Message68174
| Author | pitrou |
|---|---|
| Recipients | pitrou |
| Date | 2008-06-13.18:37:58 |
| SpamBayes Score | 0.0061075324 |
| Marked as misclassified | No |
| Message-id | <1213382283.81.0.00112227094469.issue3106@psf.upfronthosting.co.za> |
| In-reply-to |
| Content | |
|---|---|
This patch is an experiment in making faster some of the most common
comparisons (str vs. str, int vs. int). I don't know if it may bring
noticeable speedups in real-world situations, but here are the synthetic
benchmark numbers (from pybench, "this" is the patched version and
"other" is vanilla py3k):
Test minimum run-time average run-time
this other diff this other
diff
-------------------------------------------------------------------------------
CompareFloats: 182ms 173ms +5.4% 182ms 176ms
+3.4%
CompareFloatsIntegers: 238ms 232ms +2.3% 242ms 236ms
+2.5%
CompareIntegers: 237ms 277ms -14.4% 237ms 280ms
-15.2%
CompareInternedStrings: 163ms 257ms -36.7% 163ms 258ms
-36.7%
CompareLongs: 137ms 160ms -14.5% 137ms 162ms
-15.6%
CompareStrings: 149ms 170ms -12.1% 154ms 170ms
-9.5%
-------------------------------------------------------------------------------
Totals: 1105ms 1268ms -12.9% 1115ms 1281ms
-13.0% |
|
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2008-06-13 18:38:04 | pitrou | set | spambayes_score: 0.00610753 -> 0.0061075324 recipients: + pitrou |
| 2008-06-13 18:38:04 | pitrou | set | spambayes_score: 0.00610753 -> 0.00610753 messageid: <1213382283.81.0.00112227094469.issue3106@psf.upfronthosting.co.za> |
| 2008-06-13 18:38:02 | pitrou | link | issue3106 messages |
| 2008-06-13 18:38:01 | pitrou | create | |