xref: /freebsd/lib/libc/db/btree/bt_delete.c (revision afe61c15161c324a7af299a9b8457aba5afc92db)
1 /*-
2  * Copyright (c) 1990, 1993
3  *	The Regents of the University of California.  All rights reserved.
4  *
5  * This code is derived from software contributed to Berkeley by
6  * Mike Olson.
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.
16  * 3. All advertising materials mentioning features or use of this software
17  *    must display the following acknowledgement:
18  *	This product includes software developed by the University of
19  *	California, Berkeley and its contributors.
20  * 4. Neither the name of the University nor the names of its contributors
21  *    may be used to endorse or promote products derived from this software
22  *    without specific prior written permission.
23  *
24  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
25  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
28  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
29  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
30  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
31  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
33  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34  * SUCH DAMAGE.
35  */
36 
37 #if defined(LIBC_SCCS) && !defined(lint)
38 static char sccsid[] = "@(#)bt_delete.c	8.3 (Berkeley) 2/21/94";
39 #endif /* LIBC_SCCS and not lint */
40 
41 #include <sys/types.h>
42 
43 #include <errno.h>
44 #include <stdio.h>
45 #include <string.h>
46 
47 #include <db.h>
48 #include "btree.h"
49 
50 static int bt_bdelete __P((BTREE *, const DBT *));
51 
52 /*
53  * __BT_DELETE -- Delete the item(s) referenced by a key.
54  *
55  * Parameters:
56  *	dbp:	pointer to access method
57  *	key:	key to delete
58  *	flags:	R_CURSOR if deleting what the cursor references
59  *
60  * Returns:
61  *	RET_ERROR, RET_SUCCESS and RET_SPECIAL if the key not found.
62  */
63 int
64 __bt_delete(dbp, key, flags)
65 	const DB *dbp;
66 	const DBT *key;
67 	u_int flags;
68 {
69 	BTREE *t;
70 	int status;
71 
72 	t = dbp->internal;
73 
74 	/* Toss any page pinned across calls. */
75 	if (t->bt_pinned != NULL) {
76 		mpool_put(t->bt_mp, t->bt_pinned, 0);
77 		t->bt_pinned = NULL;
78 	}
79 
80 	if (ISSET(t, B_RDONLY)) {
81 		errno = EPERM;
82 		return (RET_ERROR);
83 	}
84 
85 	switch(flags) {
86 	case 0:
87 		status = bt_bdelete(t, key);
88 		break;
89 	case R_CURSOR:
90 		/*
91 		 * If flags is R_CURSOR, delete the cursor; must already have
92 		 * started a scan and not have already deleted the record.  For
93 		 * the delete cursor bit to have been set requires that the
94 		 * scan be initialized, so no reason to check.
95 		 */
96 		if (!ISSET(t, B_SEQINIT))
97                         goto einval;
98 		status = ISSET(t, B_DELCRSR) ?
99 		    RET_SPECIAL : __bt_crsrdel(t, &t->bt_bcursor);
100 		break;
101 	default:
102 einval:		errno = EINVAL;
103 		return (RET_ERROR);
104 	}
105 	if (status == RET_SUCCESS)
106 		SET(t, B_MODIFIED);
107 	return (status);
108 }
109 
110 /*
111  * BT_BDELETE -- Delete all key/data pairs matching the specified key.
112  *
113  * Parameters:
114  *	tree:	tree
115  *	key:	key to delete
116  *
117  * Returns:
118  *	RET_ERROR, RET_SUCCESS and RET_SPECIAL if the key not found.
119  */
120 static int
121 bt_bdelete(t, key)
122 	BTREE *t;
123 	const DBT *key;
124 {
125 	EPG *e, save;
126 	PAGE *h;
127 	pgno_t cpgno, pg;
128 	indx_t cindex;
129 	int deleted, dirty1, dirty2, exact;
130 
131 	/* Find any matching record; __bt_search pins the page. */
132 	if ((e = __bt_search(t, key, &exact)) == NULL)
133 		return (RET_ERROR);
134 	if (!exact) {
135 		mpool_put(t->bt_mp, e->page, 0);
136 		return (RET_SPECIAL);
137 	}
138 
139 	/*
140 	 * Delete forward, then delete backward, from the found key.  The
141 	 * ordering is so that the deletions don't mess up the page refs.
142 	 * The first loop deletes the key from the original page, the second
143 	 * unpins the original page.  In the first loop, dirty1 is set if
144 	 * the original page is modified, and dirty2 is set if any subsequent
145 	 * pages are modified.  In the second loop, dirty1 starts off set if
146 	 * the original page has been modified, and is set if any subsequent
147 	 * pages are modified.
148 	 *
149 	 * If find the key referenced by the cursor, don't delete it, just
150 	 * flag it for future deletion.  The cursor page number is P_INVALID
151 	 * unless the sequential scan is initialized, so no reason to check.
152 	 * A special case is when the already deleted cursor record was the
153 	 * only record found.  If so, then the delete opertion fails as no
154 	 * records were deleted.
155 	 *
156 	 * Cycle in place in the current page until the current record doesn't
157 	 * match the key or the page is empty.  If the latter, walk forward,
158 	 * skipping empty pages and repeating until a record doesn't match
159 	 * the key or the end of the tree is reached.
160 	 */
161 	cpgno = t->bt_bcursor.pgno;
162 	cindex = t->bt_bcursor.index;
163 	save = *e;
164 	dirty1 = 0;
165 	for (h = e->page, deleted = 0;;) {
166 		dirty2 = 0;
167 		do {
168 			if (h->pgno == cpgno && e->index == cindex) {
169 				if (!ISSET(t, B_DELCRSR)) {
170 					SET(t, B_DELCRSR);
171 					deleted = 1;
172 				}
173 				++e->index;
174 			} else {
175 				if (__bt_dleaf(t, h, e->index)) {
176 					if (h->pgno != save.page->pgno)
177 						mpool_put(t->bt_mp, h, dirty2);
178 					mpool_put(t->bt_mp, save.page, dirty1);
179 					return (RET_ERROR);
180 				}
181 				if (h->pgno == save.page->pgno)
182 					dirty1 = MPOOL_DIRTY;
183 				else
184 					dirty2 = MPOOL_DIRTY;
185 				deleted = 1;
186 			}
187 		} while (e->index < NEXTINDEX(h) && __bt_cmp(t, key, e) == 0);
188 
189 		/*
190 		 * Quit if didn't find a match, no next page, or first key on
191 		 * the next page doesn't match.  Don't unpin the original page
192 		 * unless an error occurs.
193 		 */
194 		if (e->index < NEXTINDEX(h))
195 			break;
196 		for (;;) {
197 			if ((pg = h->nextpg) == P_INVALID)
198 				goto done1;
199 			if (h->pgno != save.page->pgno)
200 				mpool_put(t->bt_mp, h, dirty2);
201 			if ((h = mpool_get(t->bt_mp, pg, 0)) == NULL) {
202 				mpool_put(t->bt_mp, save.page, dirty1);
203 				return (RET_ERROR);
204 			}
205 			if (NEXTINDEX(h) != 0) {
206 				e->page = h;
207 				e->index = 0;
208 				break;
209 			}
210 		}
211 
212 		if (__bt_cmp(t, key, e) != 0)
213 			break;
214 	}
215 
216 	/*
217 	 * Reach here with the original page and the last page referenced
218 	 * pinned (they may be the same).  Release it if not the original.
219 	 */
220 done1:	if (h->pgno != save.page->pgno)
221 		mpool_put(t->bt_mp, h, dirty2);
222 
223 	/*
224 	 * Walk backwards from the record previous to the record returned by
225 	 * __bt_search, skipping empty pages, until a record doesn't match
226 	 * the key or reach the beginning of the tree.
227 	 */
228 	*e = save;
229 	for (;;) {
230 		if (e->index)
231 			--e->index;
232 		for (h = e->page; e->index; --e->index) {
233 			if (__bt_cmp(t, key, e) != 0)
234 				goto done2;
235 			if (h->pgno == cpgno && e->index == cindex) {
236 				if (!ISSET(t, B_DELCRSR)) {
237 					SET(t, B_DELCRSR);
238 					deleted = 1;
239 				}
240 			} else {
241 				if (__bt_dleaf(t, h, e->index) == RET_ERROR) {
242 					mpool_put(t->bt_mp, h, dirty1);
243 					return (RET_ERROR);
244 				}
245 				if (h->pgno == save.page->pgno)
246 					dirty1 = MPOOL_DIRTY;
247 				deleted = 1;
248 			}
249 		}
250 
251 		if ((pg = h->prevpg) == P_INVALID)
252 			goto done2;
253 		mpool_put(t->bt_mp, h, dirty1);
254 		dirty1 = 0;
255 		if ((e->page = mpool_get(t->bt_mp, pg, 0)) == NULL)
256 			return (RET_ERROR);
257 		e->index = NEXTINDEX(e->page);
258 	}
259 
260 	/*
261 	 * Reach here with the last page that was looked at pinned.  Release
262 	 * it.
263 	 */
264 done2:	mpool_put(t->bt_mp, h, dirty1);
265 	return (deleted ? RET_SUCCESS : RET_SPECIAL);
266 }
267 
268 /*
269  * __BT_DLEAF -- Delete a single record from a leaf page.
270  *
271  * Parameters:
272  *	t:	tree
273  *	index:	index on current page to delete
274  *
275  * Returns:
276  *	RET_SUCCESS, RET_ERROR.
277  */
278 int
279 __bt_dleaf(t, h, index)
280 	BTREE *t;
281 	PAGE *h;
282 	indx_t index;
283 {
284 	register BLEAF *bl;
285 	register indx_t cnt, *ip, offset;
286 	register size_t nbytes;
287 	char *from;
288 	void *to;
289 
290 	/*
291 	 * Delete a record from a btree leaf page.  Internal records are never
292 	 * deleted from internal pages, regardless of the records that caused
293 	 * them to be added being deleted.  Pages made empty by deletion are
294 	 * not reclaimed.  They are, however, made available for reuse.
295 	 *
296 	 * Pack the remaining entries at the end of the page, shift the indices
297 	 * down, overwriting the deleted record and its index.  If the record
298 	 * uses overflow pages, make them available for reuse.
299 	 */
300 	to = bl = GETBLEAF(h, index);
301 	if (bl->flags & P_BIGKEY && __ovfl_delete(t, bl->bytes) == RET_ERROR)
302 		return (RET_ERROR);
303 	if (bl->flags & P_BIGDATA &&
304 	    __ovfl_delete(t, bl->bytes + bl->ksize) == RET_ERROR)
305 		return (RET_ERROR);
306 	nbytes = NBLEAF(bl);
307 
308 	/*
309 	 * Compress the key/data pairs.  Compress and adjust the [BR]LEAF
310 	 * offsets.  Reset the headers.
311 	 */
312 	from = (char *)h + h->upper;
313 	memmove(from + nbytes, from, (char *)to - from);
314 	h->upper += nbytes;
315 
316 	offset = h->linp[index];
317 	for (cnt = index, ip = &h->linp[0]; cnt--; ++ip)
318 		if (ip[0] < offset)
319 			ip[0] += nbytes;
320 	for (cnt = NEXTINDEX(h) - index; --cnt; ++ip)
321 		ip[0] = ip[1] < offset ? ip[1] + nbytes : ip[1];
322 	h->lower -= sizeof(indx_t);
323 	return (RET_SUCCESS);
324 }
325