summaryrefslogtreecommitdiffstats
path: root/Objects/dictobject.c
diff options
context:
space:
mode:
authorGuido van Rossum <guido@python.org>1997-08-18 21:52:47 (GMT)
committerGuido van Rossum <guido@python.org>1997-08-18 21:52:47 (GMT)
commitfd7a0b871f9114bc989c4f9b2ea4ee3a8b6dd662 (patch)
treedaf81f59aa253cdaf5c2cf44ad66561e0d0385f8 /Objects/dictobject.c
parent2da391f3876a79d2662454babe064bcb5e3524ae (diff)
downloadcpython-fd7a0b871f9114bc989c4f9b2ea4ee3a8b6dd662.zip
cpython-fd7a0b871f9114bc989c4f9b2ea4ee3a8b6dd662.tar.gz
cpython-fd7a0b871f9114bc989c4f9b2ea4ee3a8b6dd662.tar.bz2
Made lookdict nearly twice as fast, resulting in a 5% overall
improvement of pystone. Vladimir Marangozov.
Diffstat (limited to 'Objects/dictobject.c')
-rw-r--r--Objects/dictobject.c24
1 files changed, 13 insertions, 11 deletions
diff --git a/Objects/dictobject.c b/Objects/dictobject.c
index 018762b..66cec0c 100644
--- a/Objects/dictobject.c
+++ b/Objects/dictobject.c
@@ -151,26 +151,25 @@ Subsequent probe indices use the values of x^i in GF(2^n) as an offset,
where x is a root. The initial value is derived from sum, too.
(This version is due to Reimer Behrends, some ideas are also due to
-Jyrki Alakuijala.)
+Jyrki Alakuijala and Vladimir Marangozov.)
*/
static dictentry *lookdict Py_PROTO((dictobject *, PyObject *, long));
static dictentry *
lookdict(mp, key, hash)
dictobject *mp;
PyObject *key;
- long hash;
+ register long hash;
{
register int i;
register unsigned incr;
- register unsigned long sum = (unsigned long) hash;
- register dictentry *freeslot = NULL;
+ register dictentry *freeslot;
register unsigned int mask = mp->ma_size-1;
dictentry *ep0 = mp->ma_table;
register dictentry *ep;
/* We must come up with (i, incr) such that 0 <= i < ma_size
and 0 < incr < ma_size and both are a function of hash */
- i = (~sum) & mask;
- /* We use ~sum instead if sum, as degenerate hash functions, such
+ i = (~hash) & mask;
+ /* We use ~hash instead of hash, as degenerate hash functions, such
as for ints <sigh>, can have lots of leading zeros. It's not
really a performance risk, but better safe than sorry. */
ep = &ep0[i];
@@ -178,19 +177,22 @@ lookdict(mp, key, hash)
return ep;
if (ep->me_key == dummy)
freeslot = ep;
- else if (ep->me_key == key ||
+ else {
+ if (ep->me_key == key ||
(ep->me_hash == hash &&
PyObject_Compare(ep->me_key, key) == 0))
- {
- return ep;
+ {
+ return ep;
+ }
+ freeslot = NULL;
}
/* XXX What if PyObject_Compare returned an exception? */
/* Derive incr from sum, just to make it more arbitrary. Note that
incr must not be 0, or we will get into an infinite loop.*/
- incr = (sum ^ (sum >> 3)) & mask;
+ incr = (hash ^ ((unsigned long)hash >> 3)) & mask;
if (!incr)
incr = mask;
- if (incr > mask) /* Cycle through GF(2^n)-{0} */
+ else if (incr > mask) /* Cycle through GF(2^n)-{0} */
incr ^= mp->ma_poly; /* This will implicitly clear the
highest bit */
for (;;) {