1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
|
/*
--------------------------------------------------------------------
By Bob Jenkins, 1996. hashtab.c. Public Domain.
This implements a hash table.
* Keys are unique. Adding an item fails if the key is already there.
* Keys and items are pointed at, not copied. If you change the value
of the key after it is inserted then hfind will not be able to find it.
* The hash table maintains a position that can be set and queried.
* The table length doubles dynamically and never shrinks. The insert
that causes table doubling may take a long time.
* The table length splits when the table length equals the number of items
Comparisons usually take 7 instructions.
Computing a hash value takes 35+6n instructions for an n-byte key.
hcreate - create a hash table
hdestroy - destroy a hash table
hcount - The number of items in the hash table
hkey - key at the current position
hkeyl - key length at the current position
hstuff - stuff at the current position
hfind - find an item in the table
hadd - insert an item into the table
hdel - delete an item from the table
hstat - print statistics about the table
hfirst - position at the first item in the table
hnext - move the position to the next item in the table
--------------------------------------------------------------------
*/
#include "generic.h"
#ifndef LOOKUPA
#include "lookupa.h"
#endif
#ifndef HASHTAB
#include "hashtab.h"
#endif
#ifndef RECYCLE
#include "recycle.h"
#endif
#ifndef FALSE
#define FALSE 0
#endif
#ifndef TRUE
#define TRUE !FALSE
#endif
/* sanity check -- make sure ipos, apos, and count make sense */
static void hsanity(t)
htab *t;
{
Uint32 i, end, counter;
hitem *h;
/* test that apos makes sense */
end = (Uint32)1<<(t->logsize);
if (end < t->apos)
printf("error: end %ld apos %ld\n", end, t->apos);
/* test that ipos is in bucket apos */
if (t->ipos)
{
for (h=t->table[t->apos]; h && h != t->ipos; h = h->next)
;
if (h != t->ipos)
printf("error:ipos not in apos, apos is %ld\n", t->apos);
}
/* test that t->count is the number of elements in the table */
counter=0;
for (counter=0, i=0; i<end; ++i)
for (h=t->table[i]; h; h=h->next)
++counter;
if (counter != t->count)
printf("error: counter %ld t->count %ld\n", counter, t->count);
}
/*
* hgrow - Double the size of a hash table.
* Allocate a new, 2x bigger array,
* move everything from the old array to the new array,
* then free the old array.
*/
static void hgrow( t)
htab *t; /* table */
{
register Uint32 newsize = (Uint32)1<<(++t->logsize);
register Uint32 newmask = newsize-1;
register Uint32 i;
register hitem **oldtab = t->table;
register hitem **newtab = (hitem **)malloc(newsize*sizeof(hitem *));
/* make sure newtab is cleared */
for (i=0; i<newsize; ++i) newtab[i] = (hitem *)0;
t->table = newtab;
t->mask = newmask;
/* Walk through old table putting entries in new table */
for (i=newsize>>1; i--;)
{
register hitem *this, *that, **newplace;
for (this = oldtab[i]; this;)
{
that = this;
this = this->next;
newplace = &newtab[(that->hval & newmask)];
that->next = *newplace;
*newplace = that;
}
}
/* position the hash table on some existing item */
hfirst(t);
/* free the old array */
free((char *)oldtab);
}
/* hcreate - create a hash table initially of size power(2,logsize) */
htab *hcreate(logsize)
int logsize; /* log base 2 of the size of the hash table */
{
Uint32 i,len;
htab *t = (htab *)malloc(sizeof(htab));
len = ((Uint32)1<<logsize);
t->table = (hitem **)malloc(sizeof(hitem *)*(Uint32)len);
for (i=0; i<len; ++i) t->table[i] = (hitem *)0;
t->logsize = logsize;
t->mask = len-1;
t->count = 0;
t->apos = (Uint32)0;
t->ipos = (hitem *)0;
t->space = remkroot(sizeof(hitem));
t->bcount = 0;
return t;
}
/* hdestroy - destroy the hash table and free all its memory */
void hdestroy( t)
htab *t; /* the table */
{
hitem *h;
refree(t->space);
free((char *)t->table);
free((char *)t);
}
/* hcount() is a macro, see hashtab.h */
/* hkey() is a macro, see hashtab.h */
/* hkeyl() is a macro, see hashtab.h */
/* hstuff() is a macro, see hashtab.h */
/* hfind - find an item with a given key in a hash table */
int hfind( t, key, keyl )
htab *t; /* table */
Uint8 *key; /* key to find */
Uint32 keyl; /* key length */
{
hitem *h;
Uint32 x = lookup(key,keyl,0);
Uint32 y;
for (h = t->table[y=(x&t->mask)]; h; h = h->next)
{
if ((x == h->hval) &&
(keyl == h->keyl) &&
!memcmp(key, h->key, keyl))
{
t->apos = y;
t->ipos = h;
return TRUE;
}
}
return FALSE;
}
/*
* hadd - add an item to a hash table.
* return FALSE if the key is already there, otherwise TRUE.
*/
int hadd( t, key, keyl, stuff)
htab *t; /* table */
Uint8 *key; /* key to add to hash table */
Uint32 keyl; /* key length */
void *stuff; /* stuff to associate with this key */
{
register hitem *h,**hp;
register Uint32 y, x = lookup(key,keyl,0);
/* make sure the key is not already there */
for (h = t->table[(y=(x&t->mask))]; h; h = h->next)
{
if ((x == h->hval) &&
(keyl == h->keyl) &&
!memcmp(key, h->key, keyl))
{
t->apos = y;
t->ipos = h;
return FALSE;
}
}
/* find space for a new item */
h = (hitem *)renew(t->space);
/* make the hash table bigger if it is getting full */
if (++t->count > (Uint32)1<<(t->logsize))
{
hgrow(t);
y = (x&t->mask);
}
/* add the new key to the table */
h->key = key;
h->keyl = keyl;
h->stuff = stuff;
h->hval = x;
hp = &t->table[y];
h->next = *hp;
*hp = h;
t->ipos = h;
t->apos = y;
#ifdef HSANITY
hsanity(t);
#endif /* HSANITY */
return TRUE;
}
/* hdel - delete the item at the current position */
int hdel(t)
htab *t; /* the hash table */
{
hitem *h; /* item being deleted */
hitem **ip; /* a counter */
/* check for item not existing */
if (!(h = t->ipos)) return FALSE;
/* remove item from its list */
for (ip = &t->table[t->apos]; *ip != h; ip = &(*ip)->next)
;
*ip = (*ip)->next;
--(t->count);
/* adjust position to something that exists */
if (!(t->ipos = h->next)) hnbucket(t);
/* recycle the deleted hitem node */
redel(t->space, h);
#ifdef HSANITY
hsanity(t);
#endif /* HSANITY */
return TRUE;
}
/* hfirst - position on the first element in the table */
int hfirst(t)
htab *t; /* the hash table */
{
t->apos = t->mask;
(void)hnbucket(t);
return (t->ipos != (hitem *)0);
}
/* hnext() is a macro, see hashtab.h */
/*
* hnbucket - Move position to the first item in the next bucket.
* Return TRUE if we did not wrap around to the beginning of the table
*/
int hnbucket(t)
htab *t;
{
Uint32 oldapos = t->apos;
Uint32 end = (Uint32)1<<(t->logsize);
Uint32 i;
/* see if the element can be found without wrapping around */
for (i=oldapos+1; i<end; ++i)
{
if (t->table[i&t->mask])
{
t->apos = i;
t->ipos = t->table[i];
return TRUE;
}
}
/* must have to wrap around to find the last element */
for (i=0; i<=oldapos; ++i)
{
if (t->table[i])
{
t->apos = i;
t->ipos = t->table[i];
return FALSE;
}
}
return FALSE;
}
void hstat(t)
htab *t;
{
Uint32 i,j;
double total = 0.0;
hitem *h;
hitem *walk, *walk2, *stat = (hitem *)0;
/* in stat, keyl will store length of list, hval the number of buckets */
for (i=0; i<=t->mask; ++i)
{
for (h=t->table[i], j=0; h; ++j, h=h->next)
;
for (walk=stat; walk && (walk->keyl != j); walk=walk->next)
;
if (walk)
{
++(walk->hval);
}
else
{
walk = (hitem *)renew(t->space);
walk->keyl = j;
walk->hval = 1;
if (!stat || stat->keyl > j) {walk->next=stat; stat=walk;}
else
{
for (walk2=stat;
walk2->next && (walk2->next->keyl<j);
walk2=walk2->next)
;
walk->next = walk2->next;
walk2->next = walk;
}
}
}
/* figure out average list length for existing elements */
for (walk=stat; walk; walk=walk->next)
{
total+=(double)walk->hval*(double)walk->keyl*(double)walk->keyl;
}
if (t->count) total /= (double)t->count;
else total = (double)0;
/* print statistics */
printf("\n");
for (walk=stat; walk; walk=walk->next)
{
printf("items %ld: %ld buckets\n", walk->keyl, walk->hval);
}
printf("\nbuckets: %ld items: %ld existing: %g\n\n",
((Uint32)1<<t->logsize), t->count, total);
/* clean up */
while (stat)
{
walk = stat->next;
redel(t->space, stat);
stat = walk;
}
}
|