{{ message }}
Repository navigation
Expand file tree
/
Copy path_heapqmodule.c
More file actions
811 lines (702 loc) · 24.8 KB
/
Copy path_heapqmodule.c
File metadata and controls
811 lines (702 loc) · 24.8 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
/* Drop in replacement for heapq.py
C implementation derived directly from heapq.py in Py2.3
which was written by Kevin O'Connor, augmented by Tim Peters,
annotated by François Pinard, and converted to C by Raymond Hettinger.
*/
#ifndef Py_BUILD_CORE_BUILTIN
# define Py_BUILD_CORE_MODULE 1
#endif
#include "Python.h"
#include "pycore_list.h" // _PyList_ITEMS(), _PyList_AppendTakeRef()
#include "pycore_pyatomic_ft_wrappers.h"
#include "clinic/_heapqmodule.c.h"
/*[clinic input]
module _heapq
[clinic start generated code]*/
/*[clinic end generated code: output=da39a3ee5e6b4b0d input=d7cca0a2e4c0ceb3]*/
static int
siftdown(PyListObject *heap, Py_ssize_t startpos, Py_ssize_t pos)
{
PyObject *newitem, *parent, **arr;
Py_ssize_t parentpos, size;
int cmp;
assert(PyList_Check(heap));
size = PyList_GET_SIZE(heap);
if (pos >= size) {
PyErr_SetString(PyExc_IndexError, "index out of range");
return -1;
}
/* Follow the path to the root, moving parents down until finding
a place newitem fits. */
arr = _PyList_ITEMS(heap);
newitem = arr[pos];
while (pos > startpos) {
parentpos = (pos - 1) >> 1;
parent = arr[parentpos];
Py_INCREF(newitem);
Py_INCREF(parent);
cmp = PyObject_RichCompareBool(newitem, parent, Py_LT);
Py_DECREF(parent);
Py_DECREF(newitem);
if (cmp < 0)
return -1;
if (size != PyList_GET_SIZE(heap)) {
PyErr_SetString(PyExc_RuntimeError,
"list changed size during iteration");
return -1;
}
if (cmp == 0)
break;
arr = _PyList_ITEMS(heap);
parent = arr[parentpos];
newitem = arr[pos];
FT_ATOMIC_STORE_PTR_RELAXED(arr[parentpos], newitem);
FT_ATOMIC_STORE_PTR_RELAXED(arr[pos], parent);
pos = parentpos;
}
return 0;
}
static int
siftup(PyListObject *heap, Py_ssize_t pos)
{
Py_ssize_t startpos, endpos, childpos, limit;
PyObject *tmp1, *tmp2, **arr;
int cmp;
assert(PyList_Check(heap));
endpos = PyList_GET_SIZE(heap);
startpos = pos;
if (pos >= endpos) {
PyErr_SetString(PyExc_IndexError, "index out of range");
return -1;
}
/* Bubble up the smaller child until hitting a leaf. */
arr = _PyList_ITEMS(heap);
limit = endpos >> 1; /* smallest pos that has no child */
while (pos < limit) {
/* Set childpos to index of smaller child. */
childpos = 2*pos + 1; /* leftmost child position */
if (childpos + 1 < endpos) {
PyObject* a = arr[childpos];
PyObject* b = arr[childpos + 1];
Py_INCREF(a);
Py_INCREF(b);
cmp = PyObject_RichCompareBool(a, b, Py_LT);
Py_DECREF(a);
Py_DECREF(b);
if (cmp < 0)
return -1;
childpos += ((unsigned)cmp ^ 1); /* increment when cmp==0 */
arr = _PyList_ITEMS(heap); /* arr may have changed */
if (endpos != PyList_GET_SIZE(heap)) {
PyErr_SetString(PyExc_RuntimeError,
"list changed size during iteration");
return -1;
}
}
/* Move the smaller child up. */
tmp1 = arr[childpos];
tmp2 = arr[pos];
FT_ATOMIC_STORE_PTR_RELAXED(arr[childpos], tmp2);
FT_ATOMIC_STORE_PTR_RELAXED(arr[pos], tmp1);
pos = childpos;
}
/* Bubble it up to its final resting place (by sifting its parents down). */
return siftdown(heap, startpos, pos);
}
/*[clinic input]
@critical_section heap
_heapq.heappush
heap: object(subclass_of='&PyList_Type')
item: object
/
Push item onto heap, maintaining the heap invariant.
[clinic start generated code]*/
static PyObject *
_heapq_heappush_impl(PyObject *module, PyObject *heap, PyObject *item)
/*[clinic end generated code: output=912c094f47663935 input=f7a4f03ef8d52e67]*/
{
if (item == NULL) {
PyErr_BadInternalCall();
return NULL;
}
// In a free-threaded build, the heap is locked at this point.
// Therefore, calling _PyList_AppendTakeRef() is safe and no overhead.
if (_PyList_AppendTakeRef((PyListObject *)heap, Py_NewRef(item))) {
return NULL;
}
if (siftdown((PyListObject *)heap, 0, PyList_GET_SIZE(heap)-1)) {
return NULL;
}
Py_RETURN_NONE;
}
static PyObject *
heappop_internal(PyObject *heap, int siftup_func(PyListObject *, Py_ssize_t))
{
PyObject *lastelt, *returnitem;
Py_ssize_t n;
/* raises IndexError if the heap is empty */
n = PyList_GET_SIZE(heap);
if (n == 0) {
PyErr_SetString(PyExc_IndexError, "index out of range");
return NULL;
}
lastelt = PyList_GET_ITEM(heap, n-1) ;
Py_INCREF(lastelt);
if (PyList_SetSlice(heap, n-1, n, NULL)) {
Py_DECREF(lastelt);
return NULL;
}
n--;
if (!n)
return lastelt;
returnitem = PyList_GET_ITEM(heap, 0);
PyListObject *list = _PyList_CAST(heap);
FT_ATOMIC_STORE_PTR_RELAXED(list->ob_item[0], lastelt);
if (siftup_func(list, 0)) {
Py_DECREF(returnitem);
return NULL;
}
return returnitem;
}
/*[clinic input]
@critical_section heap
_heapq.heappop
heap: object(subclass_of='&PyList_Type')
/
Pop the smallest item off the heap, maintaining the heap invariant.
[clinic start generated code]*/
static PyObject *
_heapq_heappop_impl(PyObject *module, PyObject *heap)
/*[clinic end generated code: output=96dfe82d37d9af76 input=ed396461b153dd51]*/
{
return heappop_internal(heap, siftup);
}
static PyObject *
heapreplace_internal(PyObject *heap, PyObject *item, int siftup_func(PyListObject *, Py_ssize_t))
{
PyObject *returnitem;
if (PyList_GET_SIZE(heap) == 0) {
PyErr_SetString(PyExc_IndexError, "index out of range");
return NULL;
}
returnitem = PyList_GET_ITEM(heap, 0);
PyListObject *list = _PyList_CAST(heap);
FT_ATOMIC_STORE_PTR_RELAXED(list->ob_item[0], Py_NewRef(item));
if (siftup_func(list, 0)) {
Py_DECREF(returnitem);
return NULL;
}
return returnitem;
}
/*[clinic input]
@critical_section heap
_heapq.heapreplace
heap: object(subclass_of='&PyList_Type')
item: object
/
Pop and return the current smallest value, and add the new item.
This is more efficient than heappop() followed by heappush(), and can be
more appropriate when using a fixed-size heap. Note that the value
returned may be larger than item! That constrains reasonable uses of
this routine unless written as part of a conditional replacement:
if item > heap[0]:
item = heapreplace(heap, item)
[clinic start generated code]*/
static PyObject *
_heapq_heapreplace_impl(PyObject *module, PyObject *heap, PyObject *item)
/*[clinic end generated code: output=82ea55be8fbe24b4 input=9be1678b817ef1a9]*/
{
return heapreplace_internal(heap, item, siftup);
}
/*[clinic input]
@critical_section heap
_heapq.heappushpop
heap: object(subclass_of='&PyList_Type')
item: object
/
Push item on the heap, then pop and return the smallest item from the heap.
The combined action runs more efficiently than heappush() followed by
a separate call to heappop().
[clinic start generated code]*/
static PyObject *
_heapq_heappushpop_impl(PyObject *module, PyObject *heap, PyObject *item)
/*[clinic end generated code: output=67231dc98ed5774f input=db05c81b1dd92c44]*/
{
PyObject *returnitem;
int cmp;
if (PyList_GET_SIZE(heap) == 0) {
return Py_NewRef(item);
}
PyObject* top = PyList_GET_ITEM(heap, 0);
Py_INCREF(top);
cmp = PyObject_RichCompareBool(top, item, Py_LT);
Py_DECREF(top);
if (cmp < 0)
return NULL;
if (cmp == 0) {
return Py_NewRef(item);
}
if (PyList_GET_SIZE(heap) == 0) {
PyErr_SetString(PyExc_IndexError, "index out of range");
return NULL;
}
returnitem = PyList_GET_ITEM(heap, 0);
PyListObject *list = _PyList_CAST(heap);
FT_ATOMIC_STORE_PTR_RELAXED(list->ob_item[0], Py_NewRef(item));
if (siftup(list, 0)) {
Py_DECREF(returnitem);
return NULL;
}
return returnitem;
}
static Py_ssize_t
keep_top_bit(Py_ssize_t n)
{
int i = 0;
while (n > 1) {
n >>= 1;
i++;
}
return n << i;
}
/* Cache friendly version of heapify()
-----------------------------------
Build-up a heap in O(n) time by performing siftup() operations
on nodes whose children are already heaps.
The simplest way is to sift the nodes in reverse order from
n//2-1 to 0 inclusive. The downside is that children may be
out of cache by the time their parent is reached.
A better way is to not wait for the children to go out of cache.
Once a sibling pair of child nodes have been sifted, immediately
sift their parent node (while the children are still in cache).
Both ways build child heaps before their parents, so both ways
do the exact same number of comparisons and produce exactly
the same heap. The only difference is that the traversal
order is optimized for cache efficiency.
*/
static PyObject *
cache_friendly_heapify(PyObject *heap, int siftup_func(PyListObject *, Py_ssize_t))
{
Py_ssize_t i, j, m, mhalf, leftmost;
m = PyList_GET_SIZE(heap) >> 1; /* index of first childless node */
leftmost = keep_top_bit(m + 1) - 1; /* leftmost node in row of m */
mhalf = m >> 1; /* parent of first childless node */
for (i = leftmost - 1 ; i >= mhalf ; i--) {
j = i;
while (1) {
if (siftup_func((PyListObject *)heap, j))
return NULL;
if (!(j & 1))
break;
j >>= 1;
}
}
for (i = m - 1 ; i >= leftmost ; i--) {
j = i;
while (1) {
if (siftup_func((PyListObject *)heap, j))
return NULL;
if (!(j & 1))
break;
j >>= 1;
}
}
Py_RETURN_NONE;
}
static PyObject *
heapify_internal(PyObject *heap, int siftup_func(PyListObject *, Py_ssize_t))
{
Py_ssize_t i, n;
/* For heaps likely to be bigger than L1 cache, we use the cache
friendly heapify function. For smaller heaps that fit entirely
in cache, we prefer the simpler algorithm with less branching.
*/
n = PyList_GET_SIZE(heap);
if (n > 2500)
return cache_friendly_heapify(heap, siftup_func);
/* Transform bottom-up. The largest index there's any point to
looking at is the largest with a child index in-range, so must
have 2*i + 1 < n, or i < (n-1)/2. If n is even = 2*j, this is
(2*j-1)/2 = j-1/2 so j-1 is the largest, which is n//2 - 1. If
n is odd = 2*j+1, this is (2*j+1-1)/2 = j so j-1 is the largest,
and that's again n//2-1.
*/
for (i = (n >> 1) - 1 ; i >= 0 ; i--)
if (siftup_func((PyListObject *)heap, i))
return NULL;
Py_RETURN_NONE;
}
/*[clinic input]
@critical_section heap
_heapq.heapify
heap: object(subclass_of='&PyList_Type')
/
Transform list into a heap, in-place, in O(len(heap)) time.
[clinic start generated code]*/
static PyObject *
_heapq_heapify_impl(PyObject *module, PyObject *heap)
/*[clinic end generated code: output=e63a636fcf83d6d0 input=aaaaa028b9b6af08]*/
{
return heapify_internal(heap, siftup);
}
static int
siftdown_max(PyListObject *heap, Py_ssize_t startpos, Py_ssize_t pos)
{
PyObject *newitem, *parent, **arr;
Py_ssize_t parentpos, size;
int cmp;
assert(PyList_Check(heap));
size = PyList_GET_SIZE(heap);
if (pos >= size) {
PyErr_SetString(PyExc_IndexError, "index out of range");
return -1;
}
/* Follow the path to the root, moving parents down until finding
a place newitem fits. */
arr = _PyList_ITEMS(heap);
newitem = arr[pos];
while (pos > startpos) {
parentpos = (pos - 1) >> 1;
parent = Py_NewRef(arr[parentpos]);
Py_INCREF(newitem);
cmp = PyObject_RichCompareBool(parent, newitem, Py_LT);
Py_DECREF(parent);
Py_DECREF(newitem);
if (cmp < 0)
return -1;
if (size != PyList_GET_SIZE(heap)) {
PyErr_SetString(PyExc_RuntimeError,
"list changed size during iteration");
return -1;
}
if (cmp == 0)
break;
arr = _PyList_ITEMS(heap);
parent = arr[parentpos];
newitem = arr[pos];
FT_ATOMIC_STORE_PTR_RELAXED(arr[parentpos], newitem);
FT_ATOMIC_STORE_PTR_RELAXED(arr[pos], parent);
pos = parentpos;
}
return 0;
}
static int
siftup_max(PyListObject *heap, Py_ssize_t pos)
{
Py_ssize_t startpos, endpos, childpos, limit;
PyObject *tmp1, *tmp2, **arr;
int cmp;
assert(PyList_Check(heap));
endpos = PyList_GET_SIZE(heap);
startpos = pos;
if (pos >= endpos) {
PyErr_SetString(PyExc_IndexError, "index out of range");
return -1;
}
/* Bubble up the larger child until hitting a leaf. */
arr = _PyList_ITEMS(heap);
limit = endpos >> 1; /* smallest pos that has no child */
while (pos < limit) {
/* Set childpos to index of larger child. */
childpos = 2*pos + 1; /* leftmost child position */
if (childpos + 1 < endpos) {
PyObject* a = arr[childpos + 1];
PyObject* b = arr[childpos];
Py_INCREF(a);
Py_INCREF(b);
cmp = PyObject_RichCompareBool(a, b, Py_LT);
Py_DECREF(a);
Py_DECREF(b);
if (cmp < 0)
return -1;
childpos += ((unsigned)cmp ^ 1); /* increment when cmp==0 */
arr = _PyList_ITEMS(heap); /* arr may have changed */
if (endpos != PyList_GET_SIZE(heap)) {
PyErr_SetString(PyExc_RuntimeError,
"list changed size during iteration");
return -1;
}
}
/* Move the larger child up. */
tmp1 = arr[childpos];
tmp2 = arr[pos];
FT_ATOMIC_STORE_PTR_RELAXED(arr[childpos], tmp2);
FT_ATOMIC_STORE_PTR_RELAXED(arr[pos], tmp1);
pos = childpos;
}
/* Bubble it up to its final resting place (by sifting its parents down). */
return siftdown_max(heap, startpos, pos);
}
/*[clinic input]
@critical_section heap
_heapq.heappush_max
heap: object(subclass_of='&PyList_Type')
item: object
/
Push item onto max heap, maintaining the heap invariant.
[clinic start generated code]*/
static PyObject *
_heapq_heappush_max_impl(PyObject *module, PyObject *heap, PyObject *item)
/*[clinic end generated code: output=c869d5f9deb08277 input=c437e3d1ff8dcb70]*/
{
if (item == NULL) {
PyErr_BadInternalCall();
return NULL;
}
// In a free-threaded build, the heap is locked at this point.
// Therefore, calling _PyList_AppendTakeRef() is safe and no overhead.
if (_PyList_AppendTakeRef((PyListObject *)heap, Py_NewRef(item))) {
return NULL;
}
if (siftdown_max((PyListObject *)heap, 0, PyList_GET_SIZE(heap)-1)) {
return NULL;
}
Py_RETURN_NONE;
}
/*[clinic input]
@critical_section heap
_heapq.heappop_max
heap: object(subclass_of='&PyList_Type')
/
Maxheap variant of heappop.
[clinic start generated code]*/
static PyObject *
_heapq_heappop_max_impl(PyObject *module, PyObject *heap)
/*[clinic end generated code: output=2f051195ab404b77 input=5d70c997798aec64]*/
{
return heappop_internal(heap, siftup_max);
}
/*[clinic input]
@critical_section heap
_heapq.heapreplace_max
heap: object(subclass_of='&PyList_Type')
item: object
/
Maxheap variant of heapreplace.
[clinic start generated code]*/
static PyObject *
_heapq_heapreplace_max_impl(PyObject *module, PyObject *heap, PyObject *item)
/*[clinic end generated code: output=8770778b5a9cbe9b input=fe70175356e4a649]*/
{
return heapreplace_internal(heap, item, siftup_max);
}
/*[clinic input]
@critical_section heap
_heapq.heapify_max
heap: object(subclass_of='&PyList_Type')
/
Maxheap variant of heapify.
[clinic start generated code]*/
static PyObject *
_heapq_heapify_max_impl(PyObject *module, PyObject *heap)
/*[clinic end generated code: output=8401af3856529807 input=4eee63231e7d1573]*/
{
return heapify_internal(heap, siftup_max);
}
/*[clinic input]
@critical_section heap
_heapq.heappushpop_max
heap: object(subclass_of='&PyList_Type')
item: object
/
Maxheap variant of heappushpop.
The combined action runs more efficiently than heappush_max()
followed by a separate call to heappop_max().
[clinic start generated code]*/
static PyObject *
_heapq_heappushpop_max_impl(PyObject *module, PyObject *heap, PyObject *item)
/*[clinic end generated code: output=ff0019f0941aca0d input=52030929667a4c08]*/
{
PyObject *returnitem;
int cmp;
if (PyList_GET_SIZE(heap) == 0) {
return Py_NewRef(item);
}
PyObject *top = PyList_GET_ITEM(heap, 0);
Py_INCREF(top);
cmp = PyObject_RichCompareBool(item, top, Py_LT);
Py_DECREF(top);
if (cmp < 0) {
return NULL;
}
if (cmp == 0) {
return Py_NewRef(item);
}
if (PyList_GET_SIZE(heap) == 0) {
PyErr_SetString(PyExc_IndexError, "index out of range");
return NULL;
}
returnitem = PyList_GET_ITEM(heap, 0);
PyListObject *list = _PyList_CAST(heap);
FT_ATOMIC_STORE_PTR_RELAXED(list->ob_item[0], Py_NewRef(item));
if (siftup_max(list, 0) < 0) {
Py_DECREF(returnitem);
return NULL;
}
return returnitem;
}
static PyMethodDef heapq_methods[] = {
_HEAPQ_HEAPPUSH_METHODDEF
_HEAPQ_HEAPPUSHPOP_METHODDEF
_HEAPQ_HEAPPOP_METHODDEF
_HEAPQ_HEAPREPLACE_METHODDEF
_HEAPQ_HEAPIFY_METHODDEF
_HEAPQ_HEAPPUSH_MAX_METHODDEF
_HEAPQ_HEAPPUSHPOP_MAX_METHODDEF
_HEAPQ_HEAPPOP_MAX_METHODDEF
_HEAPQ_HEAPREPLACE_MAX_METHODDEF
_HEAPQ_HEAPIFY_MAX_METHODDEF
{NULL, NULL} /* sentinel */
};
PyDoc_STRVAR(module_doc,
"Heap queue algorithm (a.k.a. priority queue).\n\
\n\
Heaps are arrays for which a[k] <= a[2*k+1] and a[k] <= a[2*k+2] for\n\
all k, counting elements from 0. For the sake of comparison,\n\
non-existing elements are considered to be infinite. The interesting\n\
property of a heap is that a[0] is always its smallest element.\n\
\n\
Usage:\n\
\n\
heap = [] # creates an empty heap\n\
heappush(heap, item) # pushes a new item on the heap\n\
item = heappop(heap) # pops the smallest item from the heap\n\
item = heap[0] # smallest item on the heap without popping it\n\
heapify(x) # transforms list into a heap, in-place, in linear time\n\
item = heapreplace(heap, item) # pops and returns smallest item, and adds\n\
# new item; the heap size is unchanged\n\
\n\
Our API differs from textbook heap algorithms as follows:\n\
\n\
- We use 0-based indexing. This makes the relationship between the\n\
index for a node and the indexes for its children slightly less\n\
obvious, but is more suitable since Python uses 0-based indexing.\n\
\n\
- Our heappop() method returns the smallest item, not the largest.\n\
\n\
These two make it possible to view the heap as a regular Python list\n\
without surprises: heap[0] is the smallest item, and heap.sort()\n\
maintains the heap invariant!\n");
PyDoc_STRVAR(__about__,
"Heap queues\n\
\n\
[explanation by Fran\xc3\xa7ois Pinard]\n\
\n\
Heaps are arrays for which a[k] <= a[2*k+1] and a[k] <= a[2*k+2] for\n\
all k, counting elements from 0. For the sake of comparison,\n\
non-existing elements are considered to be infinite. The interesting\n\
property of a heap is that a[0] is always its smallest element.\n"
"\n\
The strange invariant above is meant to be an efficient memory\n\
representation for a tournament. The numbers below are 'k', not a[k]:\n\
\n\
0\n\
\n\
1 2\n\
\n\
3 4 5 6\n\
\n\
7 8 9 10 11 12 13 14\n\
\n\
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30\n\
\n\
\n\
In the tree above, each cell 'k' is topping '2*k+1' and '2*k+2'. In\n\
a usual binary tournament we see in sports, each cell is the winner\n\
over the two cells it tops, and we can trace the winner down the tree\n\
to see all opponents s/he had. However, in many computer applications\n\
of such tournaments, we do not need to trace the history of a winner.\n\
To be more memory efficient, when a winner is promoted, we try to\n\
replace it by something else at a lower level, and the rule becomes\n\
that a cell and the two cells it tops contain three different items,\n\
but the top cell \"wins\" over the two topped cells.\n"
"\n\
If this heap invariant is protected at all time, index 0 is clearly\n\
the overall winner. The simplest algorithmic way to remove it and\n\
find the \"next\" winner is to move some loser (let's say cell 30 in the\n\
diagram above) into the 0 position, and then percolate this new 0 down\n\
the tree, exchanging values, until the invariant is re-established.\n\
This is clearly logarithmic on the total number of items in the tree.\n\
By iterating over all items, you get an O(n ln n) sort.\n"
"\n\
A nice feature of this sort is that you can efficiently insert new\n\
items while the sort is going on, provided that the inserted items are\n\
not \"better\" than the last 0'th element you extracted. This is\n\
especially useful in simulation contexts, where the tree holds all\n\
incoming events, and the \"win\" condition means the smallest scheduled\n\
time. When an event schedule other events for execution, they are\n\
scheduled into the future, so they can easily go into the heap. So, a\n\
heap is a good structure for implementing schedulers (this is what I\n\
used for my MIDI sequencer :-).\n"
"\n\
Various structures for implementing schedulers have been extensively\n\
studied, and heaps are good for this, as they are reasonably speedy,\n\
the speed is almost constant, and the worst case is not much different\n\
than the average case. However, there are other representations which\n\
are more efficient overall, yet the worst cases might be terrible.\n"
"\n\
Heaps are also very useful in big disk sorts. You most probably all\n\
know that a big sort implies producing \"runs\" (which are pre-sorted\n\
sequences, which size is usually related to the amount of CPU memory),\n\
followed by a merging passes for these runs, which merging is often\n\
very cleverly organised[1]. It is very important that the initial\n\
sort produces the longest runs possible. Tournaments are a good way\n\
to that. If, using all the memory available to hold a tournament, you\n\
replace and percolate items that happen to fit the current run, you'll\n\
produce runs which are twice the size of the memory for random input,\n\
and much better for input fuzzily ordered.\n"
"\n\
Moreover, if you output the 0'th item on disk and get an input which\n\
may not fit in the current tournament (because the value \"wins\" over\n\
the last output value), it cannot fit in the heap, so the size of the\n\
heap decreases. The freed memory could be cleverly reused immediately\n\
for progressively building a second heap, which grows at exactly the\n\
same rate the first heap is melting. When the first heap completely\n\
vanishes, you switch heaps and start a new run. Clever and quite\n\
effective!\n\
\n\
In a word, heaps are useful memory structures to know. I use them in\n\
a few applications, and I think it is good to keep a 'heap' module\n\
around. :-)\n"
"\n\
--------------------\n\
[1] The disk balancing algorithms which are current, nowadays, are\n\
more annoying than clever, and this is a consequence of the seeking\n\
capabilities of the disks. On devices which cannot seek, like big\n\
tape drives, the story was quite different, and one had to be very\n\
clever to ensure (far in advance) that each tape movement will be the\n\
most effective possible (that is, will best participate at\n\
\"progressing\" the merge). Some tapes were even able to read\n\
backwards, and this was also used to avoid the rewinding time.\n\
Believe me, real good tape sorts were quite spectacular to watch!\n\
From all times, sorting has always been a Great Art! :-)\n");
static int
heapq_exec(PyObject *m)
{
if (PyModule_Add(m, "__about__", PyUnicode_FromString(__about__)) < 0) {
return -1;
}
return 0;
}
static struct PyModuleDef_Slot heapq_slots[] = {
{Py_mod_exec, heapq_exec},
{Py_mod_multiple_interpreters, Py_MOD_PER_INTERPRETER_GIL_SUPPORTED},
{Py_mod_gil, Py_MOD_GIL_NOT_USED},
{0, NULL}
};
static struct PyModuleDef _heapqmodule = {
PyModuleDef_HEAD_INIT,
"_heapq",
module_doc,
0,
heapq_methods,
heapq_slots,
NULL,
NULL,
NULL
};
PyMODINIT_FUNC
PyInit__heapq(void)
{
return PyModuleDef_Init(&_heapqmodule);
}
You can’t perform that action at this time.