6b1e113f9f
Hoist the deque->maxlen lookup out of the inner-loop.
2015-10-11 09:43:50 -07:00
a7f63009d6
Minor tweak. Make the maxlen comparisons a little more clear and consistent.
2015-10-10 23:56:02 -04:00
848f2b595d
Backport early-out 91259f061cfb to reduce the cost of bb1a2944bcb6
2015-10-06 23:06:17 -04:00
0443ac2860
Eliminate unnecessary test
2015-10-05 22:52:37 -04:00
fd265f4a18
Factor out common iterator finalization code
2015-10-02 23:17:33 -07:00
0d30940dd2
Add fast paths to deque_init() for the common cases
2015-09-30 23:15:02 -07:00
38031143fb
Add an early-out for deque_clear()
2015-09-29 22:45:05 -07:00
8299e9b59e
Move the copy and clear functions upwards to eliminate unnecessary forward references.
2015-09-26 21:31:23 -07:00
7c0b70f419
Minor tweak to the order of variable updates.
2015-09-26 21:11:05 -07:00
f2b02ced7e
Bump up the maximum number of freeblocks
2015-09-26 17:47:02 -07:00
c22eee6b59
Precomputing the number iterations allows the inner-loop to be vectorizable.
2015-09-26 02:14:50 -07:00
7a84552c84
Hoist constant expression out of an inner loop.
2015-09-26 01:30:51 -07:00
e055b88937
merge
2015-09-26 00:15:46 -07:00
bf49fee125
Issue #25135 : Avoid possible reentrancy issues in deque_clear.
2015-09-26 00:14:59 -07:00
2b0d646b75
Replace an unpredictable branch with a simple addition.
2015-09-23 19:15:44 -07:00
3a1a8d0424
Eliminate unnecessary variables
2015-09-23 02:42:02 -07:00
7a237230d1
Eliminate unnecessary variable
2015-09-22 01:20:36 -07:00
d3d2b2c50c
Minor consistency improvements to negative value comparisons.
2015-09-21 23:41:56 -07:00
aed8830af3
Add a fast path (no iterator creation) for a common case for repeating deques of size 1
2015-09-19 09:05:42 -07:00
0e14e6610b
Hoist constant expression out of an inner loop
2015-09-19 00:21:33 -06:00
ad26225e1a
Tighten inner-loop for deque_inplace_repeat().
2015-09-14 01:03:04 -04:00
e4f3467df1
Add an exact type match fast path for deque_copy().
2015-09-13 19:27:01 -04:00
95e2cc5d12
Fix refcount.
2015-09-13 02:41:18 -04:00
67c78b5421
In-line the append operations inside deque_inplace_repeat().
2015-09-12 11:00:20 -04:00
f5d72f35e8
Simply deque repeat by reusing code in in-line repeat. Avoid unnecessary division.
2015-09-09 22:39:44 -04:00
318c5aef0e
merge
2015-09-05 17:06:18 -07:00
2b2b75374e
Prevent reentrant badness by deferring the decrefs as long as possible.
2015-09-05 17:05:52 -07:00
7a1a0bbbf1
merge
2015-08-26 08:09:50 -07:00
87674ec7d5
Issue #24913 : Fix overrun error in deque.index().
2015-08-26 08:08:38 -07:00
28c995d03b
Minor cleanup.
2015-08-14 02:07:41 -07:00
12f896c521
Fix minor typo in a comment
2015-07-31 12:03:20 -07:00
aa23fa2e21
merge
2015-07-20 03:11:16 -04:00
043e800860
merge
2015-07-20 03:10:48 -04:00
239aba7874
Issue #19663 : Improve error message for defaultdict.
2015-07-20 03:09:22 -04:00
cfe5b6ca04
Divisions-by-two for a positive Py_ssize_t compile more cleanly with >>1 than /2.
2015-07-20 00:25:50 -04:00
96c6af9b20
Issue #16991 : Add a C implementation of collections.OrderedDict.
2015-05-29 22:21:39 -06:00
47db71756d
Issue #16991 : Add a C implementation of collections.OrderedDict.
2015-05-29 22:21:39 -06:00
c6249a6268
Defer deleted item decref until after the deque is restored to a consistent state.
2015-05-02 10:44:17 -07:00
1a6292194b
fix refleak in deque_concat
2015-04-04 10:52:36 -04:00
41290a68ba
Issue 23793: Add deque support for __add__(), __mul__(), and __imul__().
2015-03-31 08:12:23 -07:00
fef9c1b368
Fix typo
2015-03-24 21:12:57 -07:00
8dbbae2e52
Minor code and comment cleanups.
2015-03-24 21:01:50 -07:00
551350a79f
Improve and fix-up comments.
2015-03-24 00:19:53 -07:00
0f1451c8c0
Issue 23744: Minor speed-up for deque.__bool__().
2015-03-23 23:23:55 -07:00
6921c13bbb
Minor nit. Make the rotate() success/fail tests consistent.
2015-03-21 02:03:40 -07:00
ac13ad6a32
For safety, wait to decref deleted values until the deque state has been restored.
2015-03-21 01:53:16 -07:00
0f6f94778a
Fix minor formatting nits and remove unnecessary comment.
2015-03-21 01:42:10 -07:00
32ea16577d
Issue 23704: Add index(), copy(), and insert() to deques. Register deques as a MutableSequence.
2015-03-21 01:37:37 -07:00
39dadf7abf
Issue 23705: Improve the performance of __contains__ checks for deques.
2015-03-20 16:38:56 -07:00
87e691240b
Minor neatening-up. Make assignments in same order a struct fields. Line-up comments.
2015-03-02 23:32:02 -08:00
f9d9c79aa8
Switch the state variable to unsigned for defined wrap-around behavior.
2015-03-02 22:47:46 -08:00
30c9074b96
Minor beautification. Move struct definitions to the top. Fix-up a comment.
2015-03-02 22:31:35 -08:00
f30f5b9ba6
Minor code beautification. Replace macro with in-lineable functions.
2015-03-02 22:23:37 -08:00
3c186ba441
Beautify and better document the use of the size_t cast for bounds checking.
2015-03-02 21:45:02 -08:00
7f9ea7543e
Issue #23553 : Use an unsigned cast to tighten-up the bounds checking logic.
2015-03-01 00:38:00 -08:00
c20830804d
Need a (size_t) cast instead of (unsigned) to be big enough for a Py_ssize_t.
2015-02-28 23:29:16 -08:00
a473b9da15
Use unsigned division and modulo for item assignment as well.
2015-02-28 17:49:47 -08:00
63d1ff2a0b
Convert one more division to unsigned arithmetic to speed-up deque_item().
2015-02-28 07:41:30 -08:00
7e8c7956a7
Line missed in last checkin
2015-02-27 16:59:29 -08:00
da2850f932
Since the index is always non-negative, use faster unsigned division and modulo.
2015-02-27 12:42:54 -08:00
daf57f25e5
Bump the blocksize up from 62 to 64 to speed up the modulo calculation.
...
Remove the old comment suggesting that it was desireable to have
blocksize+2 as a multiple of the cache line length. That would
have made sense only if the block structure start point was always
aligned to a cache line boundary. However, the memory allocations
are 16 byte aligned, so we don't really have control over whether
the struct spills across cache line boundaries.
2015-02-26 23:21:29 -08:00
eb6b554fbc
Update copyright.
2015-02-10 22:37:22 -06:00
0e259f18f7
Optimization guides suggest copying memory in an ascending direction when possible.
2015-02-01 22:53:41 -08:00
507d997714
Add comment and make minor code clean-up to improve clarity.
2014-05-18 21:32:40 +01:00
4b0b1accb5
Issue #21101 : Eliminate double hashing in the C code for collections.Counter().
2014-05-03 16:41:19 -07:00
5402315626
Add implementation notes
2014-04-23 00:58:48 -07:00
e19d9d1467
merge 3.3 ( #20250 )
2014-01-13 23:56:30 -05:00
9cb33b7d03
correct defaultdict signature in docstring ( closes #20250 )
...
Patch from Andrew Barnert.
2014-01-13 23:56:05 -05:00
e7f516cbb8
Issue #19512 : _count_elements() of _collections reuses PyId_get identifier
...
instead of literal "get" string
2013-11-06 23:52:55 +01:00
07573d7b24
merge
2013-10-04 16:52:39 -07:00
cb1d96f782
Issue #18594 : Make the C code more closely match the pure python code.
2013-10-04 16:51:02 -07:00
75f65e368e
merge
2013-10-01 21:38:37 -07:00
224c87d60c
Issue #18594 : Fix the fallback path in collections.Counter().
2013-10-01 21:36:09 -07:00
c13516b0a0
merge
2013-10-01 01:00:59 -07:00
2ff2190b62
Issue #18594 : Fix the fast path for collections.Counter().
...
The path wasn't being taken due to an over-restrictive type check.
2013-10-01 00:55:43 -07:00
77578204d6
Restore the data block size to 62.
...
The former block size traded away good fit within cache lines in
order to gain faster division in deque_item(). However, compilers
are getting smarter and can now replace the slow division operation
with a fast integer multiply and right shift. Accordingly, it makes
sense to go back to a size that lets blocks neatly fill entire
cache-lines.
GCC-4.8 and CLANG 4.0 both compute "x // 62" with something
roughly equivalent to "x * 9520900167075897609 >> 69".
2013-07-28 02:39:49 -07:00
3223dd5c22
Assertions key off NDEBUG
2013-07-26 23:14:22 -07:00
b97cc49c3a
Minor code simplification by eliminating an unnecessary temporary variable.
2013-07-21 01:51:07 -07:00
90dea4ce43
Tweak the deque struct by moving the least used fields (maxlen and weakref) to the end.
2013-07-13 22:30:25 -07:00
840533bf1c
Use a do-while loop in the inner loop for rotate (m is always greater than zero).
2013-07-13 17:03:58 -07:00
3959af9b2a
Move the freeblock() call outside the main loop to speed-up and simplify the block re-use logic.
2013-07-13 02:34:08 -07:00
d9c116ca40
Add a spacing saving heuristic to deque's extend methods
2013-07-09 00:13:21 -07:00
b385529ddf
Fix #ifdef
2013-07-07 02:07:23 -10:00
82df925451
Use macros for marking and checking endpoints in the doubly-linked list of blocks.
...
* Add comment explaining the endpoint checks
* Only do the checks in a debug build
* Simplify newblock() to only require a length argument
and leave the link updates to the calling code.
* Also add comment for the freelisting logic.
2013-07-07 01:43:42 -10:00
f3a67b7e57
Improve variable names in deque_count()
2013-07-06 17:49:06 -10:00
df715ba54d
Apply the PyObject_VAR_HEAD and Py_SIZE macros
...
to be consistent with practices in other modules.
2013-07-06 13:01:13 -10:00
5bfa8671bc
Refactor deque_traverse().
...
Hoist conditional expression out of the loop.
Use rightblock as the guard instead of checking for NULL.
2013-07-06 11:58:09 -10:00
98054b4c1b
Remove unnecessary branches from count() and reverse().
2013-07-06 09:07:06 -10:00
de68e0cf0e
Speed-up deque indexing by changing the deque block length to a power of two.
...
The division and modulo calculation in deque_item() can be compiled
to fast bitwise operations when the BLOCKLEN is a power of two.
Timing before:
~/cpython $ py -m timeit -r7 -s 'from collections import deque' -s 'd=deque(range(10))' 'd[5]'
10000000 loops, best of 7: 0.0627 usec per loop
Timing after:
~/cpython $ py -m timeit -r7 -s 'from collections import deque' -s 'd=deque(range(10))' 'd[5]'
10000000 loops, best of 7: 0.0581 usec per loop
2013-07-05 18:05:29 -10:00
20b0f87e1d
Misc improvements to collections.deque()
...
* Clarified comment on the impact of BLOCKLEN on deque_index
(with a power-of-two, the division and modulo
computations are done with a right-shift and bitwise-and).
* Clarified comment on the overflow check to note that
it is general and not just applicable the 64-bit builds.
* In deque._rotate(), the "deque->" indirections are
factored-out of the loop (loop invariant code motion),
leaving the code cleaner looking and slightly faster.
* In deque._rotate(), replaced the memcpy() with an
equivalent loop. That saved the memcpy setup time
and allowed the pointers to move in their natural
leftward and rightward directions.
See comparative timings at: http://pastebin.com/p0RJnT5N
2013-06-23 15:44:33 -07:00
986bbfc079
Backport deque.rotate() improvements.
2013-02-09 20:00:55 -05:00
59cf23ab07
Minor tweaks to varnames, declarations, and comments.
2013-02-07 00:57:19 -05:00
1f0044c473
Minor variable access clean-ups for deque.rotate().
2013-02-05 01:30:46 -05:00
a4409c18eb
Minor edits: Tighten-up the halflen logic and touch-up the assertions and comments.
2013-02-04 00:08:12 -05:00
3a9ae7fd98
Issue 16398: One more assertion for good measure.
2013-02-02 12:26:37 -08:00
231ee4dc9d
Issue 16398: Add assertions to show why memcmp is safe.
2013-02-02 11:24:43 -08:00
21777acd68
Issue 16398: Use memcpy() in deque.rotate().
2013-02-02 09:56:08 -08:00
fa3965ab76
merge 3.3
2013-01-12 21:22:33 -05:00
0e5c48a917
make deque_clear void, since it's infallible
2013-01-12 21:22:18 -05:00
464d89b3ce
Issue #16398 : Optimize deque.rotate()
2013-01-11 22:29:50 -08:00
Raymond Hettinger
Eric Snow
Benjamin Peterson
Victor Stinner