rotoscope 0.3.0.pre.6 → 0.3.0.pre.7

Sign up to get free protection for your applications and to get access to all the features.
Files changed (10) hide show
  1. checksums.yaml +4 -4
  2. data/ext/rotoscope/extconf.rb +0 -3
  3. data/ext/rotoscope/rotoscope.c +95 -120
  4. data/ext/rotoscope/rotoscope.h +6 -18
  5. data/lib/rotoscope.rb +34 -35
  6. data/lib/rotoscope/version.rb +1 -1
  7. metadata +1 -4
  8. data/ext/rotoscope/strmemo.c +0 -33
  9. data/ext/rotoscope/strmemo.h +0 -14
  10. data/lib/uthash/uthash.h +0 -1107
data/lib/rotoscope/version.rb CHANGED
@@ -1,4 +1,4 @@
1
1
  # frozen_string_literal: true
2
2
  class Rotoscope
3
- VERSION = "0.3.0.pre.6"
3
+ VERSION = "0.3.0.pre.7"
4
4
  end
metadata CHANGED
@@ -1,7 +1,7 @@
1
1
  --- !ruby/object:Gem::Specification
2
2
  name: rotoscope
3
3
  version: !ruby/object:Gem::Version
4
- version: 0.3.0.pre.6
4
+ version: 0.3.0.pre.7
5
5
  platform: ruby
6
6
  authors:
7
7
  - Jahfer Husain
@@ -92,13 +92,10 @@ files:
92
92
  - ext/rotoscope/rotoscope.h
93
93
  - ext/rotoscope/stack.c
94
94
  - ext/rotoscope/stack.h
95
- - ext/rotoscope/strmemo.c
96
- - ext/rotoscope/strmemo.h
97
95
  - ext/rotoscope/tracepoint.c
98
96
  - ext/rotoscope/tracepoint.h
99
97
  - lib/rotoscope.rb
100
98
  - lib/rotoscope/version.rb
101
- - lib/uthash/uthash.h
102
99
  - rotoscope.gemspec
103
100
  homepage: https://github.com/shopify/rotoscope
104
101
  licenses:
data/ext/rotoscope/strmemo.c DELETED
@@ -1,33 +0,0 @@
1
- #include <ruby.h>
2
- #include <stdbool.h>
3
-
4
- #include "strmemo.h"
5
- #include "uthash.h"
6
-
7
- static unsigned long hash_djb2(unsigned char *str) {
8
- unsigned long hash = 5381;
9
- int c;
10
- while ((c = *str++)) hash = ((hash << 5) + hash) + c;
11
- return hash;
12
- }
13
-
14
- bool rs_strmemo_uniq(rs_strmemo_t **calls, unsigned char *entry) {
15
- rs_strmemo_t *c = NULL;
16
- unsigned long hashkey = hash_djb2(entry);
17
-
18
- HASH_FIND_INT(*calls, &hashkey, c);
19
- if (c != NULL) return false;
20
-
21
- c = ALLOC(rs_strmemo_t);
22
- c->key = hashkey;
23
- HASH_ADD_INT(*calls, key, c);
24
- return true;
25
- }
26
-
27
- void rs_strmemo_free(rs_strmemo_t *calls) {
28
- rs_strmemo_t *current_call, *tmp;
29
- HASH_ITER(hh, calls, current_call, tmp) {
30
- HASH_DEL(calls, current_call);
31
- xfree(current_call);
32
- }
33
- }
data/ext/rotoscope/strmemo.h DELETED
@@ -1,14 +0,0 @@
1
- #ifndef _INC_ROTOSCOPE_STRMEMO_H_
2
- #define _INC_ROTOSCOPE_STRMEMO_H_
3
-
4
- #include "uthash.h"
5
-
6
- typedef struct {
7
- unsigned long key;
8
- UT_hash_handle hh;
9
- } rs_strmemo_t;
10
-
11
- bool rs_strmemo_uniq(rs_strmemo_t **calls, unsigned char *entry);
12
- void rs_strmemo_free(rs_strmemo_t *calls);
13
-
14
- #endif
data/lib/uthash/uthash.h DELETED
@@ -1,1107 +0,0 @@
1
- /*
2
- Copyright (c) 2003-2017, Troy D. Hanson http://troydhanson.github.com/uthash/
3
- All rights reserved.
4
-
5
- Redistribution and use in source and binary forms, with or without
6
- modification, are permitted provided that the following conditions are met:
7
-
8
- * Redistributions of source code must retain the above copyright
9
- notice, this list of conditions and the following disclaimer.
10
-
11
- THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS
12
- IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
13
- TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
14
- PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
15
- OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
16
- EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
17
- PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
18
- PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
19
- LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
20
- NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
21
- SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
22
- */
23
-
24
- #ifndef UTHASH_H
25
- #define UTHASH_H
26
-
27
- #define UTHASH_VERSION 2.0.2
28
-
29
- #include <string.h> /* memcmp, memset, strlen */
30
- #include <stddef.h> /* ptrdiff_t */
31
- #include <stdlib.h> /* exit */
32
-
33
- /* These macros use decltype or the earlier __typeof GNU extension.
34
- As decltype is only available in newer compilers (VS2010 or gcc 4.3+
35
- when compiling c++ source) this code uses whatever method is needed
36
- or, for VS2008 where neither is available, uses casting workarounds. */
37
- #if !defined(DECLTYPE) && !defined(NO_DECLTYPE)
38
- #if defined(_MSC_VER) /* MS compiler */
39
- #if _MSC_VER >= 1600 && defined(__cplusplus) /* VS2010 or newer in C++ mode */
40
- #define DECLTYPE(x) (decltype(x))
41
- #else /* VS2008 or older (or VS2010 in C mode) */
42
- #define NO_DECLTYPE
43
- #endif
44
- #elif defined(__BORLANDC__) || defined(__ICCARM__) || defined(__LCC__) || defined(__WATCOMC__)
45
- #define NO_DECLTYPE
46
- #else /* GNU, Sun and other compilers */
47
- #define DECLTYPE(x) (__typeof(x))
48
- #endif
49
- #endif
50
-
51
- #ifdef NO_DECLTYPE
52
- #define DECLTYPE(x)
53
- #define DECLTYPE_ASSIGN(dst,src) \
54
- do { \
55
- char **_da_dst = (char**)(&(dst)); \
56
- *_da_dst = (char*)(src); \
57
- } while (0)
58
- #else
59
- #define DECLTYPE_ASSIGN(dst,src) \
60
- do { \
61
- (dst) = DECLTYPE(dst)(src); \
62
- } while (0)
63
- #endif
64
-
65
- /* a number of the hash function use uint32_t which isn't defined on Pre VS2010 */
66
- #if defined(_WIN32)
67
- #if defined(_MSC_VER) && _MSC_VER >= 1600
68
- #include <stdint.h>
69
- #elif defined(__WATCOMC__) || defined(__MINGW32__) || defined(__CYGWIN__)
70
- #include <stdint.h>
71
- #else
72
- typedef unsigned int uint32_t;
73
- typedef unsigned char uint8_t;
74
- #endif
75
- #elif defined(__GNUC__) && !defined(__VXWORKS__)
76
- #include <stdint.h>
77
- #else
78
- typedef unsigned int uint32_t;
79
- typedef unsigned char uint8_t;
80
- #endif
81
-
82
- #ifndef uthash_fatal
83
- #define uthash_fatal(msg) exit(-1) /* fatal error (out of memory,etc) */
84
- #endif
85
- #ifndef uthash_malloc
86
- #define uthash_malloc(sz) malloc(sz) /* malloc fcn */
87
- #endif
88
- #ifndef uthash_free
89
- #define uthash_free(ptr,sz) free(ptr) /* free fcn */
90
- #endif
91
- #ifndef uthash_bzero
92
- #define uthash_bzero(a,n) memset(a,'\0',n)
93
- #endif
94
- #ifndef uthash_memcmp
95
- #define uthash_memcmp(a,b,n) memcmp(a,b,n)
96
- #endif
97
- #ifndef uthash_strlen
98
- #define uthash_strlen(s) strlen(s)
99
- #endif
100
-
101
- #ifndef uthash_noexpand_fyi
102
- #define uthash_noexpand_fyi(tbl) /* can be defined to log noexpand */
103
- #endif
104
- #ifndef uthash_expand_fyi
105
- #define uthash_expand_fyi(tbl) /* can be defined to log expands */
106
- #endif
107
-
108
- /* initial number of buckets */
109
- #define HASH_INITIAL_NUM_BUCKETS 32U /* initial number of buckets */
110
- #define HASH_INITIAL_NUM_BUCKETS_LOG2 5U /* lg2 of initial number of buckets */
111
- #define HASH_BKT_CAPACITY_THRESH 10U /* expand when bucket count reaches */
112
-
113
- /* calculate the element whose hash handle address is hhp */
114
- #define ELMT_FROM_HH(tbl,hhp) ((void*)(((char*)(hhp)) - ((tbl)->hho)))
115
- /* calculate the hash handle from element address elp */
116
- #define HH_FROM_ELMT(tbl,elp) ((UT_hash_handle *)(((char*)(elp)) + ((tbl)->hho)))
117
-
118
- #define HASH_VALUE(keyptr,keylen,hashv) \
119
- do { \
120
- HASH_FCN(keyptr, keylen, hashv); \
121
- } while (0)
122
-
123
- #define HASH_FIND_BYHASHVALUE(hh,head,keyptr,keylen,hashval,out) \
124
- do { \
125
- (out) = NULL; \
126
- if (head) { \
127
- unsigned _hf_bkt; \
128
- HASH_TO_BKT(hashval, (head)->hh.tbl->num_buckets, _hf_bkt); \
129
- if (HASH_BLOOM_TEST((head)->hh.tbl, hashval) != 0) { \
130
- HASH_FIND_IN_BKT((head)->hh.tbl, hh, (head)->hh.tbl->buckets[ _hf_bkt ], keyptr, keylen, hashval, out); \
131
- } \
132
- } \
133
- } while (0)
134
-
135
- #define HASH_FIND(hh,head,keyptr,keylen,out) \
136
- do { \
137
- unsigned _hf_hashv; \
138
- HASH_VALUE(keyptr, keylen, _hf_hashv); \
139
- HASH_FIND_BYHASHVALUE(hh, head, keyptr, keylen, _hf_hashv, out); \
140
- } while (0)
141
-
142
- #ifdef HASH_BLOOM
143
- #define HASH_BLOOM_BITLEN (1UL << HASH_BLOOM)
144
- #define HASH_BLOOM_BYTELEN (HASH_BLOOM_BITLEN/8UL) + (((HASH_BLOOM_BITLEN%8UL)!=0UL) ? 1UL : 0UL)
145
- #define HASH_BLOOM_MAKE(tbl) \
146
- do { \
147
- (tbl)->bloom_nbits = HASH_BLOOM; \
148
- (tbl)->bloom_bv = (uint8_t*)uthash_malloc(HASH_BLOOM_BYTELEN); \
149
- if (!(tbl)->bloom_bv) { \
150
- uthash_fatal("out of memory"); \
151
- } \
152
- uthash_bzero((tbl)->bloom_bv, HASH_BLOOM_BYTELEN); \
153
- (tbl)->bloom_sig = HASH_BLOOM_SIGNATURE; \
154
- } while (0)
155
-
156
- #define HASH_BLOOM_FREE(tbl) \
157
- do { \
158
- uthash_free((tbl)->bloom_bv, HASH_BLOOM_BYTELEN); \
159
- } while (0)
160
-
161
- #define HASH_BLOOM_BITSET(bv,idx) (bv[(idx)/8U] |= (1U << ((idx)%8U)))
162
- #define HASH_BLOOM_BITTEST(bv,idx) (bv[(idx)/8U] & (1U << ((idx)%8U)))
163
-
164
- #define HASH_BLOOM_ADD(tbl,hashv) \
165
- HASH_BLOOM_BITSET((tbl)->bloom_bv, (hashv & (uint32_t)((1UL << (tbl)->bloom_nbits) - 1U)))
166
-
167
- #define HASH_BLOOM_TEST(tbl,hashv) \
168
- HASH_BLOOM_BITTEST((tbl)->bloom_bv, (hashv & (uint32_t)((1UL << (tbl)->bloom_nbits) - 1U)))
169
-
170
- #else
171
- #define HASH_BLOOM_MAKE(tbl)
172
- #define HASH_BLOOM_FREE(tbl)
173
- #define HASH_BLOOM_ADD(tbl,hashv)
174
- #define HASH_BLOOM_TEST(tbl,hashv) (1)
175
- #define HASH_BLOOM_BYTELEN 0U
176
- #endif
177
-
178
- #define HASH_MAKE_TABLE(hh,head) \
179
- do { \
180
- (head)->hh.tbl = (UT_hash_table*)uthash_malloc(sizeof(UT_hash_table)); \
181
- if (!(head)->hh.tbl) { \
182
- uthash_fatal("out of memory"); \
183
- } \
184
- uthash_bzero((head)->hh.tbl, sizeof(UT_hash_table)); \
185
- (head)->hh.tbl->tail = &((head)->hh); \
186
- (head)->hh.tbl->num_buckets = HASH_INITIAL_NUM_BUCKETS; \
187
- (head)->hh.tbl->log2_num_buckets = HASH_INITIAL_NUM_BUCKETS_LOG2; \
188
- (head)->hh.tbl->hho = (char*)(&(head)->hh) - (char*)(head); \
189
- (head)->hh.tbl->buckets = (UT_hash_bucket*)uthash_malloc( \
190
- HASH_INITIAL_NUM_BUCKETS * sizeof(struct UT_hash_bucket)); \
191
- if (!(head)->hh.tbl->buckets) { \
192
- uthash_fatal("out of memory"); \
193
- } \
194
- uthash_bzero((head)->hh.tbl->buckets, \
195
- HASH_INITIAL_NUM_BUCKETS * sizeof(struct UT_hash_bucket)); \
196
- HASH_BLOOM_MAKE((head)->hh.tbl); \
197
- (head)->hh.tbl->signature = HASH_SIGNATURE; \
198
- } while (0)
199
-
200
- #define HASH_REPLACE_BYHASHVALUE_INORDER(hh,head,fieldname,keylen_in,hashval,add,replaced,cmpfcn) \
201
- do { \
202
- (replaced) = NULL; \
203
- HASH_FIND_BYHASHVALUE(hh, head, &((add)->fieldname), keylen_in, hashval, replaced); \
204
- if (replaced) { \
205
- HASH_DELETE(hh, head, replaced); \
206
- } \
207
- HASH_ADD_KEYPTR_BYHASHVALUE_INORDER(hh, head, &((add)->fieldname), keylen_in, hashval, add, cmpfcn); \
208
- } while (0)
209
-
210
- #define HASH_REPLACE_BYHASHVALUE(hh,head,fieldname,keylen_in,hashval,add,replaced) \
211
- do { \
212
- (replaced) = NULL; \
213
- HASH_FIND_BYHASHVALUE(hh, head, &((add)->fieldname), keylen_in, hashval, replaced); \
214
- if (replaced) { \
215
- HASH_DELETE(hh, head, replaced); \
216
- } \
217
- HASH_ADD_KEYPTR_BYHASHVALUE(hh, head, &((add)->fieldname), keylen_in, hashval, add); \
218
- } while (0)
219
-
220
- #define HASH_REPLACE(hh,head,fieldname,keylen_in,add,replaced) \
221
- do { \
222
- unsigned _hr_hashv; \
223
- HASH_VALUE(&((add)->fieldname), keylen_in, _hr_hashv); \
224
- HASH_REPLACE_BYHASHVALUE(hh, head, fieldname, keylen_in, _hr_hashv, add, replaced); \
225
- } while (0)
226
-
227
- #define HASH_REPLACE_INORDER(hh,head,fieldname,keylen_in,add,replaced,cmpfcn) \
228
- do { \
229
- unsigned _hr_hashv; \
230
- HASH_VALUE(&((add)->fieldname), keylen_in, _hr_hashv); \
231
- HASH_REPLACE_BYHASHVALUE_INORDER(hh, head, fieldname, keylen_in, _hr_hashv, add, replaced, cmpfcn); \
232
- } while (0)
233
-
234
- #define HASH_APPEND_LIST(hh, head, add) \
235
- do { \
236
- (add)->hh.next = NULL; \
237
- (add)->hh.prev = ELMT_FROM_HH((head)->hh.tbl, (head)->hh.tbl->tail); \
238
- (head)->hh.tbl->tail->next = (add); \
239
- (head)->hh.tbl->tail = &((add)->hh); \
240
- } while (0)
241
-
242
- #define HASH_AKBI_INNER_LOOP(hh,head,add,cmpfcn) \
243
- do { \
244
- do { \
245
- if (cmpfcn(DECLTYPE(head)(_hs_iter), add) > 0) { \
246
- break; \
247
- } \
248
- } while ((_hs_iter = HH_FROM_ELMT((head)->hh.tbl, _hs_iter)->next)); \
249
- } while (0)
250
-
251
- #ifdef NO_DECLTYPE
252
- #undef HASH_AKBI_INNER_LOOP
253
- #define HASH_AKBI_INNER_LOOP(hh,head,add,cmpfcn) \
254
- do { \
255
- char *_hs_saved_head = (char*)(head); \
256
- do { \
257
- DECLTYPE_ASSIGN(head, _hs_iter); \
258
- if (cmpfcn(head, add) > 0) { \
259
- DECLTYPE_ASSIGN(head, _hs_saved_head); \
260
- break; \
261
- } \
262
- DECLTYPE_ASSIGN(head, _hs_saved_head); \
263
- } while ((_hs_iter = HH_FROM_ELMT((head)->hh.tbl, _hs_iter)->next)); \
264
- } while (0)
265
- #endif
266
-
267
- #define HASH_ADD_KEYPTR_BYHASHVALUE_INORDER(hh,head,keyptr,keylen_in,hashval,add,cmpfcn) \
268
- do { \
269
- unsigned _ha_bkt; \
270
- (add)->hh.hashv = (hashval); \
271
- (add)->hh.key = (char*) (keyptr); \
272
- (add)->hh.keylen = (unsigned) (keylen_in); \
273
- if (!(head)) { \
274
- (add)->hh.next = NULL; \
275
- (add)->hh.prev = NULL; \
276
- (head) = (add); \
277
- HASH_MAKE_TABLE(hh, head); \
278
- } else { \
279
- void *_hs_iter = (head); \
280
- (add)->hh.tbl = (head)->hh.tbl; \
281
- HASH_AKBI_INNER_LOOP(hh, head, add, cmpfcn); \
282
- if (_hs_iter) { \
283
- (add)->hh.next = _hs_iter; \
284
- if (((add)->hh.prev = HH_FROM_ELMT((head)->hh.tbl, _hs_iter)->prev)) { \
285
- HH_FROM_ELMT((head)->hh.tbl, (add)->hh.prev)->next = (add); \
286
- } else { \
287
- (head) = (add); \
288
- } \
289
- HH_FROM_ELMT((head)->hh.tbl, _hs_iter)->prev = (add); \
290
- } else { \
291
- HASH_APPEND_LIST(hh, head, add); \
292
- } \
293
- } \
294
- (head)->hh.tbl->num_items++; \
295
- HASH_TO_BKT(hashval, (head)->hh.tbl->num_buckets, _ha_bkt); \
296
- HASH_ADD_TO_BKT((head)->hh.tbl->buckets[_ha_bkt], &(add)->hh); \
297
- HASH_BLOOM_ADD((head)->hh.tbl, hashval); \
298
- HASH_EMIT_KEY(hh, head, keyptr, keylen_in); \
299
- HASH_FSCK(hh, head, "HASH_ADD_KEYPTR_BYHASHVALUE_INORDER"); \
300
- } while (0)
301
-
302
- #define HASH_ADD_KEYPTR_INORDER(hh,head,keyptr,keylen_in,add,cmpfcn) \
303
- do { \
304
- unsigned _hs_hashv; \
305
- HASH_VALUE(keyptr, keylen_in, _hs_hashv); \
306
- HASH_ADD_KEYPTR_BYHASHVALUE_INORDER(hh, head, keyptr, keylen_in, _hs_hashv, add, cmpfcn); \
307
- } while (0)
308
-
309
- #define HASH_ADD_BYHASHVALUE_INORDER(hh,head,fieldname,keylen_in,hashval,add,cmpfcn) \
310
- HASH_ADD_KEYPTR_BYHASHVALUE_INORDER(hh, head, &((add)->fieldname), keylen_in, hashval, add, cmpfcn)
311
-
312
- #define HASH_ADD_INORDER(hh,head,fieldname,keylen_in,add,cmpfcn) \
313
- HASH_ADD_KEYPTR_INORDER(hh, head, &((add)->fieldname), keylen_in, add, cmpfcn)
314
-
315
- #define HASH_ADD_KEYPTR_BYHASHVALUE(hh,head,keyptr,keylen_in,hashval,add) \
316
- do { \
317
- unsigned _ha_bkt; \
318
- (add)->hh.hashv = (hashval); \
319
- (add)->hh.key = (char*) (keyptr); \
320
- (add)->hh.keylen = (unsigned) (keylen_in); \
321
- if (!(head)) { \
322
- (add)->hh.next = NULL; \
323
- (add)->hh.prev = NULL; \
324
- (head) = (add); \
325
- HASH_MAKE_TABLE(hh, head); \
326
- } else { \
327
- (add)->hh.tbl = (head)->hh.tbl; \
328
- HASH_APPEND_LIST(hh, head, add); \
329
- } \
330
- (head)->hh.tbl->num_items++; \
331
- HASH_TO_BKT(hashval, (head)->hh.tbl->num_buckets, _ha_bkt); \
332
- HASH_ADD_TO_BKT((head)->hh.tbl->buckets[_ha_bkt], &(add)->hh); \
333
- HASH_BLOOM_ADD((head)->hh.tbl, hashval); \
334
- HASH_EMIT_KEY(hh, head, keyptr, keylen_in); \
335
- HASH_FSCK(hh, head, "HASH_ADD_KEYPTR_BYHASHVALUE"); \
336
- } while (0)
337
-
338
- #define HASH_ADD_KEYPTR(hh,head,keyptr,keylen_in,add) \
339
- do { \
340
- unsigned _ha_hashv; \
341
- HASH_VALUE(keyptr, keylen_in, _ha_hashv); \
342
- HASH_ADD_KEYPTR_BYHASHVALUE(hh, head, keyptr, keylen_in, _ha_hashv, add); \
343
- } while (0)
344
-
345
- #define HASH_ADD_BYHASHVALUE(hh,head,fieldname,keylen_in,hashval,add) \
346
- HASH_ADD_KEYPTR_BYHASHVALUE(hh, head, &((add)->fieldname), keylen_in, hashval, add)
347
-
348
- #define HASH_ADD(hh,head,fieldname,keylen_in,add) \
349
- HASH_ADD_KEYPTR(hh, head, &((add)->fieldname), keylen_in, add)
350
-
351
- #define HASH_TO_BKT(hashv,num_bkts,bkt) \
352
- do { \
353
- bkt = ((hashv) & ((num_bkts) - 1U)); \
354
- } while (0)
355
-
356
- /* delete "delptr" from the hash table.
357
- * "the usual" patch-up process for the app-order doubly-linked-list.
358
- * The use of _hd_hh_del below deserves special explanation.
359
- * These used to be expressed using (delptr) but that led to a bug
360
- * if someone used the same symbol for the head and deletee, like
361
- * HASH_DELETE(hh,users,users);
362
- * We want that to work, but by changing the head (users) below
363
- * we were forfeiting our ability to further refer to the deletee (users)
364
- * in the patch-up process. Solution: use scratch space to
365
- * copy the deletee pointer, then the latter references are via that
366
- * scratch pointer rather than through the repointed (users) symbol.
367
- */
368
- #define HASH_DELETE(hh,head,delptr) \
369
- HASH_DELETE_HH(hh, head, &(delptr)->hh)
370
-
371
- #define HASH_DELETE_HH(hh,head,delptrhh) \
372
- do { \
373
- struct UT_hash_handle *_hd_hh_del = (delptrhh); \
374
- if ((_hd_hh_del->prev == NULL) && (_hd_hh_del->next == NULL)) { \
375
- HASH_BLOOM_FREE((head)->hh.tbl); \
376
- uthash_free((head)->hh.tbl->buckets, \
377
- (head)->hh.tbl->num_buckets * sizeof(struct UT_hash_bucket)); \
378
- uthash_free((head)->hh.tbl, sizeof(UT_hash_table)); \
379
- (head) = NULL; \
380
- } else { \
381
- unsigned _hd_bkt; \
382
- if (_hd_hh_del == (head)->hh.tbl->tail) { \
383
- (head)->hh.tbl->tail = HH_FROM_ELMT((head)->hh.tbl, _hd_hh_del->prev); \
384
- } \
385
- if (_hd_hh_del->prev != NULL) { \
386
- HH_FROM_ELMT((head)->hh.tbl, _hd_hh_del->prev)->next = _hd_hh_del->next; \
387
- } else { \
388
- DECLTYPE_ASSIGN(head, _hd_hh_del->next); \
389
- } \
390
- if (_hd_hh_del->next != NULL) { \
391
- HH_FROM_ELMT((head)->hh.tbl, _hd_hh_del->next)->prev = _hd_hh_del->prev; \
392
- } \
393
- HASH_TO_BKT(_hd_hh_del->hashv, (head)->hh.tbl->num_buckets, _hd_bkt); \
394
- HASH_DEL_IN_BKT((head)->hh.tbl->buckets[_hd_bkt], _hd_hh_del); \
395
- (head)->hh.tbl->num_items--; \
396
- } \
397
- HASH_FSCK(hh, head, "HASH_DELETE"); \
398
- } while (0)
399
-
400
-
401
- /* convenience forms of HASH_FIND/HASH_ADD/HASH_DEL */
402
- #define HASH_FIND_STR(head,findstr,out) \
403
- HASH_FIND(hh,head,findstr,(unsigned)uthash_strlen(findstr),out)
404
- #define HASH_ADD_STR(head,strfield,add) \
405
- HASH_ADD(hh,head,strfield[0],(unsigned)uthash_strlen(add->strfield),add)
406
- #define HASH_REPLACE_STR(head,strfield,add,replaced) \
407
- HASH_REPLACE(hh,head,strfield[0],(unsigned)uthash_strlen(add->strfield),add,replaced)
408
- #define HASH_FIND_INT(head,findint,out) \
409
- HASH_FIND(hh,head,findint,sizeof(int),out)
410
- #define HASH_ADD_INT(head,intfield,add) \
411
- HASH_ADD(hh,head,intfield,sizeof(int),add)
412
- #define HASH_REPLACE_INT(head,intfield,add,replaced) \
413
- HASH_REPLACE(hh,head,intfield,sizeof(int),add,replaced)
414
- #define HASH_FIND_PTR(head,findptr,out) \
415
- HASH_FIND(hh,head,findptr,sizeof(void *),out)
416
- #define HASH_ADD_PTR(head,ptrfield,add) \
417
- HASH_ADD(hh,head,ptrfield,sizeof(void *),add)
418
- #define HASH_REPLACE_PTR(head,ptrfield,add,replaced) \
419
- HASH_REPLACE(hh,head,ptrfield,sizeof(void *),add,replaced)
420
- #define HASH_DEL(head,delptr) \
421
- HASH_DELETE(hh,head,delptr)
422
-
423
- /* HASH_FSCK checks hash integrity on every add/delete when HASH_DEBUG is defined.
424
- * This is for uthash developer only; it compiles away if HASH_DEBUG isn't defined.
425
- */
426
- #ifdef HASH_DEBUG
427
- #define HASH_OOPS(...) do { fprintf(stderr,__VA_ARGS__); exit(-1); } while (0)
428
- #define HASH_FSCK(hh,head,where) \
429
- do { \
430
- struct UT_hash_handle *_thh; \
431
- if (head) { \
432
- unsigned _bkt_i; \
433
- unsigned _count = 0; \
434
- char *_prev; \
435
- for (_bkt_i = 0; _bkt_i < (head)->hh.tbl->num_buckets; ++_bkt_i) { \
436
- unsigned _bkt_count = 0; \
437
- _thh = (head)->hh.tbl->buckets[_bkt_i].hh_head; \
438
- _prev = NULL; \
439
- while (_thh) { \
440
- if (_prev != (char*)(_thh->hh_prev)) { \
441
- HASH_OOPS("%s: invalid hh_prev %p, actual %p\n", \
442
- (where), (void*)_thh->hh_prev, (void*)_prev); \
443
- } \
444
- _bkt_count++; \
445
- _prev = (char*)(_thh); \
446
- _thh = _thh->hh_next; \
447
- } \
448
- _count += _bkt_count; \
449
- if ((head)->hh.tbl->buckets[_bkt_i].count != _bkt_count) { \
450
- HASH_OOPS("%s: invalid bucket count %u, actual %u\n", \
451
- (where), (head)->hh.tbl->buckets[_bkt_i].count, _bkt_count); \
452
- } \
453
- } \
454
- if (_count != (head)->hh.tbl->num_items) { \
455
- HASH_OOPS("%s: invalid hh item count %u, actual %u\n", \
456
- (where), (head)->hh.tbl->num_items, _count); \
457
- } \
458
- _count = 0; \
459
- _prev = NULL; \
460
- _thh = &(head)->hh; \
461
- while (_thh) { \
462
- _count++; \
463
- if (_prev != (char*)_thh->prev) { \
464
- HASH_OOPS("%s: invalid prev %p, actual %p\n", \
465
- (where), (void*)_thh->prev, (void*)_prev); \
466
- } \
467
- _prev = (char*)ELMT_FROM_HH((head)->hh.tbl, _thh); \
468
- _thh = (_thh->next ? HH_FROM_ELMT((head)->hh.tbl, _thh->next) : NULL); \
469
- } \
470
- if (_count != (head)->hh.tbl->num_items) { \
471
- HASH_OOPS("%s: invalid app item count %u, actual %u\n", \
472
- (where), (head)->hh.tbl->num_items, _count); \
473
- } \
474
- } \
475
- } while (0)
476
- #else
477
- #define HASH_FSCK(hh,head,where)
478
- #endif
479
-
480
- /* When compiled with -DHASH_EMIT_KEYS, length-prefixed keys are emitted to
481
- * the descriptor to which this macro is defined for tuning the hash function.
482
- * The app can #include <unistd.h> to get the prototype for write(2). */
483
- #ifdef HASH_EMIT_KEYS
484
- #define HASH_EMIT_KEY(hh,head,keyptr,fieldlen) \
485
- do { \
486
- unsigned _klen = fieldlen; \
487
- write(HASH_EMIT_KEYS, &_klen, sizeof(_klen)); \
488
- write(HASH_EMIT_KEYS, keyptr, (unsigned long)fieldlen); \
489
- } while (0)
490
- #else
491
- #define HASH_EMIT_KEY(hh,head,keyptr,fieldlen)
492
- #endif
493
-
494
- /* default to Jenkin's hash unless overridden e.g. DHASH_FUNCTION=HASH_SAX */
495
- #ifdef HASH_FUNCTION
496
- #define HASH_FCN HASH_FUNCTION
497
- #else
498
- #define HASH_FCN HASH_JEN
499
- #endif
500
-
501
- /* The Bernstein hash function, used in Perl prior to v5.6. Note (x<<5+x)=x*33. */
502
- #define HASH_BER(key,keylen,hashv) \
503
- do { \
504
- unsigned _hb_keylen = (unsigned)keylen; \
505
- const unsigned char *_hb_key = (const unsigned char*)(key); \
506
- (hashv) = 0; \
507
- while (_hb_keylen-- != 0U) { \
508
- (hashv) = (((hashv) << 5) + (hashv)) + *_hb_key++; \
509
- } \
510
- } while (0)
511
-
512
-
513
- /* SAX/FNV/OAT/JEN hash functions are macro variants of those listed at
514
- * http://eternallyconfuzzled.com/tuts/algorithms/jsw_tut_hashing.aspx */
515
- #define HASH_SAX(key,keylen,hashv) \
516
- do { \
517
- unsigned _sx_i; \
518
- const unsigned char *_hs_key = (const unsigned char*)(key); \
519
- hashv = 0; \
520
- for (_sx_i=0; _sx_i < keylen; _sx_i++) { \
521
- hashv ^= (hashv << 5) + (hashv >> 2) + _hs_key[_sx_i]; \
522
- } \
523
- } while (0)
524
- /* FNV-1a variation */
525
- #define HASH_FNV(key,keylen,hashv) \
526
- do { \
527
- unsigned _fn_i; \
528
- const unsigned char *_hf_key = (const unsigned char*)(key); \
529
- (hashv) = 2166136261U; \
530
- for (_fn_i=0; _fn_i < keylen; _fn_i++) { \
531
- hashv = hashv ^ _hf_key[_fn_i]; \
532
- hashv = hashv * 16777619U; \
533
- } \
534
- } while (0)
535
-
536
- #define HASH_OAT(key,keylen,hashv) \
537
- do { \
538
- unsigned _ho_i; \
539
- const unsigned char *_ho_key=(const unsigned char*)(key); \
540
- hashv = 0; \
541
- for(_ho_i=0; _ho_i < keylen; _ho_i++) { \
542
- hashv += _ho_key[_ho_i]; \
543
- hashv += (hashv << 10); \
544
- hashv ^= (hashv >> 6); \
545
- } \
546
- hashv += (hashv << 3); \
547
- hashv ^= (hashv >> 11); \
548
- hashv += (hashv << 15); \
549
- } while (0)
550
-
551
- #define HASH_JEN_MIX(a,b,c) \
552
- do { \
553
- a -= b; a -= c; a ^= ( c >> 13 ); \
554
- b -= c; b -= a; b ^= ( a << 8 ); \
555
- c -= a; c -= b; c ^= ( b >> 13 ); \
556
- a -= b; a -= c; a ^= ( c >> 12 ); \
557
- b -= c; b -= a; b ^= ( a << 16 ); \
558
- c -= a; c -= b; c ^= ( b >> 5 ); \
559
- a -= b; a -= c; a ^= ( c >> 3 ); \
560
- b -= c; b -= a; b ^= ( a << 10 ); \
561
- c -= a; c -= b; c ^= ( b >> 15 ); \
562
- } while (0)
563
-
564
- #define HASH_JEN(key,keylen,hashv) \
565
- do { \
566
- unsigned _hj_i,_hj_j,_hj_k; \
567
- unsigned const char *_hj_key=(unsigned const char*)(key); \
568
- hashv = 0xfeedbeefu; \
569
- _hj_i = _hj_j = 0x9e3779b9u; \
570
- _hj_k = (unsigned)(keylen); \
571
- while (_hj_k >= 12U) { \
572
- _hj_i += (_hj_key[0] + ( (unsigned)_hj_key[1] << 8 ) \
573
- + ( (unsigned)_hj_key[2] << 16 ) \
574
- + ( (unsigned)_hj_key[3] << 24 ) ); \
575
- _hj_j += (_hj_key[4] + ( (unsigned)_hj_key[5] << 8 ) \
576
- + ( (unsigned)_hj_key[6] << 16 ) \
577
- + ( (unsigned)_hj_key[7] << 24 ) ); \
578
- hashv += (_hj_key[8] + ( (unsigned)_hj_key[9] << 8 ) \
579
- + ( (unsigned)_hj_key[10] << 16 ) \
580
- + ( (unsigned)_hj_key[11] << 24 ) ); \
581
- \
582
- HASH_JEN_MIX(_hj_i, _hj_j, hashv); \
583
- \
584
- _hj_key += 12; \
585
- _hj_k -= 12U; \
586
- } \
587
- hashv += (unsigned)(keylen); \
588
- switch ( _hj_k ) { \
589
- case 11: hashv += ( (unsigned)_hj_key[10] << 24 ); /* FALLTHROUGH */ \
590
- case 10: hashv += ( (unsigned)_hj_key[9] << 16 ); /* FALLTHROUGH */ \
591
- case 9: hashv += ( (unsigned)_hj_key[8] << 8 ); /* FALLTHROUGH */ \
592
- case 8: _hj_j += ( (unsigned)_hj_key[7] << 24 ); /* FALLTHROUGH */ \
593
- case 7: _hj_j += ( (unsigned)_hj_key[6] << 16 ); /* FALLTHROUGH */ \
594
- case 6: _hj_j += ( (unsigned)_hj_key[5] << 8 ); /* FALLTHROUGH */ \
595
- case 5: _hj_j += _hj_key[4]; /* FALLTHROUGH */ \
596
- case 4: _hj_i += ( (unsigned)_hj_key[3] << 24 ); /* FALLTHROUGH */ \
597
- case 3: _hj_i += ( (unsigned)_hj_key[2] << 16 ); /* FALLTHROUGH */ \
598
- case 2: _hj_i += ( (unsigned)_hj_key[1] << 8 ); /* FALLTHROUGH */ \
599
- case 1: _hj_i += _hj_key[0]; \
600
- } \
601
- HASH_JEN_MIX(_hj_i, _hj_j, hashv); \
602
- } while (0)
603
-
604
- /* The Paul Hsieh hash function */
605
- #undef get16bits
606
- #if (defined(__GNUC__) && defined(__i386__)) || defined(__WATCOMC__) \
607
- || defined(_MSC_VER) || defined (__BORLANDC__) || defined (__TURBOC__)
608
- #define get16bits(d) (*((const uint16_t *) (d)))
609
- #endif
610
-
611
- #if !defined (get16bits)
612
- #define get16bits(d) ((((uint32_t)(((const uint8_t *)(d))[1])) << 8) \
613
- +(uint32_t)(((const uint8_t *)(d))[0]) )
614
- #endif
615
- #define HASH_SFH(key,keylen,hashv) \
616
- do { \
617
- unsigned const char *_sfh_key=(unsigned const char*)(key); \
618
- uint32_t _sfh_tmp, _sfh_len = (uint32_t)keylen; \
619
- \
620
- unsigned _sfh_rem = _sfh_len & 3U; \
621
- _sfh_len >>= 2; \
622
- hashv = 0xcafebabeu; \
623
- \
624
- /* Main loop */ \
625
- for (;_sfh_len > 0U; _sfh_len--) { \
626
- hashv += get16bits (_sfh_key); \
627
- _sfh_tmp = ((uint32_t)(get16bits (_sfh_key+2)) << 11) ^ hashv; \
628
- hashv = (hashv << 16) ^ _sfh_tmp; \
629
- _sfh_key += 2U*sizeof (uint16_t); \
630
- hashv += hashv >> 11; \
631
- } \
632
- \
633
- /* Handle end cases */ \
634
- switch (_sfh_rem) { \
635
- case 3: hashv += get16bits (_sfh_key); \
636
- hashv ^= hashv << 16; \
637
- hashv ^= (uint32_t)(_sfh_key[sizeof (uint16_t)]) << 18; \
638
- hashv += hashv >> 11; \
639
- break; \
640
- case 2: hashv += get16bits (_sfh_key); \
641
- hashv ^= hashv << 11; \
642
- hashv += hashv >> 17; \
643
- break; \
644
- case 1: hashv += *_sfh_key; \
645
- hashv ^= hashv << 10; \
646
- hashv += hashv >> 1; \
647
- } \
648
- \
649
- /* Force "avalanching" of final 127 bits */ \
650
- hashv ^= hashv << 3; \
651
- hashv += hashv >> 5; \
652
- hashv ^= hashv << 4; \
653
- hashv += hashv >> 17; \
654
- hashv ^= hashv << 25; \
655
- hashv += hashv >> 6; \
656
- } while (0)
657
-
658
- #ifdef HASH_USING_NO_STRICT_ALIASING
659
- /* The MurmurHash exploits some CPU's (x86,x86_64) tolerance for unaligned reads.
660
- * For other types of CPU's (e.g. Sparc) an unaligned read causes a bus error.
661
- * MurmurHash uses the faster approach only on CPU's where we know it's safe.
662
- *
663
- * Note the preprocessor built-in defines can be emitted using:
664
- *
665
- * gcc -m64 -dM -E - < /dev/null (on gcc)
666
- * cc -## a.c (where a.c is a simple test file) (Sun Studio)
667
- */
668
- #if (defined(__i386__) || defined(__x86_64__) || defined(_M_IX86))
669
- #define MUR_GETBLOCK(p,i) p[i]
670
- #else /* non intel */
671
- #define MUR_PLUS0_ALIGNED(p) (((unsigned long)p & 3UL) == 0UL)
672
- #define MUR_PLUS1_ALIGNED(p) (((unsigned long)p & 3UL) == 1UL)
673
- #define MUR_PLUS2_ALIGNED(p) (((unsigned long)p & 3UL) == 2UL)
674
- #define MUR_PLUS3_ALIGNED(p) (((unsigned long)p & 3UL) == 3UL)
675
- #define WP(p) ((uint32_t*)((unsigned long)(p) & ~3UL))
676
- #if (defined(__BIG_ENDIAN__) || defined(SPARC) || defined(__ppc__) || defined(__ppc64__))
677
- #define MUR_THREE_ONE(p) ((((*WP(p))&0x00ffffff) << 8) | (((*(WP(p)+1))&0xff000000) >> 24))
678
- #define MUR_TWO_TWO(p) ((((*WP(p))&0x0000ffff) <<16) | (((*(WP(p)+1))&0xffff0000) >> 16))
679
- #define MUR_ONE_THREE(p) ((((*WP(p))&0x000000ff) <<24) | (((*(WP(p)+1))&0xffffff00) >> 8))
680
- #else /* assume little endian non-intel */
681
- #define MUR_THREE_ONE(p) ((((*WP(p))&0xffffff00) >> 8) | (((*(WP(p)+1))&0x000000ff) << 24))
682
- #define MUR_TWO_TWO(p) ((((*WP(p))&0xffff0000) >>16) | (((*(WP(p)+1))&0x0000ffff) << 16))
683
- #define MUR_ONE_THREE(p) ((((*WP(p))&0xff000000) >>24) | (((*(WP(p)+1))&0x00ffffff) << 8))
684
- #endif
685
- #define MUR_GETBLOCK(p,i) (MUR_PLUS0_ALIGNED(p) ? ((p)[i]) : \
686
- (MUR_PLUS1_ALIGNED(p) ? MUR_THREE_ONE(p) : \
687
- (MUR_PLUS2_ALIGNED(p) ? MUR_TWO_TWO(p) : \
688
- MUR_ONE_THREE(p))))
689
- #endif
690
- #define MUR_ROTL32(x,r) (((x) << (r)) | ((x) >> (32 - (r))))
691
- #define MUR_FMIX(_h) \
692
- do { \
693
- _h ^= _h >> 16; \
694
- _h *= 0x85ebca6bu; \
695
- _h ^= _h >> 13; \
696
- _h *= 0xc2b2ae35u; \
697
- _h ^= _h >> 16; \
698
- } while (0)
699
-
700
- #define HASH_MUR(key,keylen,hashv) \
701
- do { \
702
- const uint8_t *_mur_data = (const uint8_t*)(key); \
703
- const int _mur_nblocks = (int)(keylen) / 4; \
704
- uint32_t _mur_h1 = 0xf88D5353u; \
705
- uint32_t _mur_c1 = 0xcc9e2d51u; \
706
- uint32_t _mur_c2 = 0x1b873593u; \
707
- uint32_t _mur_k1 = 0; \
708
- const uint8_t *_mur_tail; \
709
- const uint32_t *_mur_blocks = (const uint32_t*)(_mur_data+(_mur_nblocks*4)); \
710
- int _mur_i; \
711
- for (_mur_i = -_mur_nblocks; _mur_i != 0; _mur_i++) { \
712
- _mur_k1 = MUR_GETBLOCK(_mur_blocks,_mur_i); \
713
- _mur_k1 *= _mur_c1; \
714
- _mur_k1 = MUR_ROTL32(_mur_k1,15); \
715
- _mur_k1 *= _mur_c2; \
716
- \
717
- _mur_h1 ^= _mur_k1; \
718
- _mur_h1 = MUR_ROTL32(_mur_h1,13); \
719
- _mur_h1 = (_mur_h1*5U) + 0xe6546b64u; \
720
- } \
721
- _mur_tail = (const uint8_t*)(_mur_data + (_mur_nblocks*4)); \
722
- _mur_k1=0; \
723
- switch ((keylen) & 3U) { \
724
- case 0: break; \
725
- case 3: _mur_k1 ^= (uint32_t)_mur_tail[2] << 16; /* FALLTHROUGH */ \
726
- case 2: _mur_k1 ^= (uint32_t)_mur_tail[1] << 8; /* FALLTHROUGH */ \
727
- case 1: _mur_k1 ^= (uint32_t)_mur_tail[0]; \
728
- _mur_k1 *= _mur_c1; \
729
- _mur_k1 = MUR_ROTL32(_mur_k1,15); \
730
- _mur_k1 *= _mur_c2; \
731
- _mur_h1 ^= _mur_k1; \
732
- } \
733
- _mur_h1 ^= (uint32_t)(keylen); \
734
- MUR_FMIX(_mur_h1); \
735
- hashv = _mur_h1; \
736
- } while (0)
737
- #endif /* HASH_USING_NO_STRICT_ALIASING */
738
-
739
- /* iterate over items in a known bucket to find desired item */
740
- #define HASH_FIND_IN_BKT(tbl,hh,head,keyptr,keylen_in,hashval,out) \
741
- do { \
742
- if ((head).hh_head != NULL) { \
743
- DECLTYPE_ASSIGN(out, ELMT_FROM_HH(tbl, (head).hh_head)); \
744
- } else { \
745
- (out) = NULL; \
746
- } \
747
- while ((out) != NULL) { \
748
- if ((out)->hh.hashv == (hashval) && (out)->hh.keylen == (keylen_in)) { \
749
- if (uthash_memcmp((out)->hh.key, keyptr, keylen_in) == 0) { \
750
- break; \
751
- } \
752
- } \
753
- if ((out)->hh.hh_next != NULL) { \
754
- DECLTYPE_ASSIGN(out, ELMT_FROM_HH(tbl, (out)->hh.hh_next)); \
755
- } else { \
756
- (out) = NULL; \
757
- } \
758
- } \
759
- } while (0)
760
-
761
- /* add an item to a bucket */
762
- #define HASH_ADD_TO_BKT(head,addhh) \
763
- do { \
764
- UT_hash_bucket *_ha_head = &(head); \
765
- _ha_head->count++; \
766
- (addhh)->hh_next = _ha_head->hh_head; \
767
- (addhh)->hh_prev = NULL; \
768
- if (_ha_head->hh_head != NULL) { \
769
- _ha_head->hh_head->hh_prev = (addhh); \
770
- } \
771
- _ha_head->hh_head = (addhh); \
772
- if ((_ha_head->count >= ((_ha_head->expand_mult + 1U) * HASH_BKT_CAPACITY_THRESH)) \
773
- && !(addhh)->tbl->noexpand) { \
774
- HASH_EXPAND_BUCKETS((addhh)->tbl); \
775
- } \
776
- } while (0)
777
-
778
- /* remove an item from a given bucket */
779
- #define HASH_DEL_IN_BKT(head,delhh) \
780
- do { \
781
- UT_hash_bucket *_hd_head = &(head); \
782
- _hd_head->count--; \
783
- if (_hd_head->hh_head == (delhh)) { \
784
- _hd_head->hh_head = (delhh)->hh_next; \
785
- } \
786
- if ((delhh)->hh_prev) { \
787
- (delhh)->hh_prev->hh_next = (delhh)->hh_next; \
788
- } \
789
- if ((delhh)->hh_next) { \
790
- (delhh)->hh_next->hh_prev = (delhh)->hh_prev; \
791
- } \
792
- } while (0)
793
-
794
- /* Bucket expansion has the effect of doubling the number of buckets
795
- * and redistributing the items into the new buckets. Ideally the
796
- * items will distribute more or less evenly into the new buckets
797
- * (the extent to which this is true is a measure of the quality of
798
- * the hash function as it applies to the key domain).
799
- *
800
- * With the items distributed into more buckets, the chain length
801
- * (item count) in each bucket is reduced. Thus by expanding buckets
802
- * the hash keeps a bound on the chain length. This bounded chain
803
- * length is the essence of how a hash provides constant time lookup.
804
- *
805
- * The calculation of tbl->ideal_chain_maxlen below deserves some
806
- * explanation. First, keep in mind that we're calculating the ideal
807
- * maximum chain length based on the *new* (doubled) bucket count.
808
- * In fractions this is just n/b (n=number of items,b=new num buckets).
809
- * Since the ideal chain length is an integer, we want to calculate
810
- * ceil(n/b). We don't depend on floating point arithmetic in this
811
- * hash, so to calculate ceil(n/b) with integers we could write
812
- *
813
- * ceil(n/b) = (n/b) + ((n%b)?1:0)
814
- *
815
- * and in fact a previous version of this hash did just that.
816
- * But now we have improved things a bit by recognizing that b is
817
- * always a power of two. We keep its base 2 log handy (call it lb),
818
- * so now we can write this with a bit shift and logical AND:
819
- *
820
- * ceil(n/b) = (n>>lb) + ( (n & (b-1)) ? 1:0)
821
- *
822
- */
823
- #define HASH_EXPAND_BUCKETS(tbl) \
824
- do { \
825
- unsigned _he_bkt; \
826
- unsigned _he_bkt_i; \
827
- struct UT_hash_handle *_he_thh, *_he_hh_nxt; \
828
- UT_hash_bucket *_he_new_buckets, *_he_newbkt; \
829
- _he_new_buckets = (UT_hash_bucket*)uthash_malloc( \
830
- 2UL * (tbl)->num_buckets * sizeof(struct UT_hash_bucket)); \
831
- if (!_he_new_buckets) { \
832
- uthash_fatal("out of memory"); \
833
- } \
834
- uthash_bzero(_he_new_buckets, \
835
- 2UL * (tbl)->num_buckets * sizeof(struct UT_hash_bucket)); \
836
- (tbl)->ideal_chain_maxlen = \
837
- ((tbl)->num_items >> ((tbl)->log2_num_buckets+1U)) + \
838
- ((((tbl)->num_items & (((tbl)->num_buckets*2U)-1U)) != 0U) ? 1U : 0U); \
839
- (tbl)->nonideal_items = 0; \
840
- for (_he_bkt_i = 0; _he_bkt_i < (tbl)->num_buckets; _he_bkt_i++) { \
841
- _he_thh = (tbl)->buckets[ _he_bkt_i ].hh_head; \
842
- while (_he_thh != NULL) { \
843
- _he_hh_nxt = _he_thh->hh_next; \
844
- HASH_TO_BKT(_he_thh->hashv, (tbl)->num_buckets * 2U, _he_bkt); \
845
- _he_newbkt = &(_he_new_buckets[_he_bkt]); \
846
- if (++(_he_newbkt->count) > (tbl)->ideal_chain_maxlen) { \
847
- (tbl)->nonideal_items++; \
848
- _he_newbkt->expand_mult = _he_newbkt->count / (tbl)->ideal_chain_maxlen; \
849
- } \
850
- _he_thh->hh_prev = NULL; \
851
- _he_thh->hh_next = _he_newbkt->hh_head; \
852
- if (_he_newbkt->hh_head != NULL) { \
853
- _he_newbkt->hh_head->hh_prev = _he_thh; \
854
- } \
855
- _he_newbkt->hh_head = _he_thh; \
856
- _he_thh = _he_hh_nxt; \
857
- } \
858
- } \
859
- uthash_free((tbl)->buckets, (tbl)->num_buckets * sizeof(struct UT_hash_bucket)); \
860
- (tbl)->num_buckets *= 2U; \
861
- (tbl)->log2_num_buckets++; \
862
- (tbl)->buckets = _he_new_buckets; \
863
- (tbl)->ineff_expands = ((tbl)->nonideal_items > ((tbl)->num_items >> 1)) ? \
864
- ((tbl)->ineff_expands+1U) : 0U; \
865
- if ((tbl)->ineff_expands > 1U) { \
866
- (tbl)->noexpand = 1; \
867
- uthash_noexpand_fyi(tbl); \
868
- } \
869
- uthash_expand_fyi(tbl); \
870
- } while (0)
871
-
872
-
873
- /* This is an adaptation of Simon Tatham's O(n log(n)) mergesort */
874
- /* Note that HASH_SORT assumes the hash handle name to be hh.
875
- * HASH_SRT was added to allow the hash handle name to be passed in. */
876
- #define HASH_SORT(head,cmpfcn) HASH_SRT(hh,head,cmpfcn)
877
- #define HASH_SRT(hh,head,cmpfcn) \
878
- do { \
879
- unsigned _hs_i; \
880
- unsigned _hs_looping,_hs_nmerges,_hs_insize,_hs_psize,_hs_qsize; \
881
- struct UT_hash_handle *_hs_p, *_hs_q, *_hs_e, *_hs_list, *_hs_tail; \
882
- if (head != NULL) { \
883
- _hs_insize = 1; \
884
- _hs_looping = 1; \
885
- _hs_list = &((head)->hh); \
886
- while (_hs_looping != 0U) { \
887
- _hs_p = _hs_list; \
888
- _hs_list = NULL; \
889
- _hs_tail = NULL; \
890
- _hs_nmerges = 0; \
891
- while (_hs_p != NULL) { \
892
- _hs_nmerges++; \
893
- _hs_q = _hs_p; \
894
- _hs_psize = 0; \
895
- for (_hs_i = 0; _hs_i < _hs_insize; ++_hs_i) { \
896
- _hs_psize++; \
897
- _hs_q = ((_hs_q->next != NULL) ? \
898
- HH_FROM_ELMT((head)->hh.tbl, _hs_q->next) : NULL); \
899
- if (_hs_q == NULL) { \
900
- break; \
901
- } \
902
- } \
903
- _hs_qsize = _hs_insize; \
904
- while ((_hs_psize != 0U) || ((_hs_qsize != 0U) && (_hs_q != NULL))) { \
905
- if (_hs_psize == 0U) { \
906
- _hs_e = _hs_q; \
907
- _hs_q = ((_hs_q->next != NULL) ? \
908
- HH_FROM_ELMT((head)->hh.tbl, _hs_q->next) : NULL); \
909
- _hs_qsize--; \
910
- } else if ((_hs_qsize == 0U) || (_hs_q == NULL)) { \
911
- _hs_e = _hs_p; \
912
- if (_hs_p != NULL) { \
913
- _hs_p = ((_hs_p->next != NULL) ? \
914
- HH_FROM_ELMT((head)->hh.tbl, _hs_p->next) : NULL); \
915
- } \
916
- _hs_psize--; \
917
- } else if ((cmpfcn( \
918
- DECLTYPE(head)(ELMT_FROM_HH((head)->hh.tbl, _hs_p)), \
919
- DECLTYPE(head)(ELMT_FROM_HH((head)->hh.tbl, _hs_q)) \
920
- )) <= 0) { \
921
- _hs_e = _hs_p; \
922
- if (_hs_p != NULL) { \
923
- _hs_p = ((_hs_p->next != NULL) ? \
924
- HH_FROM_ELMT((head)->hh.tbl, _hs_p->next) : NULL); \
925
- } \
926
- _hs_psize--; \
927
- } else { \
928
- _hs_e = _hs_q; \
929
- _hs_q = ((_hs_q->next != NULL) ? \
930
- HH_FROM_ELMT((head)->hh.tbl, _hs_q->next) : NULL); \
931
- _hs_qsize--; \
932
- } \
933
- if ( _hs_tail != NULL ) { \
934
- _hs_tail->next = ((_hs_e != NULL) ? \
935
- ELMT_FROM_HH((head)->hh.tbl, _hs_e) : NULL); \
936
- } else { \
937
- _hs_list = _hs_e; \
938
- } \
939
- if (_hs_e != NULL) { \
940
- _hs_e->prev = ((_hs_tail != NULL) ? \
941
- ELMT_FROM_HH((head)->hh.tbl, _hs_tail) : NULL); \
942
- } \
943
- _hs_tail = _hs_e; \
944
- } \
945
- _hs_p = _hs_q; \
946
- } \
947
- if (_hs_tail != NULL) { \
948
- _hs_tail->next = NULL; \
949
- } \
950
- if (_hs_nmerges <= 1U) { \
951
- _hs_looping = 0; \
952
- (head)->hh.tbl->tail = _hs_tail; \
953
- DECLTYPE_ASSIGN(head, ELMT_FROM_HH((head)->hh.tbl, _hs_list)); \
954
- } \
955
- _hs_insize *= 2U; \
956
- } \
957
- HASH_FSCK(hh, head, "HASH_SRT"); \
958
- } \
959
- } while (0)
960
-
961
- /* This function selects items from one hash into another hash.
962
- * The end result is that the selected items have dual presence
963
- * in both hashes. There is no copy of the items made; rather
964
- * they are added into the new hash through a secondary hash
965
- * hash handle that must be present in the structure. */
966
- #define HASH_SELECT(hh_dst, dst, hh_src, src, cond) \
967
- do { \
968
- unsigned _src_bkt, _dst_bkt; \
969
- void *_last_elt = NULL, *_elt; \
970
- UT_hash_handle *_src_hh, *_dst_hh, *_last_elt_hh=NULL; \
971
- ptrdiff_t _dst_hho = ((char*)(&(dst)->hh_dst) - (char*)(dst)); \
972
- if ((src) != NULL) { \
973
- for (_src_bkt=0; _src_bkt < (src)->hh_src.tbl->num_buckets; _src_bkt++) { \
974
- for (_src_hh = (src)->hh_src.tbl->buckets[_src_bkt].hh_head; \
975
- _src_hh != NULL; \
976
- _src_hh = _src_hh->hh_next) { \
977
- _elt = ELMT_FROM_HH((src)->hh_src.tbl, _src_hh); \
978
- if (cond(_elt)) { \
979
- _dst_hh = (UT_hash_handle*)(((char*)_elt) + _dst_hho); \
980
- _dst_hh->key = _src_hh->key; \
981
- _dst_hh->keylen = _src_hh->keylen; \
982
- _dst_hh->hashv = _src_hh->hashv; \
983
- _dst_hh->prev = _last_elt; \
984
- _dst_hh->next = NULL; \
985
- if (_last_elt_hh != NULL) { \
986
- _last_elt_hh->next = _elt; \
987
- } \
988
- if ((dst) == NULL) { \
989
- DECLTYPE_ASSIGN(dst, _elt); \
990
- HASH_MAKE_TABLE(hh_dst, dst); \
991
- } else { \
992
- _dst_hh->tbl = (dst)->hh_dst.tbl; \
993
- } \
994
- HASH_TO_BKT(_dst_hh->hashv, _dst_hh->tbl->num_buckets, _dst_bkt); \
995
- HASH_ADD_TO_BKT(_dst_hh->tbl->buckets[_dst_bkt], _dst_hh); \
996
- HASH_BLOOM_ADD(_dst_hh->tbl, _dst_hh->hashv); \
997
- (dst)->hh_dst.tbl->num_items++; \
998
- _last_elt = _elt; \
999
- _last_elt_hh = _dst_hh; \
1000
- } \
1001
- } \
1002
- } \
1003
- } \
1004
- HASH_FSCK(hh_dst, dst, "HASH_SELECT"); \
1005
- } while (0)
1006
-
1007
- #define HASH_CLEAR(hh,head) \
1008
- do { \
1009
- if ((head) != NULL) { \
1010
- HASH_BLOOM_FREE((head)->hh.tbl); \
1011
- uthash_free((head)->hh.tbl->buckets, \
1012
- (head)->hh.tbl->num_buckets*sizeof(struct UT_hash_bucket)); \
1013
- uthash_free((head)->hh.tbl, sizeof(UT_hash_table)); \
1014
- (head) = NULL; \
1015
- } \
1016
- } while (0)
1017
-
1018
- #define HASH_OVERHEAD(hh,head) \
1019
- (((head) != NULL) ? ( \
1020
- (size_t)(((head)->hh.tbl->num_items * sizeof(UT_hash_handle)) + \
1021
- ((head)->hh.tbl->num_buckets * sizeof(UT_hash_bucket)) + \
1022
- sizeof(UT_hash_table) + \
1023
- (HASH_BLOOM_BYTELEN))) : 0U)
1024
-
1025
- #ifdef NO_DECLTYPE
1026
- #define HASH_ITER(hh,head,el,tmp) \
1027
- for(((el)=(head)), ((*(char**)(&(tmp)))=(char*)((head!=NULL)?(head)->hh.next:NULL)); \
1028
- (el) != NULL; ((el)=(tmp)), ((*(char**)(&(tmp)))=(char*)((tmp!=NULL)?(tmp)->hh.next:NULL)))
1029
- #else
1030
- #define HASH_ITER(hh,head,el,tmp) \
1031
- for(((el)=(head)), ((tmp)=DECLTYPE(el)((head!=NULL)?(head)->hh.next:NULL)); \
1032
- (el) != NULL; ((el)=(tmp)), ((tmp)=DECLTYPE(el)((tmp!=NULL)?(tmp)->hh.next:NULL)))
1033
- #endif
1034
-
1035
- /* obtain a count of items in the hash */
1036
- #define HASH_COUNT(head) HASH_CNT(hh,head)
1037
- #define HASH_CNT(hh,head) ((head != NULL)?((head)->hh.tbl->num_items):0U)
1038
-
1039
- typedef struct UT_hash_bucket {
1040
- struct UT_hash_handle *hh_head;
1041
- unsigned count;
1042
-
1043
- /* expand_mult is normally set to 0. In this situation, the max chain length
1044
- * threshold is enforced at its default value, HASH_BKT_CAPACITY_THRESH. (If
1045
- * the bucket's chain exceeds this length, bucket expansion is triggered).
1046
- * However, setting expand_mult to a non-zero value delays bucket expansion
1047
- * (that would be triggered by additions to this particular bucket)
1048
- * until its chain length reaches a *multiple* of HASH_BKT_CAPACITY_THRESH.
1049
- * (The multiplier is simply expand_mult+1). The whole idea of this
1050
- * multiplier is to reduce bucket expansions, since they are expensive, in
1051
- * situations where we know that a particular bucket tends to be overused.
1052
- * It is better to let its chain length grow to a longer yet-still-bounded
1053
- * value, than to do an O(n) bucket expansion too often.
1054
- */
1055
- unsigned expand_mult;
1056
-
1057
- } UT_hash_bucket;
1058
-
1059
- /* random signature used only to find hash tables in external analysis */
1060
- #define HASH_SIGNATURE 0xa0111fe1u
1061
- #define HASH_BLOOM_SIGNATURE 0xb12220f2u
1062
-
1063
- typedef struct UT_hash_table {
1064
- UT_hash_bucket *buckets;
1065
- unsigned num_buckets, log2_num_buckets;
1066
- unsigned num_items;
1067
- struct UT_hash_handle *tail; /* tail hh in app order, for fast append */
1068
- ptrdiff_t hho; /* hash handle offset (byte pos of hash handle in element */
1069
-
1070
- /* in an ideal situation (all buckets used equally), no bucket would have
1071
- * more than ceil(#items/#buckets) items. that's the ideal chain length. */
1072
- unsigned ideal_chain_maxlen;
1073
-
1074
- /* nonideal_items is the number of items in the hash whose chain position
1075
- * exceeds the ideal chain maxlen. these items pay the penalty for an uneven
1076
- * hash distribution; reaching them in a chain traversal takes >ideal steps */
1077
- unsigned nonideal_items;
1078
-
1079
- /* ineffective expands occur when a bucket doubling was performed, but
1080
- * afterward, more than half the items in the hash had nonideal chain
1081
- * positions. If this happens on two consecutive expansions we inhibit any
1082
- * further expansion, as it's not helping; this happens when the hash
1083
- * function isn't a good fit for the key domain. When expansion is inhibited
1084
- * the hash will still work, albeit no longer in constant time. */
1085
- unsigned ineff_expands, noexpand;
1086
-
1087
- uint32_t signature; /* used only to find hash tables in external analysis */
1088
- #ifdef HASH_BLOOM
1089
- uint32_t bloom_sig; /* used only to test bloom exists in external analysis */
1090
- uint8_t *bloom_bv;
1091
- uint8_t bloom_nbits;
1092
- #endif
1093
-
1094
- } UT_hash_table;
1095
-
1096
- typedef struct UT_hash_handle {
1097
- struct UT_hash_table *tbl;
1098
- void *prev; /* prev element in app order */
1099
- void *next; /* next element in app order */
1100
- struct UT_hash_handle *hh_prev; /* previous hh in bucket order */
1101
- struct UT_hash_handle *hh_next; /* next hh in bucket order */
1102
- void *key; /* ptr to enclosing struct's key */
1103
- unsigned keylen; /* enclosing struct's key len */
1104
- unsigned hashv; /* result of hash-fcn(key) */
1105
- } UT_hash_handle;
1106
-
1107
- #endif /* UTHASH_H */