1.\" $OpenBSD: btree.3,v 1.19 2007/05/31 19:19:27 jmc Exp $ 2.\" $NetBSD: btree.3,v 1.6 1996/05/03 21:26:48 cgd Exp $ 3.\" 4.\" Copyright (c) 1997, Phillip F Knaack. All rights reserved. 5.\" 6.\" Copyright (c) 1990, 1993 7.\" The Regents of the University of California. All rights reserved. 8.\" 9.\" Redistribution and use in source and binary forms, with or without 10.\" modification, are permitted provided that the following conditions 11.\" are met: 12.\" 1. Redistributions of source code must retain the above copyright 13.\" notice, this list of conditions and the following disclaimer. 14.\" 2. Redistributions in binary form must reproduce the above copyright 15.\" notice, this list of conditions and the following disclaimer in the 16.\" documentation and/or other materials provided with the distribution. 17.\" 3. Neither the name of the University nor the names of its contributors 18.\" may be used to endorse or promote products derived from this software 19.\" without specific prior written permission. 20.\" 21.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 22.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 23.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 24.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 25.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 26.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 27.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 28.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 29.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 30.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 31.\" SUCH DAMAGE. 32.\" 33.\" @(#)btree.3 8.4 (Berkeley) 8/18/94 34.\" 35.Dd $Mdocdate: May 31 2007 $ 36.Dt BTREE 3 37.Os 38.Sh NAME 39.Nm btree 40.Nd btree database access method 41.Sh SYNOPSIS 42.Fd #include <sys/types.h> 43.Fd #include <db.h> 44.Sh DESCRIPTION 45The 46.Fn dbopen 47routine is the library interface to database files. 48One of the supported file formats is btree files. 49The general description of the database access methods is in 50.Xr dbopen 3 . 51This manual page describes only the btree specific information. 52.Pp 53The btree data structure is a sorted, balanced tree structure storing 54associated key/data pairs. 55.Pp 56The btree access method specific data structure provided to 57.Fn dbopen 58is defined in the 59.Aq Pa db.h 60include file as follows: 61.Bd -literal -offset indent 62typedef struct { 63 unsigned long flags; 64 unsigned int cachesize; 65 int maxkeypage; 66 int minkeypage; 67 unsigned int psize; 68 int (*compare)(const DBT *key1, const DBT *key2); 69 size_t (*prefix)(const DBT *key1, const DBT *key2); 70 int lorder; 71} BTREEINFO; 72.Ed 73.Pp 74The elements of this structure are as follows: 75.Bl -tag -width "XXXXXX" 76.It Fa flags 77The flag value is specified by 78.Tn OR Ns 'ing 79any of the following values: 80.Bl -tag -width XXXXX 81.It Dv R_DUP 82Permit duplicate keys in the tree, i.e., permit insertion if the key to be 83inserted already exists in the tree. 84The default behavior, as described in 85.Xr dbopen 3 , 86is to overwrite a matching key when inserting a new key or to fail if 87the 88.Dv R_NOOVERWRITE 89flag is specified. 90The 91.Dv R_DUP 92flag is overridden by the 93.Dv R_NOOVERWRITE 94flag, and if the 95.Dv R_NOOVERWRITE 96flag is specified, attempts to insert duplicate keys into 97the tree will fail. 98.Pp 99If the database contains duplicate keys, the order of retrieval of 100key/data pairs is undefined if the 101.Fn get 102routine is used; however, 103.Fn seq 104routine calls with the 105.Dv R_CURSOR 106flag set will always return the logical 107``first'' of any group of duplicate keys. 108.El 109.It Fa cachesize 110A suggested maximum size (in bytes) of the memory cache. 111This value is 112.Em only 113advisory, and the access method will allocate more memory rather than fail. 114Since every search examines the root page of the tree, caching the most 115recently used pages substantially improves access time. 116In addition, physical writes are delayed as long as possible, so a moderate 117cache can reduce the number of I/O operations significantly. 118Obviously, using a cache increases (but only increases) the likelihood of 119corruption or lost data if the system crashes while a tree is being modified. 120If 121.Fa cachesize 122is 0 (no size is specified) a default cache is used. 123.It Fa maxkeypage 124The maximum number of keys which will be stored on any single page. 125Not currently implemented. 126.It Fa minkeypage 127The minimum number of keys which will be stored on any single page. 128This value is used to determine which keys will be stored on overflow 129pages, i.e., if a key or data item is longer than the pagesize divided 130by the minkeypage value, it will be stored on overflow pages instead 131of in the page itself. 132If 133.Fa minkeypage 134is 0 (no minimum number of keys is specified) a value of 2 is used. 135.It Fa psize 136Page size is the size (in bytes) of the pages used for nodes in the tree. 137The minimum page size is 512 bytes and the maximum page size is 64K. 138If 139.Fa psize 140is 0 (no page size is specified) a page size is chosen based on the 141underlying file system I/O block size. 142.It Fa compare 143Compare is the key comparison function. 144It must return an integer less than, equal to, or greater than zero if the 145first key argument is considered to be respectively less than, equal to, 146or greater than the second key argument. 147The same comparison function must be used on a given tree every time it 148is opened. 149If 150.Fa compare 151is 152.Dv NULL 153(no comparison function is specified), the keys are compared 154lexically, with shorter keys considered less than longer keys. 155.It Fa prefix 156Prefix is the prefix comparison function. 157If specified, this routine must return the number of bytes of the second key 158argument which are necessary to determine that it is greater than the first 159key argument. 160If the keys are equal, the key length should be returned. 161Note, the usefulness of this routine is very data dependent, but in some 162data sets it can produce significantly reduced tree sizes and search times. 163If 164.Fa prefix 165is 166.Dv NULL 167(no prefix function is specified), 168.Em and 169no comparison function is specified, a default lexical comparison routine 170is used. 171If 172.Fa prefix 173is 174.Dv NULL 175and a comparison routine is specified, no prefix comparison is done. 176.It Fa lorder 177The byte order for integers in the stored database metadata. 178The number should represent the order as an integer; for example, 179big endian order would be the number 4,321. 180If 181.Fa lorder 182is 0 (no order is specified) the current host order is used. 183.El 184.Pp 185If the file already exists (and the 186.Dv O_TRUNC 187flag is not specified), the 188values specified for the parameters 189.Fa flags , 190.Fa lorder , 191and 192.Fa psize 193are ignored in favor of the values used when the tree was created. 194.Pp 195Forward sequential scans of a tree are from the least key to the greatest. 196.Pp 197Space freed up by deleting key/data pairs from the tree is never reclaimed, 198although it is normally made available for reuse. 199This means that the btree storage structure is grow-only. 200The only solutions are to avoid excessive deletions, or to create a fresh 201tree periodically from a scan of an existing one. 202.Pp 203Searches, insertions, and deletions in a btree will all complete in 204O(lg\ base\ N) where base is the average fill factor. 205Often, inserting ordered data into btrees results in a low fill factor. 206This implementation has been modified to make ordered insertion the best 207case, resulting in a much better than normal page fill factor. 208.Sh ERRORS 209The 210.Nm 211access method routines may fail and set 212.Va errno 213for any of the errors specified for the library routine 214.Xr dbopen 3 . 215.Sh SEE ALSO 216.Xr dbopen 3 , 217.Xr hash 3 , 218.Xr mpool 3 , 219.Xr recno 3 220.Rs 221.%T "The Ubiquitous B-tree" 222.%A Douglas Comer 223.%J ACM Comput. Surv. 11 224.%D June 1979 225.%P pp 121-138 226.Re 227.Rs 228.%T "Prefix B-trees" 229.%A Rudolf Bayer 230.%A Karl Unterauer 231.%J ACM Transactions on Database Systems 232.%V Vol. 2 , 1 233.%D March 1977 234.%P pp 11-26 235.Re 236.Rs 237.%B "The Art of Computer Programming Vol. 3: Sorting and Searching" 238.%A D. E. Knuth 239.%D 1968 240.%P pp 471-480 241.Re 242.Sh BUGS 243Only big and little endian byte order is supported. 244