1996-07-09 08:22:35 +02:00
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
*
|
1999-02-14 00:22:53 +01:00
|
|
|
* hashfunc.c
|
1997-09-07 07:04:48 +02:00
|
|
|
* Comparison functions for hash access method.
|
1996-07-09 08:22:35 +02:00
|
|
|
*
|
2001-01-24 20:43:33 +01:00
|
|
|
* Portions Copyright (c) 1996-2001, PostgreSQL Global Development Group
|
2000-01-26 06:58:53 +01:00
|
|
|
* Portions Copyright (c) 1994, Regents of the University of California
|
1996-07-09 08:22:35 +02:00
|
|
|
*
|
|
|
|
*
|
|
|
|
* IDENTIFICATION
|
2002-03-06 21:49:46 +01:00
|
|
|
* $Header: /cvsroot/pgsql/src/backend/access/hash/hashfunc.c,v 1.32 2002/03/06 20:49:38 momjian Exp $
|
1996-07-09 08:22:35 +02:00
|
|
|
*
|
|
|
|
* NOTES
|
1997-09-07 07:04:48 +02:00
|
|
|
* These functions are stored in pg_amproc. For each operator class
|
|
|
|
* defined on hash tables, they compute the hash value of the argument.
|
1996-07-09 08:22:35 +02:00
|
|
|
*
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
*/
|
1996-10-20 08:34:30 +02:00
|
|
|
|
1996-07-09 08:22:35 +02:00
|
|
|
#include "postgres.h"
|
|
|
|
|
1996-11-10 04:06:38 +01:00
|
|
|
#include "access/hash.h"
|
|
|
|
|
2000-06-19 05:55:01 +02:00
|
|
|
|
|
|
|
Datum
|
|
|
|
hashchar(PG_FUNCTION_ARGS)
|
|
|
|
{
|
2001-03-22 05:01:46 +01:00
|
|
|
PG_RETURN_UINT32(~((uint32) PG_GETARG_CHAR(0)));
|
2000-06-19 05:55:01 +02:00
|
|
|
}
|
|
|
|
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
|
|
|
hashint2(PG_FUNCTION_ARGS)
|
1996-07-09 08:22:35 +02:00
|
|
|
{
|
2001-03-22 05:01:46 +01:00
|
|
|
PG_RETURN_UINT32(~((uint32) PG_GETARG_INT16(0)));
|
1996-07-09 08:22:35 +02:00
|
|
|
}
|
|
|
|
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
|
|
|
hashint4(PG_FUNCTION_ARGS)
|
1996-07-09 08:22:35 +02:00
|
|
|
{
|
2001-03-22 05:01:46 +01:00
|
|
|
PG_RETURN_UINT32(~PG_GETARG_UINT32(0));
|
1996-07-09 08:22:35 +02:00
|
|
|
}
|
|
|
|
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
|
|
|
hashint8(PG_FUNCTION_ARGS)
|
1999-03-14 06:09:05 +01:00
|
|
|
{
|
2000-06-05 09:29:25 +02:00
|
|
|
/* we just use the low 32 bits... */
|
2001-03-22 05:01:46 +01:00
|
|
|
PG_RETURN_UINT32(~((uint32) PG_GETARG_INT64(0)));
|
2000-06-19 05:55:01 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
Datum
|
|
|
|
hashoid(PG_FUNCTION_ARGS)
|
|
|
|
{
|
2001-03-22 05:01:46 +01:00
|
|
|
PG_RETURN_UINT32(~((uint32) PG_GETARG_OID(0)));
|
1999-03-14 06:09:05 +01:00
|
|
|
}
|
|
|
|
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
|
|
|
hashfloat4(PG_FUNCTION_ARGS)
|
1996-07-09 08:22:35 +02:00
|
|
|
{
|
2000-06-05 09:29:25 +02:00
|
|
|
float4 key = PG_GETARG_FLOAT4(0);
|
1997-09-07 07:04:48 +02:00
|
|
|
|
2000-06-19 05:55:01 +02:00
|
|
|
return hash_any((char *) &key, sizeof(key));
|
1997-09-07 07:04:48 +02:00
|
|
|
}
|
1996-07-09 08:22:35 +02:00
|
|
|
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
|
|
|
hashfloat8(PG_FUNCTION_ARGS)
|
1996-07-09 08:22:35 +02:00
|
|
|
{
|
2000-06-05 09:29:25 +02:00
|
|
|
float8 key = PG_GETARG_FLOAT8(0);
|
1996-07-09 08:22:35 +02:00
|
|
|
|
2000-06-19 05:55:01 +02:00
|
|
|
return hash_any((char *) &key, sizeof(key));
|
1996-07-09 08:22:35 +02:00
|
|
|
}
|
|
|
|
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
|
|
|
hashoidvector(PG_FUNCTION_ARGS)
|
1998-08-19 04:04:17 +02:00
|
|
|
{
|
2000-06-05 09:29:25 +02:00
|
|
|
Oid *key = (Oid *) PG_GETARG_POINTER(0);
|
1998-08-19 04:04:17 +02:00
|
|
|
|
2000-06-19 05:55:01 +02:00
|
|
|
return hash_any((char *) key, INDEX_MAX_KEYS * sizeof(Oid));
|
2000-02-21 04:36:59 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Note: hashint2vector currently can't be used as a user hash table
|
2000-04-12 19:17:23 +02:00
|
|
|
* hash function, because it has no pg_proc entry. We only need it
|
2000-02-21 04:36:59 +01:00
|
|
|
* for catcache indexing.
|
|
|
|
*/
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
|
|
|
hashint2vector(PG_FUNCTION_ARGS)
|
2000-02-21 04:36:59 +01:00
|
|
|
{
|
2000-06-05 09:29:25 +02:00
|
|
|
int16 *key = (int16 *) PG_GETARG_POINTER(0);
|
2000-02-21 04:36:59 +01:00
|
|
|
|
2000-06-19 05:55:01 +02:00
|
|
|
return hash_any((char *) key, INDEX_MAX_KEYS * sizeof(int16));
|
1998-08-19 04:04:17 +02:00
|
|
|
}
|
|
|
|
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
2000-06-19 05:55:01 +02:00
|
|
|
hashname(PG_FUNCTION_ARGS)
|
1996-07-09 08:22:35 +02:00
|
|
|
{
|
2001-03-22 05:01:46 +01:00
|
|
|
char *key = NameStr(*PG_GETARG_NAME(0));
|
1996-07-09 08:22:35 +02:00
|
|
|
|
2002-03-06 21:49:46 +01:00
|
|
|
Assert(strlen(key) <= NAMEDATALEN);
|
|
|
|
|
2002-02-25 05:06:52 +01:00
|
|
|
return hash_any(key, strlen(key));
|
1996-07-09 08:22:35 +02:00
|
|
|
}
|
|
|
|
|
2000-06-19 05:55:01 +02:00
|
|
|
/*
|
|
|
|
* hashvarlena() can be used for any varlena datatype in which there are
|
|
|
|
* no non-significant bits, ie, distinct bitpatterns never compare as equal.
|
|
|
|
*/
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
2000-06-19 05:55:01 +02:00
|
|
|
hashvarlena(PG_FUNCTION_ARGS)
|
1996-07-09 08:22:35 +02:00
|
|
|
{
|
2000-06-19 05:55:01 +02:00
|
|
|
struct varlena *key = PG_GETARG_VARLENA_P(0);
|
2000-12-09 00:57:03 +01:00
|
|
|
Datum result;
|
1997-09-07 07:04:48 +02:00
|
|
|
|
2000-12-09 00:57:03 +01:00
|
|
|
result = hash_any(VARDATA(key), VARSIZE(key) - VARHDRSZ);
|
|
|
|
|
|
|
|
/* Avoid leaking memory for toasted inputs */
|
|
|
|
PG_FREE_IF_COPY(key, 0);
|
|
|
|
|
|
|
|
return result;
|
1996-07-09 08:22:35 +02:00
|
|
|
}
|
|
|
|
|
2002-03-06 21:49:46 +01:00
|
|
|
/* This hash function was written by Bob Jenkins
|
|
|
|
* (bob_jenkins@burtleburtle.net), and superficially adapted
|
|
|
|
* for PostgreSQL by Neil Conway. For more information on this
|
|
|
|
* hash function, see http://burtleburtle.net/bob/hash/doobs.html
|
|
|
|
*/
|
2000-06-19 05:55:01 +02:00
|
|
|
|
1996-07-09 08:22:35 +02:00
|
|
|
/*
|
2002-03-06 21:49:46 +01:00
|
|
|
* mix -- mix 3 32-bit values reversibly.
|
|
|
|
* For every delta with one or two bits set, and the deltas of all three
|
|
|
|
* high bits or all three low bits, whether the original value of a,b,c
|
|
|
|
* is almost all zero or is uniformly distributed,
|
|
|
|
* - If mix() is run forward or backward, at least 32 bits in a,b,c
|
|
|
|
* have at least 1/4 probability of changing.
|
|
|
|
* - If mix() is run forward, every bit of c will change between 1/3 and
|
|
|
|
* 2/3 of the time. (Well, 22/100 and 78/100 for some 2-bit deltas.)
|
|
|
|
*/
|
|
|
|
#define mix(a,b,c) \
|
|
|
|
{ \
|
|
|
|
a -= b; a -= c; a ^= (c>>13); \
|
|
|
|
b -= c; b -= a; b ^= (a<<8); \
|
|
|
|
c -= a; c -= b; c ^= (b>>13); \
|
|
|
|
a -= b; a -= c; a ^= (c>>12); \
|
|
|
|
b -= c; b -= a; b ^= (a<<16); \
|
|
|
|
c -= a; c -= b; c ^= (b>>5); \
|
|
|
|
a -= b; a -= c; a ^= (c>>3); \
|
|
|
|
b -= c; b -= a; b ^= (a<<10); \
|
|
|
|
c -= a; c -= b; c ^= (b>>15); \
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* hash_any() -- hash a variable-length key into a 32-bit value
|
|
|
|
* k : the key (the unaligned variable-length array of bytes)
|
|
|
|
* len : the length of the key, counting by bytes
|
|
|
|
* Returns a 32-bit value. Every bit of the key affects every bit of
|
|
|
|
* the return value. Every 1-bit and 2-bit delta achieves avalanche.
|
|
|
|
* About 6*len+35 instructions. The best hash table sizes are powers
|
|
|
|
* of 2. There is no need to do mod a prime (mod is sooo slow!).
|
|
|
|
* If you need less than 32 bits, use a bitmask.
|
1996-07-09 08:22:35 +02:00
|
|
|
*/
|
2000-06-05 09:29:25 +02:00
|
|
|
Datum
|
2002-03-06 21:49:46 +01:00
|
|
|
hash_any(register const char *k, register int keylen)
|
1996-07-09 08:22:35 +02:00
|
|
|
{
|
2002-03-06 21:49:46 +01:00
|
|
|
register Datum a,b,c,len;
|
|
|
|
|
|
|
|
/* Set up the internal state */
|
|
|
|
len = keylen;
|
|
|
|
a = b = 0x9e3779b9; /* the golden ratio; an arbitrary value */
|
|
|
|
/* Another arbitrary value. If the hash function is called
|
|
|
|
* multiple times, this could be the previously generated
|
|
|
|
* hash value; however, the interface currently doesn't allow
|
|
|
|
* this. AFAIK this isn't a big deal.
|
|
|
|
*/
|
|
|
|
c = 3923095;
|
|
|
|
|
|
|
|
/* handle most of the key */
|
|
|
|
while (len >= 12)
|
|
|
|
{
|
|
|
|
a += (k[0] +((Datum)k[1]<<8) +((Datum)k[2]<<16) +((Datum)k[3]<<24));
|
|
|
|
b += (k[4] +((Datum)k[5]<<8) +((Datum)k[6]<<16) +((Datum)k[7]<<24));
|
|
|
|
c += (k[8] +((Datum)k[9]<<8) +((Datum)k[10]<<16)+((Datum)k[11]<<24));
|
|
|
|
mix(a,b,c);
|
|
|
|
k += 12; len -= 12;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* handle the last 11 bytes */
|
|
|
|
c += keylen;
|
|
|
|
switch(len) /* all the case statements fall through */
|
|
|
|
{
|
|
|
|
case 11: c+=((Datum)k[10]<<24);
|
|
|
|
case 10: c+=((Datum)k[9]<<16);
|
|
|
|
case 9 : c+=((Datum)k[8]<<8);
|
|
|
|
/* the first byte of c is reserved for the length */
|
|
|
|
case 8 : b+=((Datum)k[7]<<24);
|
|
|
|
case 7 : b+=((Datum)k[6]<<16);
|
|
|
|
case 6 : b+=((Datum)k[5]<<8);
|
|
|
|
case 5 : b+=k[4];
|
|
|
|
case 4 : a+=((Datum)k[3]<<24);
|
|
|
|
case 3 : a+=((Datum)k[2]<<16);
|
|
|
|
case 2 : a+=((Datum)k[1]<<8);
|
|
|
|
case 1 : a+=k[0];
|
|
|
|
/* case 0: nothing left to add */
|
|
|
|
}
|
|
|
|
mix(a,b,c);
|
|
|
|
/* report the result */
|
|
|
|
return c;
|
1997-09-07 07:04:48 +02:00
|
|
|
}
|