
61 | .Fc | |
ad3c9f2a A |
62 | #ifdef UNIFDEF_BLOCKS |
63 | .Ft int | |
64 | .Fo heapsort_b | |
9385eb3d | 65 | .Fa "void *base" |
ad3c9f2a A |
66 | .Fa "size_t nel" |
67 | .Fa "size_t width" | |
68 | .Fa "int \*[lp]^compar\*[rp]\*[lp]const void *, const void *\*[rp]" | |
9385eb3d | 69 | .Fc |
ad3c9f2a | 70 | #endif |
5b2abdfb | 71 | .Ft int |
ad3c9f2a | 72 | .Fo mergesort |
9385eb3d | 73 | .Fa "void *base" |
ad3c9f2a A |
74 | .Fa "size_t nel" |
75 | .Fa "size_t width" | |
9385eb3d A |
76 | .Fa "int \*[lp]*compar\*[rp]\*[lp]const void *, const void *\*[rp]" |
77 | .Fc | |
ad3c9f2a | 78 | #ifdef UNIFDEF_BLOCKS |
5b2abdfb | 79 | .Ft int |
ad3c9f2a A |
80 | .Fo mergesort_b |
81 | .Fa "void *base" | |
82 | .Fa "size_t nel" | |
83 | .Fa "size_t width" | |
84 | .Fa "int \*[lp]^compar\*[rp]\*[lp]const void *, const void *\*[rp]" | |
85 | .Fc | |
86 | #endif | |
87 | .Ft void | |
88 | .Fo qsort | |
9385eb3d | 89 | .Fa "void *base" |
ad3c9f2a A |
90 | .Fa "size_t nel" |
91 | .Fa "size_t width" | |
9385eb3d A |
92 | .Fa "int \*[lp]*compar\*[rp]\*[lp]const void *, const void *\*[rp]" |
93 | .Fc | |
ad3c9f2a A |
94 | #ifdef UNIFDEF_BLOCKS |
95 | .Ft void | |
96 | .Fo qsort_b | |
97 | .Fa "void *base" | |
98 | .Fa "size_t nel" | |
99 | .Fa "size_t width" | |
100 | .Fa "int \*[lp]^compar\*[rp]\*[lp]const void *, const void *\*[rp]" | |
101 | .Fc | |
102 | #endif | |
103 | .Ft void | |
104 | .Fo qsort_r | |
105 | .Fa "void *base" | |
106 | .Fa "size_t nel" | |
107 | .Fa "size_t width" | |
108 | .Fa "void *thunk" | |
109 | .Fa "int \*[lp]*compar\*[rp]\*[lp]void *, const void *, const void *\*[rp]" | |
110 | .Fc | |
5b2abdfb A |
111 | .Sh DESCRIPTION |
112 | The | |
113 | .Fn qsort | |
114 | function is a modified partition-exchange sort, or quicksort. | |
115 | The | |
116 | .Fn heapsort | |
117 | function is a modified selection sort. | |
118 | The | |
119 | .Fn mergesort | |
ad3c9f2a | 120 | function is a modified merge sort with exponential search, |
5b2abdfb A |
121 | intended for sorting data with pre-existing order. |
122 | .Pp | |
123 | The | |
124 | .Fn qsort | |
125 | and | |
126 | .Fn heapsort | |
127 | functions sort an array of | |
ad3c9f2a | 128 | .Fa nel |
5b2abdfb A |
129 | objects, the initial member of which is pointed to by |
130 | .Fa base . | |
131 | The size of each object is specified by | |
ad3c9f2a | 132 | .Fa width . |
9385eb3d A |
133 | The |
134 | .Fn mergesort | |
135 | function | |
5b2abdfb A |
136 | behaves similarly, but |
137 | .Em requires | |
138 | that | |
ad3c9f2a A |
139 | .Fa width |
140 | be greater than or equal to | |
5b2abdfb A |
141 | .Dq "sizeof(void *) / 2" . |
142 | .Pp | |
143 | The contents of the array | |
144 | .Fa base | |
145 | are sorted in ascending order according to | |
146 | a comparison function pointed to by | |
147 | .Fa compar , | |
148 | which requires two arguments pointing to the objects being | |
149 | compared. | |
150 | .Pp | |
151 | The comparison function must return an integer less than, equal to, or | |
152 | greater than zero if the first argument is considered to be respectively | |
153 | less than, equal to, or greater than the second. | |
154 | .Pp | |
9385eb3d A |
155 | The |
156 | .Fn qsort_r | |
157 | function behaves identically to | |
158 | .Fn qsort , | |
159 | except that it takes an additional argument, | |
160 | .Fa thunk , | |
161 | which is passed unchanged as the first argument to function pointed to | |
162 | .Fa compar . | |
163 | This allows the comparison function to access additional | |
164 | data without using global variables, and thus | |
165 | .Fn qsort_r | |
166 | is suitable for use in functions which must be reentrant. | |
167 | .Pp | |
168 | The algorithms implemented by | |
169 | .Fn qsort , | |
170 | .Fn qsort_r , | |
5b2abdfb A |
171 | and |
172 | .Fn heapsort | |
173 | are | |
174 | .Em not | |
ad3c9f2a | 175 | stable; that is, if two members compare as equal, their order in |
5b2abdfb | 176 | the sorted array is undefined. |
9385eb3d | 177 | The |
5b2abdfb | 178 | .Fn mergesort |
9385eb3d | 179 | algorithm is stable. |
5b2abdfb A |
180 | .Pp |
181 | The | |
182 | .Fn qsort | |
9385eb3d A |
183 | and |
184 | .Fn qsort_r | |
3d9156a7 A |
185 | functions are an implementation of C.A.R. |
186 | Hoare's | |
9385eb3d A |
187 | .Dq quicksort |
188 | algorithm, | |
3d9156a7 A |
189 | a variant of partition-exchange sorting; in particular, see |
190 | .An D.E. Knuth Ns 's | |
191 | .%T "Algorithm Q" . | |
9385eb3d | 192 | .Sy Quicksort |
5b2abdfb A |
193 | takes O N lg N average time. |
194 | This implementation uses median selection to avoid its | |
195 | O N**2 worst-case behavior. | |
196 | .Pp | |
197 | The | |
198 | .Fn heapsort | |
3d9156a7 A |
199 | function is an implementation of |
200 | .An "J.W.J. William" Ns 's | |
9385eb3d A |
201 | .Dq heapsort |
202 | algorithm, | |
3d9156a7 A |
203 | a variant of selection sorting; in particular, see |
204 | .An "D.E. Knuth" Ns 's | |
205 | .%T "Algorithm H" . | |
9385eb3d | 206 | .Sy Heapsort |
5b2abdfb A |
207 | takes O N lg N worst-case time. |
208 | Its | |
209 | .Em only | |
210 | advantage over | |
211 | .Fn qsort | |
212 | is that it uses almost no additional memory; while | |
213 | .Fn qsort | |
214 | does not allocate memory, it is implemented using recursion. | |
215 | .Pp | |
216 | The function | |
217 | .Fn mergesort | |
218 | requires additional memory of size | |
ad3c9f2a A |
219 | .Fa nel * |
220 | .Fa width | |
5b2abdfb | 221 | bytes; it should be used only when space is not at a premium. |
9385eb3d A |
222 | The |
223 | .Fn mergesort | |
224 | function | |
5b2abdfb A |
225 | is optimized for data with pre-existing order; its worst case |
226 | time is O N lg N; its best case is O N. | |
227 | .Pp | |
228 | Normally, | |
229 | .Fn qsort | |
230 | is faster than | |
231 | .Fn mergesort | |
ad3c9f2a | 232 | which is faster than |
5b2abdfb A |
233 | .Fn heapsort . |
234 | Memory availability and pre-existing order in the data can make this | |
235 | untrue. | |
ad3c9f2a A |
236 | #ifdef UNIFDEF_BLOCKS |
237 | .Pp | |
238 | The | |
239 | .Fn heapsort_b , | |
240 | .Fn mergesort_b , | |
241 | and | |
242 | .Fn qsort_b | |
243 | routines are like the corresponding routines without the _b suffix, expect | |
244 | that the | |
245 | .Fa compar | |
246 | callback is a block pointer instead of a function pointer. | |
247 | #endif | |
5b2abdfb A |
248 | .Sh RETURN VALUES |
249 | The | |
ad3c9f2a A |
250 | #ifdef UNIFDEF_BLOCKS |
251 | .Fn qsort , | |
252 | .Fn qsort_b | |
253 | #else | |
5b2abdfb | 254 | .Fn qsort |
ad3c9f2a | 255 | #endif |
9385eb3d A |
256 | and |
257 | .Fn qsort_r | |
258 | functions | |
259 | return no value. | |
5b2abdfb | 260 | .Pp |
ad3c9f2a A |
261 | #ifdef UNIFDEF_BLOCKS |
262 | .ds HEAPSORT_B heapsort_b | |
263 | .ds MERGESORT_B mergesort_b | |
264 | #endif | |
265 | .Rv -std heapsort \*[HEAPSORT_B] mergesort \*[MERGESORT_B] | |
1f2f436a A |
266 | .Sh COMPATIBILITY |
267 | Previous versions of | |
268 | .Fn qsort | |
269 | did not permit the comparison routine itself to call | |
270 | .Fn qsort 3 . | |
271 | This is no longer true. | |
5b2abdfb A |
272 | .Sh ERRORS |
273 | The | |
ad3c9f2a A |
274 | #ifdef UNIFDEF_BLOCKS |
275 | .Fn heapsort , | |
276 | .Fn heapsort_b , | |
277 | .Fn mergesort , | |
278 | and | |
279 | .Fn mergesort_b | |
280 | #else | |
5b2abdfb A |
281 | .Fn heapsort |
282 | and | |
283 | .Fn mergesort | |
ad3c9f2a | 284 | #endif |
5b2abdfb A |
285 | functions succeed unless: |
286 | .Bl -tag -width Er | |
287 | .It Bq Er EINVAL | |
288 | The | |
ad3c9f2a | 289 | .Fa width |
5b2abdfb A |
290 | argument is zero, or, |
291 | the | |
ad3c9f2a | 292 | .Fa width |
5b2abdfb A |
293 | argument to |
294 | .Fn mergesort | |
ad3c9f2a A |
295 | #ifdef UNIFDEF_BLOCKS |
296 | or | |
297 | .Fn mergesort_b | |
298 | #endif | |
5b2abdfb A |
299 | is less than |
300 | .Dq "sizeof(void *) / 2" . | |
301 | .It Bq Er ENOMEM | |
9385eb3d | 302 | The |
ad3c9f2a A |
303 | #ifdef UNIFDEF_BLOCKS |
304 | .Fn heapsort , | |
305 | .Fn heapsort_b , | |
306 | .Fn mergesort , | |
307 | or | |
308 | .Fn mergesort_b | |
309 | #else | |
9385eb3d | 310 | .Fn heapsort |
5b2abdfb A |
311 | or |
312 | .Fn mergesort | |
ad3c9f2a | 313 | #endif |
9385eb3d | 314 | functions |
5b2abdfb A |
315 | were unable to allocate memory. |
316 | .El | |
5b2abdfb A |
317 | .Sh SEE ALSO |
318 | .Xr sort 1 , | |
319 | .Xr radixsort 3 | |
320 | .Rs | |
321 | .%A Hoare, C.A.R. | |
322 | .%D 1962 | |
323 | .%T "Quicksort" | |
324 | .%J "The Computer Journal" | |
325 | .%V 5:1 | |
326 | .%P pp. 10-15 | |
327 | .Re | |
328 | .Rs | |
329 | .%A Williams, J.W.J | |
330 | .%D 1964 | |
331 | .%T "Heapsort" | |
332 | .%J "Communications of the ACM" | |
333 | .%V 7:1 | |
334 | .%P pp. 347-348 | |
335 | .Re | |
336 | .Rs | |
337 | .%A Knuth, D.E. | |
338 | .%D 1968 | |
339 | .%B "The Art of Computer Programming" | |
340 | .%V Vol. 3 | |
341 | .%T "Sorting and Searching" | |
342 | .%P pp. 114-123, 145-149 | |
343 | .Re | |
344 | .Rs | |
3d9156a7 | 345 | .%A McIlroy, P.M. |
5b2abdfb A |
346 | .%T "Optimistic Sorting and Information Theoretic Complexity" |
347 | .%J "Fourth Annual ACM-SIAM Symposium on Discrete Algorithms" | |
348 | .%V January 1992 | |
349 | .Re | |
350 | .Rs | |
351 | .%A Bentley, J.L. | |
3d9156a7 | 352 | .%A McIlroy, M.D. |
5b2abdfb | 353 | .%T "Engineering a Sort Function" |
3d9156a7 A |
354 | .%J "Software--Practice and Experience" |
355 | .%V Vol. 23(11) | |
356 | .%P pp. 1249-1265 | |
357 | .%D November\ 1993 | |
5b2abdfb A |
358 | .Re |
359 | .Sh STANDARDS | |
360 | The | |
361 | .Fn qsort | |
362 | function | |
363 | conforms to | |
364 | .St -isoC . |