Message183365
| Author | pitrou |
|---|---|
| Recipients | alex, benjamin.peterson, pitrou, rhettinger |
| Date | 2013-03-03.10:29:59 |
| SpamBayes Score | -1.0 |
| Marked as misclassified | Yes |
| Message-id | <1362306380.3470.2.camel@localhost.localdomain> |
| In-reply-to | <1362282521.75.0.367136904932.issue17338@psf.upfronthosting.co.za> |
| Content | |
|---|---|
> There's an obvious need for this, CPython uses optimizations like this
> internally, as does PyPy.
I don't know if it's a need or if it's just "nice to have".
By the way, in the list case it also makes C code simpler, since once
the list is preallocated you just have to call PyList_SET_ITEM to
populate it, and there's no error return to worry about.
Also, lists are easy to pre-allocate in pure Python as well:
l = [None] * N
# populate
for i in range(N):
l[i] = ... |
|
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2013-03-03 10:29:59 | pitrou | set | recipients: + pitrou, rhettinger, benjamin.peterson, alex |
| 2013-03-03 10:29:59 | pitrou | link | issue17338 messages |
| 2013-03-03 10:29:59 | pitrou | create | |