db/btree/bt_put.c

   1 /*
   2  * Copyright (c) 1999 Apple Computer, Inc. All rights reserved.
   3  *
   4  * @APPLE_LICENSE_HEADER_START@
   5  *
   6  * Copyright (c) 1999-2003 Apple Computer, Inc.  All Rights Reserved.
   7  *
   8  * This file contains Original Code and/or Modifications of Original Code
   9  * as defined in and that are subject to the Apple Public Source License
  10  * Version 2.0 (the 'License'). You may not use this file except in
  11  * compliance with the License. Please obtain a copy of the License at
  12  * http://www.opensource.apple.com/apsl/ and read it before using this
  13  * file.
  14  *
  15  * The Original Code and all software distributed under the License are
  16  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  17  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  18  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  19  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  20  * Please see the License for the specific language governing rights and
  21  * limitations under the License.
  22  *
  23  * @APPLE_LICENSE_HEADER_END@
  24  */
  25 /*
  26  * Copyright (c) 1990, 1993
  27  *      The Regents of the University of California.  All rights reserved.
  28  *
  29  * This code is derived from software contributed to Berkeley by
  30  * Mike Olson.
  31  *
  32  * Redistribution and use in source and binary forms, with or without
  33  * modification, are permitted provided that the following conditions
  34  * are met:
  35  * 1. Redistributions of source code must retain the above copyright
  36  *    notice, this list of conditions and the following disclaimer.
  37  * 2. Redistributions in binary form must reproduce the above copyright
  38  *    notice, this list of conditions and the following disclaimer in the
  39  *    documentation and/or other materials provided with the distribution.
  40  * 3. All advertising materials mentioning features or use of this software
  41  *    must display the following acknowledgement:
  42  *      This product includes software developed by the University of
  43  *      California, Berkeley and its contributors.
  44  * 4. Neither the name of the University nor the names of its contributors
  45  *    may be used to endorse or promote products derived from this software
  46  *    without specific prior written permission.
  47  *
  48  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  49  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  50  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  51  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  52  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  53  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  54  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  55  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  56  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  57  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  58  * SUCH DAMAGE.
  59  */
  60
  61
  62 #include <sys/types.h>
  63
  64 #include <errno.h>
  65 #include <stdio.h>
  66 #include <stdlib.h>
  67 #include <string.h>
  68
  69 #include <db.h>
  70 #include "btree.h"
  71
  72 static EPG *bt_fast __P((BTREE *, const DBT *, const DBT *, int *));
  73
  74 /*
  75  * __BT_PUT -- Add a btree item to the tree.
  76  *
  77  * Parameters:
  78  *      dbp:    pointer to access method
  79  *      key:    key
  80  *      data:   data
  81  *      flag:   R_NOOVERWRITE
  82  *
  83  * Returns:
  84  *      RET_ERROR, RET_SUCCESS and RET_SPECIAL if the key is already in the
  85  *      tree and R_NOOVERWRITE specified.
  86  */
  87 int
  88 __bt_put(dbp, key, data, flags)
  89         const DB *dbp;
  90         DBT *key;
  91         const DBT *data;
  92         u_int flags;
  93 {
  94         BTREE *t;
  95         DBT tkey, tdata;
  96         EPG *e;
  97         PAGE *h;
  98         indx_t index, nxtindex;
  99         pgno_t pg;
 100         size_t nbytes;
 101         int dflags, exact, status;
 102         char *dest, db[NOVFLSIZE], kb[NOVFLSIZE];
 103
 104         t = dbp->internal;
 105
 106         /* Toss any page pinned across calls. */
 107         if (t->bt_pinned != NULL) {
 108                 mpool_put(t->bt_mp, t->bt_pinned, 0);
 109                 t->bt_pinned = NULL;
 110         }
 111
 112         switch (flags) {
 113         case R_CURSOR:
 114                 if (!ISSET(t, B_SEQINIT))
 115                         goto einval;
 116                 if (ISSET(t, B_DELCRSR))
 117                         goto einval;
 118                 break;
 119         case 0:
 120         case R_NOOVERWRITE:
 121                 break;
 122         default:
 123 einval:         errno = EINVAL;
 124                 return (RET_ERROR);
 125         }
 126
 127         if (ISSET(t, B_RDONLY)) {
 128                 errno = EPERM;
 129                 return (RET_ERROR);
 130         }
 131
 132         /*
 133          * If the key/data won't fit on a page, store it on indirect pages.
 134          * Only store the key on the overflow page if it's too big after the
 135          * data is on an overflow page.
 136          *
 137          * XXX
 138          * If the insert fails later on, these pages aren't recovered.
 139          */
 140         dflags = 0;
 141         if (key->size + data->size > t->bt_ovflsize) {
 142                 if (key->size > t->bt_ovflsize) {
 143 storekey:               if (__ovfl_put(t, key, &pg) == RET_ERROR)
 144                                 return (RET_ERROR);
 145                         tkey.data = kb;
 146                         tkey.size = NOVFLSIZE;
 147                         memmove(kb, &pg, sizeof(pgno_t));
 148                         memmove(kb + sizeof(pgno_t),
 149                             &key->size, sizeof(size_t));
 150                         dflags |= P_BIGKEY;
 151                         key = &tkey;
 152                 }
 153                 if (key->size + data->size > t->bt_ovflsize) {
 154                         if (__ovfl_put(t, data, &pg) == RET_ERROR)
 155                                 return (RET_ERROR);
 156                         tdata.data = db;
 157                         tdata.size = NOVFLSIZE;
 158                         memmove(db, &pg, sizeof(pgno_t));
 159                         memmove(db + sizeof(pgno_t),
 160                             &data->size, sizeof(size_t));
 161                         dflags |= P_BIGDATA;
 162                         data = &tdata;
 163                 }
 164                 if (key->size + data->size > t->bt_ovflsize)
 165                         goto storekey;
 166         }
 167
 168         /* Replace the cursor. */
 169         if (flags == R_CURSOR) {
 170                 if ((h = mpool_get(t->bt_mp, t->bt_bcursor.pgno, 0)) == NULL)
 171                         return (RET_ERROR);
 172                 index = t->bt_bcursor.index;
 173                 goto delete;
 174         }
 175
 176         /*
 177          * Find the key to delete, or, the location at which to insert.  Bt_fast
 178          * and __bt_search pin the returned page.
 179          */
 180         if (t->bt_order == NOT || (e = bt_fast(t, key, data, &exact)) == NULL)
 181                 if ((e = __bt_search(t, key, &exact)) == NULL)
 182                         return (RET_ERROR);
 183         h = e->page;
 184         index = e->index;
 185
 186         /*
 187          * Add the specified key/data pair to the tree.  If an identical key
 188          * is already in the tree, and R_NOOVERWRITE is set, an error is
 189          * returned.  If R_NOOVERWRITE is not set, the key is either added (if
 190          * duplicates are permitted) or an error is returned.
 191          *
 192          * Pages are split as required.
 193          */
 194         switch (flags) {
 195         case R_NOOVERWRITE:
 196                 if (!exact)
 197                         break;
 198                 /*
 199                  * One special case is if the cursor references the record and
 200                  * it's been flagged for deletion.  Then, we delete the record,
 201                  * leaving the cursor there -- this means that the inserted
 202                  * record will not be seen in a cursor scan.
 203                  */
 204                 if (ISSET(t, B_DELCRSR) && t->bt_bcursor.pgno == h->pgno &&
 205                     t->bt_bcursor.index == index) {
 206                         CLR(t, B_DELCRSR);
 207                         goto delete;
 208                 }
 209                 mpool_put(t->bt_mp, h, 0);
 210                 return (RET_SPECIAL);
 211         default:
 212                 if (!exact || !ISSET(t, B_NODUPS))
 213                         break;
 214 delete:         if (__bt_dleaf(t, h, index) == RET_ERROR) {
 215                         mpool_put(t->bt_mp, h, 0);
 216                         return (RET_ERROR);
 217                 }
 218                 break;
 219         }
 220
 221         /*
 222          * If not enough room, or the user has put a ceiling on the number of
 223          * keys permitted in the page, split the page.  The split code will
 224          * insert the key and data and unpin the current page.  If inserting
 225          * into the offset array, shift the pointers up.
 226          */
 227         nbytes = NBLEAFDBT(key->size, data->size);
 228         if (h->upper - h->lower < nbytes + sizeof(indx_t)) {
 229                 if ((status = __bt_split(t, h, key,
 230                     data, dflags, nbytes, index)) != RET_SUCCESS)
 231                         return (status);
 232                 goto success;
 233         }
 234
 235         if (index < (nxtindex = NEXTINDEX(h)))
 236                 memmove(h->linp + index + 1, h->linp + index,
 237                     (nxtindex - index) * sizeof(indx_t));
 238         h->lower += sizeof(indx_t);
 239
 240         h->linp[index] = h->upper -= nbytes;
 241         dest = (char *)h + h->upper;
 242         WR_BLEAF(dest, key, data, dflags);
 243
 244         if (t->bt_order == NOT)
 245                 if (h->nextpg == P_INVALID) {
 246                         if (index == NEXTINDEX(h) - 1) {
 247                                 t->bt_order = FORWARD;
 248                                 t->bt_last.index = index;
 249                                 t->bt_last.pgno = h->pgno;
 250                         }
 251                 } else if (h->prevpg == P_INVALID) {
 252                         if (index == 0) {
 253                                 t->bt_order = BACK;
 254                                 t->bt_last.index = 0;
 255                                 t->bt_last.pgno = h->pgno;
 256                         }
 257                 }
 258
 259         mpool_put(t->bt_mp, h, MPOOL_DIRTY);
 260
 261 success:
 262         if (flags == R_SETCURSOR) {
 263                 t->bt_bcursor.pgno = e->page->pgno;
 264                 t->bt_bcursor.index = e->index;
 265         }
 266         SET(t, B_MODIFIED);
 267         return (RET_SUCCESS);
 268 }
 269
 270 #ifdef STATISTICS
 271 u_long bt_cache_hit, bt_cache_miss;
 272 #endif
 273
 274 /*
 275  * BT_FAST -- Do a quick check for sorted data.
 276  *
 277  * Parameters:
 278  *      t:      tree
 279  *      key:    key to insert
 280  *
 281  * Returns:
 282  *      EPG for new record or NULL if not found.
 283  */
 284 static EPG *
 285 bt_fast(t, key, data, exactp)
 286         BTREE *t;
 287         const DBT *key, *data;
 288         int *exactp;
 289 {
 290         PAGE *h;
 291         size_t nbytes;
 292         int cmp;
 293
 294         if ((h = mpool_get(t->bt_mp, t->bt_last.pgno, 0)) == NULL) {
 295                 t->bt_order = NOT;
 296                 return (NULL);
 297         }
 298         t->bt_cur.page = h;
 299         t->bt_cur.index = t->bt_last.index;
 300
 301         /*
 302          * If won't fit in this page or have too many keys in this page, have
 303          * to search to get split stack.
 304          */
 305         nbytes = NBLEAFDBT(key->size, data->size);
 306         if (h->upper - h->lower < nbytes + sizeof(indx_t))
 307                 goto miss;
 308
 309         if (t->bt_order == FORWARD) {
 310                 if (t->bt_cur.page->nextpg != P_INVALID)
 311                         goto miss;
 312                 if (t->bt_cur.index != NEXTINDEX(h) - 1)
 313                         goto miss;
 314                 if ((cmp = __bt_cmp(t, key, &t->bt_cur)) < 0)
 315                         goto miss;
 316                 t->bt_last.index = cmp ? ++t->bt_cur.index : t->bt_cur.index;
 317         } else {
 318                 if (t->bt_cur.page->prevpg != P_INVALID)
 319                         goto miss;
 320                 if (t->bt_cur.index != 0)
 321                         goto miss;
 322                 if ((cmp = __bt_cmp(t, key, &t->bt_cur)) > 0)
 323                         goto miss;
 324                 t->bt_last.index = 0;
 325         }
 326         *exactp = cmp == 0;
 327 #ifdef STATISTICS
 328         ++bt_cache_hit;
 329 #endif
 330         return (&t->bt_cur);
 331
 332 miss:
 333 #ifdef STATISTICS
 334         ++bt_cache_miss;
 335 #endif
 336         t->bt_order = NOT;
 337         mpool_put(t->bt_mp, h, 0);
 338         return (NULL);
 339 }