Commit | Line | Data |
84d4ea48 |
1 | /* pp_sort.c |
2 | * |
be3c0a43 |
3 | * Copyright (c) 1991-2002, Larry Wall |
84d4ea48 |
4 | * |
5 | * You may distribute under the terms of either the GNU General Public |
6 | * License or the Artistic License, as specified in the README file. |
7 | * |
8 | */ |
9 | |
10 | /* |
11 | * ...they shuffled back towards the rear of the line. 'No, not at the |
12 | * rear!' the slave-driver shouted. 'Three files up. And stay there... |
13 | */ |
14 | |
15 | #include "EXTERN.h" |
16 | #define PERL_IN_PP_SORT_C |
17 | #include "perl.h" |
18 | |
19 | static I32 sortcv(pTHX_ SV *a, SV *b); |
20 | static I32 sortcv_stacked(pTHX_ SV *a, SV *b); |
21 | static I32 sortcv_xsub(pTHX_ SV *a, SV *b); |
22 | static I32 sv_ncmp(pTHX_ SV *a, SV *b); |
23 | static I32 sv_i_ncmp(pTHX_ SV *a, SV *b); |
24 | static I32 amagic_ncmp(pTHX_ SV *a, SV *b); |
25 | static I32 amagic_i_ncmp(pTHX_ SV *a, SV *b); |
26 | static I32 amagic_cmp(pTHX_ SV *a, SV *b); |
27 | static I32 amagic_cmp_locale(pTHX_ SV *a, SV *b); |
28 | |
29 | #define sv_cmp_static Perl_sv_cmp |
30 | #define sv_cmp_locale_static Perl_sv_cmp_locale |
31 | |
32 | #define SORTHINTS(hintsvp) \ |
33 | ((PL_hintgv && \ |
34 | (hintsvp = hv_fetch(GvHV(PL_hintgv), "SORT", 4, FALSE))) ? \ |
35 | (I32)SvIV(*hintsvp) : 0) |
36 | |
c53fc8a6 |
37 | #ifndef SMALLSORT |
38 | #define SMALLSORT (200) |
39 | #endif |
40 | |
84d4ea48 |
41 | /* |
42 | * The mergesort implementation is by Peter M. Mcilroy <pmcilroy@lucent.com>. |
43 | * |
44 | * The original code was written in conjunction with BSD Computer Software |
45 | * Research Group at University of California, Berkeley. |
46 | * |
47 | * See also: "Optimistic Merge Sort" (SODA '92) |
48 | * |
49 | * The integration to Perl is by John P. Linderman <jpl@research.att.com>. |
50 | * |
51 | * The code can be distributed under the same terms as Perl itself. |
52 | * |
53 | */ |
54 | |
84d4ea48 |
55 | |
56 | typedef char * aptr; /* pointer for arithmetic on sizes */ |
57 | typedef SV * gptr; /* pointers in our lists */ |
58 | |
59 | /* Binary merge internal sort, with a few special mods |
60 | ** for the special perl environment it now finds itself in. |
61 | ** |
62 | ** Things that were once options have been hotwired |
63 | ** to values suitable for this use. In particular, we'll always |
64 | ** initialize looking for natural runs, we'll always produce stable |
65 | ** output, and we'll always do Peter McIlroy's binary merge. |
66 | */ |
67 | |
68 | /* Pointer types for arithmetic and storage and convenience casts */ |
69 | |
70 | #define APTR(P) ((aptr)(P)) |
71 | #define GPTP(P) ((gptr *)(P)) |
72 | #define GPPP(P) ((gptr **)(P)) |
73 | |
74 | |
75 | /* byte offset from pointer P to (larger) pointer Q */ |
76 | #define BYTEOFF(P, Q) (APTR(Q) - APTR(P)) |
77 | |
78 | #define PSIZE sizeof(gptr) |
79 | |
80 | /* If PSIZE is power of 2, make PSHIFT that power, if that helps */ |
81 | |
82 | #ifdef PSHIFT |
83 | #define PNELEM(P, Q) (BYTEOFF(P,Q) >> (PSHIFT)) |
84 | #define PNBYTE(N) ((N) << (PSHIFT)) |
85 | #define PINDEX(P, N) (GPTP(APTR(P) + PNBYTE(N))) |
86 | #else |
87 | /* Leave optimization to compiler */ |
88 | #define PNELEM(P, Q) (GPTP(Q) - GPTP(P)) |
89 | #define PNBYTE(N) ((N) * (PSIZE)) |
90 | #define PINDEX(P, N) (GPTP(P) + (N)) |
91 | #endif |
92 | |
93 | /* Pointer into other corresponding to pointer into this */ |
94 | #define POTHER(P, THIS, OTHER) GPTP(APTR(OTHER) + BYTEOFF(THIS,P)) |
95 | |
96 | #define FROMTOUPTO(src, dst, lim) do *dst++ = *src++; while(src<lim) |
97 | |
98 | |
99 | /* Runs are identified by a pointer in the auxilliary list. |
100 | ** The pointer is at the start of the list, |
101 | ** and it points to the start of the next list. |
102 | ** NEXT is used as an lvalue, too. |
103 | */ |
104 | |
105 | #define NEXT(P) (*GPPP(P)) |
106 | |
107 | |
108 | /* PTHRESH is the minimum number of pairs with the same sense to justify |
109 | ** checking for a run and extending it. Note that PTHRESH counts PAIRS, |
110 | ** not just elements, so PTHRESH == 8 means a run of 16. |
111 | */ |
112 | |
113 | #define PTHRESH (8) |
114 | |
115 | /* RTHRESH is the number of elements in a run that must compare low |
116 | ** to the low element from the opposing run before we justify |
117 | ** doing a binary rampup instead of single stepping. |
118 | ** In random input, N in a row low should only happen with |
119 | ** probability 2^(1-N), so we can risk that we are dealing |
120 | ** with orderly input without paying much when we aren't. |
121 | */ |
122 | |
123 | #define RTHRESH (6) |
124 | |
125 | |
126 | /* |
127 | ** Overview of algorithm and variables. |
128 | ** The array of elements at list1 will be organized into runs of length 2, |
129 | ** or runs of length >= 2 * PTHRESH. We only try to form long runs when |
130 | ** PTHRESH adjacent pairs compare in the same way, suggesting overall order. |
131 | ** |
132 | ** Unless otherwise specified, pair pointers address the first of two elements. |
133 | ** |
134 | ** b and b+1 are a pair that compare with sense ``sense''. |
135 | ** b is the ``bottom'' of adjacent pairs that might form a longer run. |
136 | ** |
137 | ** p2 parallels b in the list2 array, where runs are defined by |
138 | ** a pointer chain. |
139 | ** |
140 | ** t represents the ``top'' of the adjacent pairs that might extend |
141 | ** the run beginning at b. Usually, t addresses a pair |
142 | ** that compares with opposite sense from (b,b+1). |
143 | ** However, it may also address a singleton element at the end of list1, |
144 | ** or it may be equal to ``last'', the first element beyond list1. |
145 | ** |
146 | ** r addresses the Nth pair following b. If this would be beyond t, |
147 | ** we back it off to t. Only when r is less than t do we consider the |
148 | ** run long enough to consider checking. |
149 | ** |
150 | ** q addresses a pair such that the pairs at b through q already form a run. |
151 | ** Often, q will equal b, indicating we only are sure of the pair itself. |
152 | ** However, a search on the previous cycle may have revealed a longer run, |
153 | ** so q may be greater than b. |
154 | ** |
155 | ** p is used to work back from a candidate r, trying to reach q, |
156 | ** which would mean b through r would be a run. If we discover such a run, |
157 | ** we start q at r and try to push it further towards t. |
158 | ** If b through r is NOT a run, we detect the wrong order at (p-1,p). |
159 | ** In any event, after the check (if any), we have two main cases. |
160 | ** |
161 | ** 1) Short run. b <= q < p <= r <= t. |
162 | ** b through q is a run (perhaps trivial) |
163 | ** q through p are uninteresting pairs |
164 | ** p through r is a run |
165 | ** |
166 | ** 2) Long run. b < r <= q < t. |
167 | ** b through q is a run (of length >= 2 * PTHRESH) |
168 | ** |
169 | ** Note that degenerate cases are not only possible, but likely. |
170 | ** For example, if the pair following b compares with opposite sense, |
171 | ** then b == q < p == r == t. |
172 | */ |
173 | |
174 | |
957d8989 |
175 | static IV |
84d4ea48 |
176 | dynprep(pTHX_ gptr *list1, gptr *list2, size_t nmemb, SVCOMPARE_t cmp) |
177 | { |
957d8989 |
178 | I32 sense; |
84d4ea48 |
179 | register gptr *b, *p, *q, *t, *p2; |
180 | register gptr c, *last, *r; |
181 | gptr *savep; |
957d8989 |
182 | IV runs = 0; |
84d4ea48 |
183 | |
184 | b = list1; |
185 | last = PINDEX(b, nmemb); |
186 | sense = (cmp(aTHX_ *b, *(b+1)) > 0); |
187 | for (p2 = list2; b < last; ) { |
188 | /* We just started, or just reversed sense. |
189 | ** Set t at end of pairs with the prevailing sense. |
190 | */ |
191 | for (p = b+2, t = p; ++p < last; t = ++p) { |
192 | if ((cmp(aTHX_ *t, *p) > 0) != sense) break; |
193 | } |
194 | q = b; |
195 | /* Having laid out the playing field, look for long runs */ |
196 | do { |
197 | p = r = b + (2 * PTHRESH); |
198 | if (r >= t) p = r = t; /* too short to care about */ |
199 | else { |
200 | while (((cmp(aTHX_ *(p-1), *p) > 0) == sense) && |
201 | ((p -= 2) > q)); |
202 | if (p <= q) { |
203 | /* b through r is a (long) run. |
204 | ** Extend it as far as possible. |
205 | */ |
206 | p = q = r; |
207 | while (((p += 2) < t) && |
208 | ((cmp(aTHX_ *(p-1), *p) > 0) == sense)) q = p; |
209 | r = p = q + 2; /* no simple pairs, no after-run */ |
210 | } |
211 | } |
212 | if (q > b) { /* run of greater than 2 at b */ |
213 | savep = p; |
214 | p = q += 2; |
215 | /* pick up singleton, if possible */ |
216 | if ((p == t) && |
217 | ((t + 1) == last) && |
218 | ((cmp(aTHX_ *(p-1), *p) > 0) == sense)) |
219 | savep = r = p = q = last; |
957d8989 |
220 | p2 = NEXT(p2) = p2 + (p - b); ++runs; |
84d4ea48 |
221 | if (sense) while (b < --p) { |
222 | c = *b; |
223 | *b++ = *p; |
224 | *p = c; |
225 | } |
226 | p = savep; |
227 | } |
228 | while (q < p) { /* simple pairs */ |
957d8989 |
229 | p2 = NEXT(p2) = p2 + 2; ++runs; |
84d4ea48 |
230 | if (sense) { |
231 | c = *q++; |
232 | *(q-1) = *q; |
233 | *q++ = c; |
234 | } else q += 2; |
235 | } |
236 | if (((b = p) == t) && ((t+1) == last)) { |
957d8989 |
237 | NEXT(p2) = p2 + 1; ++runs; |
84d4ea48 |
238 | b++; |
239 | } |
240 | q = r; |
241 | } while (b < t); |
242 | sense = !sense; |
243 | } |
957d8989 |
244 | return runs; |
84d4ea48 |
245 | } |
246 | |
247 | |
3fe0b9a9 |
248 | /* The original merge sort, in use since 5.7, was as fast as, or faster than, |
957d8989 |
249 | * qsort on many platforms, but slower than qsort, conspicuously so, |
3fe0b9a9 |
250 | * on others. The most likely explanation was platform-specific |
957d8989 |
251 | * differences in cache sizes and relative speeds. |
252 | * |
253 | * The quicksort divide-and-conquer algorithm guarantees that, as the |
254 | * problem is subdivided into smaller and smaller parts, the parts |
255 | * fit into smaller (and faster) caches. So it doesn't matter how |
256 | * many levels of cache exist, quicksort will "find" them, and, |
257 | * as long as smaller is faster, take advanatge of them. |
258 | * |
3fe0b9a9 |
259 | * By contrast, consider how the original mergesort algorithm worked. |
957d8989 |
260 | * Suppose we have five runs (each typically of length 2 after dynprep). |
261 | * |
262 | * pass base aux |
263 | * 0 1 2 3 4 5 |
264 | * 1 12 34 5 |
265 | * 2 1234 5 |
266 | * 3 12345 |
267 | * 4 12345 |
268 | * |
269 | * Adjacent pairs are merged in "grand sweeps" through the input. |
270 | * This means, on pass 1, the records in runs 1 and 2 aren't revisited until |
271 | * runs 3 and 4 are merged and the runs from run 5 have been copied. |
272 | * The only cache that matters is one large enough to hold *all* the input. |
273 | * On some platforms, this may be many times slower than smaller caches. |
274 | * |
275 | * The following pseudo-code uses the same basic merge algorithm, |
276 | * but in a divide-and-conquer way. |
277 | * |
278 | * # merge $runs runs at offset $offset of list $list1 into $list2. |
279 | * # all unmerged runs ($runs == 1) originate in list $base. |
280 | * sub mgsort2 { |
281 | * my ($offset, $runs, $base, $list1, $list2) = @_; |
282 | * |
283 | * if ($runs == 1) { |
284 | * if ($list1 is $base) copy run to $list2 |
285 | * return offset of end of list (or copy) |
286 | * } else { |
287 | * $off2 = mgsort2($offset, $runs-($runs/2), $base, $list2, $list1) |
288 | * mgsort2($off2, $runs/2, $base, $list2, $list1) |
289 | * merge the adjacent runs at $offset of $list1 into $list2 |
290 | * return the offset of the end of the merged runs |
291 | * } |
292 | * } |
293 | * mgsort2(0, $runs, $base, $aux, $base); |
294 | * |
295 | * For our 5 runs, the tree of calls looks like |
296 | * |
297 | * 5 |
298 | * 3 2 |
299 | * 2 1 1 1 |
300 | * 1 1 |
301 | * |
302 | * 1 2 3 4 5 |
303 | * |
304 | * and the corresponding activity looks like |
305 | * |
306 | * copy runs 1 and 2 from base to aux |
307 | * merge runs 1 and 2 from aux to base |
308 | * (run 3 is where it belongs, no copy needed) |
309 | * merge runs 12 and 3 from base to aux |
310 | * (runs 4 and 5 are where they belong, no copy needed) |
311 | * merge runs 4 and 5 from base to aux |
312 | * merge runs 123 and 45 from aux to base |
313 | * |
314 | * Note that we merge runs 1 and 2 immediately after copying them, |
315 | * while they are still likely to be in fast cache. Similarly, |
316 | * run 3 is merged with run 12 while it still may be lingering in cache. |
317 | * This implementation should therefore enjoy much of the cache-friendly |
318 | * behavior that quicksort does. In addition, it does less copying |
319 | * than the original mergesort implementation (only runs 1 and 2 are copied) |
320 | * and the "balancing" of merges is better (merged runs comprise more nearly |
321 | * equal numbers of original runs). |
322 | * |
323 | * The actual cache-friendly implementation will use a pseudo-stack |
324 | * to avoid recursion, and will unroll processing of runs of length 2, |
325 | * but it is otherwise similar to the recursive implementation. |
957d8989 |
326 | */ |
327 | |
328 | typedef struct { |
329 | IV offset; /* offset of 1st of 2 runs at this level */ |
330 | IV runs; /* how many runs must be combined into 1 */ |
331 | } off_runs; /* pseudo-stack element */ |
332 | |
333 | STATIC void |
3fe0b9a9 |
334 | S_mergesortsv(pTHX_ gptr *base, size_t nmemb, SVCOMPARE_t cmp) |
957d8989 |
335 | { |
336 | IV i, run, runs, offset; |
337 | I32 sense, level; |
338 | int iwhich; |
339 | register gptr *f1, *f2, *t, *b, *p, *tp2, *l1, *l2, *q; |
340 | gptr *aux, *list1, *list2; |
341 | gptr *p1; |
342 | gptr small[SMALLSORT]; |
343 | gptr *which[3]; |
344 | off_runs stack[60], *stackp; |
345 | |
346 | if (nmemb <= 1) return; /* sorted trivially */ |
347 | if (nmemb <= SMALLSORT) aux = small; /* use stack for aux array */ |
348 | else { New(799,aux,nmemb,gptr); } /* allocate auxilliary array */ |
349 | level = 0; |
350 | stackp = stack; |
351 | stackp->runs = dynprep(aTHX_ base, aux, nmemb, cmp); |
352 | stackp->offset = offset = 0; |
353 | which[0] = which[2] = base; |
354 | which[1] = aux; |
355 | for (;;) { |
356 | /* On levels where both runs have be constructed (stackp->runs == 0), |
357 | * merge them, and note the offset of their end, in case the offset |
358 | * is needed at the next level up. Hop up a level, and, |
359 | * as long as stackp->runs is 0, keep merging. |
360 | */ |
361 | if ((runs = stackp->runs) == 0) { |
362 | iwhich = level & 1; |
363 | list1 = which[iwhich]; /* area where runs are now */ |
364 | list2 = which[++iwhich]; /* area for merged runs */ |
365 | do { |
366 | offset = stackp->offset; |
367 | f1 = p1 = list1 + offset; /* start of first run */ |
368 | p = tp2 = list2 + offset; /* where merged run will go */ |
369 | t = NEXT(p); /* where first run ends */ |
370 | f2 = l1 = POTHER(t, list2, list1); /* ... on the other side */ |
371 | t = NEXT(t); /* where second runs ends */ |
372 | l2 = POTHER(t, list2, list1); /* ... on the other side */ |
373 | offset = PNELEM(list2, t); |
374 | while (f1 < l1 && f2 < l2) { |
375 | /* If head 1 is larger than head 2, find ALL the elements |
376 | ** in list 2 strictly less than head1, write them all, |
377 | ** then head 1. Then compare the new heads, and repeat, |
378 | ** until one or both lists are exhausted. |
379 | ** |
380 | ** In all comparisons (after establishing |
381 | ** which head to merge) the item to merge |
382 | ** (at pointer q) is the first operand of |
383 | ** the comparison. When we want to know |
384 | ** if ``q is strictly less than the other'', |
385 | ** we can't just do |
386 | ** cmp(q, other) < 0 |
387 | ** because stability demands that we treat equality |
388 | ** as high when q comes from l2, and as low when |
389 | ** q was from l1. So we ask the question by doing |
390 | ** cmp(q, other) <= sense |
391 | ** and make sense == 0 when equality should look low, |
392 | ** and -1 when equality should look high. |
393 | */ |
394 | |
395 | |
396 | if (cmp(aTHX_ *f1, *f2) <= 0) { |
397 | q = f2; b = f1; t = l1; |
398 | sense = -1; |
399 | } else { |
400 | q = f1; b = f2; t = l2; |
401 | sense = 0; |
402 | } |
403 | |
404 | |
405 | /* ramp up |
406 | ** |
407 | ** Leave t at something strictly |
408 | ** greater than q (or at the end of the list), |
409 | ** and b at something strictly less than q. |
410 | */ |
411 | for (i = 1, run = 0 ;;) { |
412 | if ((p = PINDEX(b, i)) >= t) { |
413 | /* off the end */ |
414 | if (((p = PINDEX(t, -1)) > b) && |
415 | (cmp(aTHX_ *q, *p) <= sense)) |
416 | t = p; |
417 | else b = p; |
418 | break; |
419 | } else if (cmp(aTHX_ *q, *p) <= sense) { |
420 | t = p; |
421 | break; |
422 | } else b = p; |
423 | if (++run >= RTHRESH) i += i; |
424 | } |
425 | |
426 | |
427 | /* q is known to follow b and must be inserted before t. |
428 | ** Increment b, so the range of possibilities is [b,t). |
429 | ** Round binary split down, to favor early appearance. |
430 | ** Adjust b and t until q belongs just before t. |
431 | */ |
432 | |
433 | b++; |
434 | while (b < t) { |
435 | p = PINDEX(b, (PNELEM(b, t) - 1) / 2); |
436 | if (cmp(aTHX_ *q, *p) <= sense) { |
437 | t = p; |
438 | } else b = p + 1; |
439 | } |
440 | |
441 | |
442 | /* Copy all the strictly low elements */ |
443 | |
444 | if (q == f1) { |
445 | FROMTOUPTO(f2, tp2, t); |
446 | *tp2++ = *f1++; |
447 | } else { |
448 | FROMTOUPTO(f1, tp2, t); |
449 | *tp2++ = *f2++; |
450 | } |
451 | } |
452 | |
453 | |
454 | /* Run out remaining list */ |
455 | if (f1 == l1) { |
456 | if (f2 < l2) FROMTOUPTO(f2, tp2, l2); |
457 | } else FROMTOUPTO(f1, tp2, l1); |
458 | p1 = NEXT(p1) = POTHER(tp2, list2, list1); |
459 | |
460 | if (--level == 0) goto done; |
461 | --stackp; |
462 | t = list1; list1 = list2; list2 = t; /* swap lists */ |
463 | } while ((runs = stackp->runs) == 0); |
464 | } |
465 | |
466 | |
467 | stackp->runs = 0; /* current run will finish level */ |
468 | /* While there are more than 2 runs remaining, |
469 | * turn them into exactly 2 runs (at the "other" level), |
470 | * each made up of approximately half the runs. |
471 | * Stack the second half for later processing, |
472 | * and set about producing the first half now. |
473 | */ |
474 | while (runs > 2) { |
475 | ++level; |
476 | ++stackp; |
477 | stackp->offset = offset; |
478 | runs -= stackp->runs = runs / 2; |
479 | } |
480 | /* We must construct a single run from 1 or 2 runs. |
481 | * All the original runs are in which[0] == base. |
482 | * The run we construct must end up in which[level&1]. |
483 | */ |
484 | iwhich = level & 1; |
485 | if (runs == 1) { |
486 | /* Constructing a single run from a single run. |
487 | * If it's where it belongs already, there's nothing to do. |
488 | * Otherwise, copy it to where it belongs. |
489 | * A run of 1 is either a singleton at level 0, |
490 | * or the second half of a split 3. In neither event |
491 | * is it necessary to set offset. It will be set by the merge |
492 | * that immediately follows. |
493 | */ |
494 | if (iwhich) { /* Belongs in aux, currently in base */ |
495 | f1 = b = PINDEX(base, offset); /* where list starts */ |
496 | f2 = PINDEX(aux, offset); /* where list goes */ |
497 | t = NEXT(f2); /* where list will end */ |
498 | offset = PNELEM(aux, t); /* offset thereof */ |
499 | t = PINDEX(base, offset); /* where it currently ends */ |
500 | FROMTOUPTO(f1, f2, t); /* copy */ |
501 | NEXT(b) = t; /* set up parallel pointer */ |
502 | } else if (level == 0) goto done; /* single run at level 0 */ |
503 | } else { |
504 | /* Constructing a single run from two runs. |
505 | * The merge code at the top will do that. |
506 | * We need only make sure the two runs are in the "other" array, |
507 | * so they'll end up in the correct array after the merge. |
508 | */ |
509 | ++level; |
510 | ++stackp; |
511 | stackp->offset = offset; |
512 | stackp->runs = 0; /* take care of both runs, trigger merge */ |
513 | if (!iwhich) { /* Merged runs belong in aux, copy 1st */ |
514 | f1 = b = PINDEX(base, offset); /* where first run starts */ |
515 | f2 = PINDEX(aux, offset); /* where it will be copied */ |
516 | t = NEXT(f2); /* where first run will end */ |
517 | offset = PNELEM(aux, t); /* offset thereof */ |
518 | p = PINDEX(base, offset); /* end of first run */ |
519 | t = NEXT(t); /* where second run will end */ |
520 | t = PINDEX(base, PNELEM(aux, t)); /* where it now ends */ |
521 | FROMTOUPTO(f1, f2, t); /* copy both runs */ |
522 | NEXT(b) = p; /* paralled pointer for 1st */ |
523 | NEXT(p) = t; /* ... and for second */ |
524 | } |
525 | } |
526 | } |
527 | done: |
528 | if (aux != small) Safefree(aux); /* free iff allocated */ |
529 | return; |
530 | } |
531 | |
84d4ea48 |
532 | /* |
533 | * The quicksort implementation was derived from source code contributed |
534 | * by Tom Horsley. |
535 | * |
536 | * NOTE: this code was derived from Tom Horsley's qsort replacement |
537 | * and should not be confused with the original code. |
538 | */ |
539 | |
540 | /* Copyright (C) Tom Horsley, 1997. All rights reserved. |
541 | |
542 | Permission granted to distribute under the same terms as perl which are |
543 | (briefly): |
544 | |
545 | This program is free software; you can redistribute it and/or modify |
546 | it under the terms of either: |
547 | |
548 | a) the GNU General Public License as published by the Free |
549 | Software Foundation; either version 1, or (at your option) any |
550 | later version, or |
551 | |
552 | b) the "Artistic License" which comes with this Kit. |
553 | |
554 | Details on the perl license can be found in the perl source code which |
555 | may be located via the www.perl.com web page. |
556 | |
557 | This is the most wonderfulest possible qsort I can come up with (and |
558 | still be mostly portable) My (limited) tests indicate it consistently |
559 | does about 20% fewer calls to compare than does the qsort in the Visual |
560 | C++ library, other vendors may vary. |
561 | |
562 | Some of the ideas in here can be found in "Algorithms" by Sedgewick, |
563 | others I invented myself (or more likely re-invented since they seemed |
564 | pretty obvious once I watched the algorithm operate for a while). |
565 | |
566 | Most of this code was written while watching the Marlins sweep the Giants |
567 | in the 1997 National League Playoffs - no Braves fans allowed to use this |
568 | code (just kidding :-). |
569 | |
570 | I realize that if I wanted to be true to the perl tradition, the only |
571 | comment in this file would be something like: |
572 | |
573 | ...they shuffled back towards the rear of the line. 'No, not at the |
574 | rear!' the slave-driver shouted. 'Three files up. And stay there... |
575 | |
576 | However, I really needed to violate that tradition just so I could keep |
577 | track of what happens myself, not to mention some poor fool trying to |
578 | understand this years from now :-). |
579 | */ |
580 | |
581 | /* ********************************************************** Configuration */ |
582 | |
583 | #ifndef QSORT_ORDER_GUESS |
584 | #define QSORT_ORDER_GUESS 2 /* Select doubling version of the netBSD trick */ |
585 | #endif |
586 | |
587 | /* QSORT_MAX_STACK is the largest number of partitions that can be stacked up for |
588 | future processing - a good max upper bound is log base 2 of memory size |
589 | (32 on 32 bit machines, 64 on 64 bit machines, etc). In reality can |
590 | safely be smaller than that since the program is taking up some space and |
591 | most operating systems only let you grab some subset of contiguous |
592 | memory (not to mention that you are normally sorting data larger than |
593 | 1 byte element size :-). |
594 | */ |
595 | #ifndef QSORT_MAX_STACK |
596 | #define QSORT_MAX_STACK 32 |
597 | #endif |
598 | |
599 | /* QSORT_BREAK_EVEN is the size of the largest partition we should insertion sort. |
600 | Anything bigger and we use qsort. If you make this too small, the qsort |
601 | will probably break (or become less efficient), because it doesn't expect |
602 | the middle element of a partition to be the same as the right or left - |
603 | you have been warned). |
604 | */ |
605 | #ifndef QSORT_BREAK_EVEN |
606 | #define QSORT_BREAK_EVEN 6 |
607 | #endif |
608 | |
4eb872f6 |
609 | /* QSORT_PLAY_SAFE is the size of the largest partition we're willing |
610 | to go quadratic on. We innoculate larger partitions against |
611 | quadratic behavior by shuffling them before sorting. This is not |
612 | an absolute guarantee of non-quadratic behavior, but it would take |
613 | staggeringly bad luck to pick extreme elements as the pivot |
614 | from randomized data. |
615 | */ |
616 | #ifndef QSORT_PLAY_SAFE |
617 | #define QSORT_PLAY_SAFE 255 |
618 | #endif |
619 | |
84d4ea48 |
620 | /* ************************************************************* Data Types */ |
621 | |
622 | /* hold left and right index values of a partition waiting to be sorted (the |
623 | partition includes both left and right - right is NOT one past the end or |
624 | anything like that). |
625 | */ |
626 | struct partition_stack_entry { |
627 | int left; |
628 | int right; |
629 | #ifdef QSORT_ORDER_GUESS |
630 | int qsort_break_even; |
631 | #endif |
632 | }; |
633 | |
634 | /* ******************************************************* Shorthand Macros */ |
635 | |
636 | /* Note that these macros will be used from inside the qsort function where |
637 | we happen to know that the variable 'elt_size' contains the size of an |
638 | array element and the variable 'temp' points to enough space to hold a |
639 | temp element and the variable 'array' points to the array being sorted |
640 | and 'compare' is the pointer to the compare routine. |
641 | |
642 | Also note that there are very many highly architecture specific ways |
643 | these might be sped up, but this is simply the most generally portable |
644 | code I could think of. |
645 | */ |
646 | |
647 | /* Return < 0 == 0 or > 0 as the value of elt1 is < elt2, == elt2, > elt2 |
648 | */ |
649 | #define qsort_cmp(elt1, elt2) \ |
650 | ((*compare)(aTHX_ array[elt1], array[elt2])) |
651 | |
652 | #ifdef QSORT_ORDER_GUESS |
653 | #define QSORT_NOTICE_SWAP swapped++; |
654 | #else |
655 | #define QSORT_NOTICE_SWAP |
656 | #endif |
657 | |
658 | /* swaps contents of array elements elt1, elt2. |
659 | */ |
660 | #define qsort_swap(elt1, elt2) \ |
661 | STMT_START { \ |
662 | QSORT_NOTICE_SWAP \ |
663 | temp = array[elt1]; \ |
664 | array[elt1] = array[elt2]; \ |
665 | array[elt2] = temp; \ |
666 | } STMT_END |
667 | |
668 | /* rotate contents of elt1, elt2, elt3 such that elt1 gets elt2, elt2 gets |
669 | elt3 and elt3 gets elt1. |
670 | */ |
671 | #define qsort_rotate(elt1, elt2, elt3) \ |
672 | STMT_START { \ |
673 | QSORT_NOTICE_SWAP \ |
674 | temp = array[elt1]; \ |
675 | array[elt1] = array[elt2]; \ |
676 | array[elt2] = array[elt3]; \ |
677 | array[elt3] = temp; \ |
678 | } STMT_END |
679 | |
680 | /* ************************************************************ Debug stuff */ |
681 | |
682 | #ifdef QSORT_DEBUG |
683 | |
684 | static void |
685 | break_here() |
686 | { |
687 | return; /* good place to set a breakpoint */ |
688 | } |
689 | |
690 | #define qsort_assert(t) (void)( (t) || (break_here(), 0) ) |
691 | |
692 | static void |
693 | doqsort_all_asserts( |
694 | void * array, |
695 | size_t num_elts, |
696 | size_t elt_size, |
697 | int (*compare)(const void * elt1, const void * elt2), |
698 | int pc_left, int pc_right, int u_left, int u_right) |
699 | { |
700 | int i; |
701 | |
702 | qsort_assert(pc_left <= pc_right); |
703 | qsort_assert(u_right < pc_left); |
704 | qsort_assert(pc_right < u_left); |
705 | for (i = u_right + 1; i < pc_left; ++i) { |
706 | qsort_assert(qsort_cmp(i, pc_left) < 0); |
707 | } |
708 | for (i = pc_left; i < pc_right; ++i) { |
709 | qsort_assert(qsort_cmp(i, pc_right) == 0); |
710 | } |
711 | for (i = pc_right + 1; i < u_left; ++i) { |
712 | qsort_assert(qsort_cmp(pc_right, i) < 0); |
713 | } |
714 | } |
715 | |
716 | #define qsort_all_asserts(PC_LEFT, PC_RIGHT, U_LEFT, U_RIGHT) \ |
717 | doqsort_all_asserts(array, num_elts, elt_size, compare, \ |
718 | PC_LEFT, PC_RIGHT, U_LEFT, U_RIGHT) |
719 | |
720 | #else |
721 | |
722 | #define qsort_assert(t) ((void)0) |
723 | |
724 | #define qsort_all_asserts(PC_LEFT, PC_RIGHT, U_LEFT, U_RIGHT) ((void)0) |
725 | |
726 | #endif |
727 | |
728 | /* ****************************************************************** qsort */ |
729 | |
730 | STATIC void /* the standard unstable (u) quicksort (qsort) */ |
731 | S_qsortsvu(pTHX_ SV ** array, size_t num_elts, SVCOMPARE_t compare) |
732 | { |
733 | register SV * temp; |
734 | |
735 | struct partition_stack_entry partition_stack[QSORT_MAX_STACK]; |
736 | int next_stack_entry = 0; |
737 | |
738 | int part_left; |
739 | int part_right; |
740 | #ifdef QSORT_ORDER_GUESS |
741 | int qsort_break_even; |
742 | int swapped; |
743 | #endif |
744 | |
745 | /* Make sure we actually have work to do. |
746 | */ |
747 | if (num_elts <= 1) { |
748 | return; |
749 | } |
750 | |
4eb872f6 |
751 | /* Innoculate large partitions against quadratic behavior */ |
752 | if (num_elts > QSORT_PLAY_SAFE) { |
753 | register size_t n, j; |
754 | register SV **q; |
755 | for (n = num_elts, q = array; n > 1; ) { |
756 | j = n-- * Drand01(); |
757 | temp = q[j]; |
758 | q[j] = q[n]; |
759 | q[n] = temp; |
760 | } |
761 | } |
762 | |
84d4ea48 |
763 | /* Setup the initial partition definition and fall into the sorting loop |
764 | */ |
765 | part_left = 0; |
766 | part_right = (int)(num_elts - 1); |
767 | #ifdef QSORT_ORDER_GUESS |
768 | qsort_break_even = QSORT_BREAK_EVEN; |
769 | #else |
770 | #define qsort_break_even QSORT_BREAK_EVEN |
771 | #endif |
772 | for ( ; ; ) { |
773 | if ((part_right - part_left) >= qsort_break_even) { |
774 | /* OK, this is gonna get hairy, so lets try to document all the |
775 | concepts and abbreviations and variables and what they keep |
776 | track of: |
777 | |
778 | pc: pivot chunk - the set of array elements we accumulate in the |
779 | middle of the partition, all equal in value to the original |
780 | pivot element selected. The pc is defined by: |
781 | |
782 | pc_left - the leftmost array index of the pc |
783 | pc_right - the rightmost array index of the pc |
784 | |
785 | we start with pc_left == pc_right and only one element |
786 | in the pivot chunk (but it can grow during the scan). |
787 | |
788 | u: uncompared elements - the set of elements in the partition |
789 | we have not yet compared to the pivot value. There are two |
790 | uncompared sets during the scan - one to the left of the pc |
791 | and one to the right. |
792 | |
793 | u_right - the rightmost index of the left side's uncompared set |
794 | u_left - the leftmost index of the right side's uncompared set |
795 | |
796 | The leftmost index of the left sides's uncompared set |
797 | doesn't need its own variable because it is always defined |
798 | by the leftmost edge of the whole partition (part_left). The |
799 | same goes for the rightmost edge of the right partition |
800 | (part_right). |
801 | |
802 | We know there are no uncompared elements on the left once we |
803 | get u_right < part_left and no uncompared elements on the |
804 | right once u_left > part_right. When both these conditions |
805 | are met, we have completed the scan of the partition. |
806 | |
807 | Any elements which are between the pivot chunk and the |
808 | uncompared elements should be less than the pivot value on |
809 | the left side and greater than the pivot value on the right |
810 | side (in fact, the goal of the whole algorithm is to arrange |
811 | for that to be true and make the groups of less-than and |
812 | greater-then elements into new partitions to sort again). |
813 | |
814 | As you marvel at the complexity of the code and wonder why it |
815 | has to be so confusing. Consider some of the things this level |
816 | of confusion brings: |
817 | |
818 | Once I do a compare, I squeeze every ounce of juice out of it. I |
819 | never do compare calls I don't have to do, and I certainly never |
820 | do redundant calls. |
821 | |
822 | I also never swap any elements unless I can prove there is a |
823 | good reason. Many sort algorithms will swap a known value with |
824 | an uncompared value just to get things in the right place (or |
825 | avoid complexity :-), but that uncompared value, once it gets |
826 | compared, may then have to be swapped again. A lot of the |
827 | complexity of this code is due to the fact that it never swaps |
828 | anything except compared values, and it only swaps them when the |
829 | compare shows they are out of position. |
830 | */ |
831 | int pc_left, pc_right; |
832 | int u_right, u_left; |
833 | |
834 | int s; |
835 | |
836 | pc_left = ((part_left + part_right) / 2); |
837 | pc_right = pc_left; |
838 | u_right = pc_left - 1; |
839 | u_left = pc_right + 1; |
840 | |
841 | /* Qsort works best when the pivot value is also the median value |
842 | in the partition (unfortunately you can't find the median value |
843 | without first sorting :-), so to give the algorithm a helping |
844 | hand, we pick 3 elements and sort them and use the median value |
845 | of that tiny set as the pivot value. |
846 | |
847 | Some versions of qsort like to use the left middle and right as |
848 | the 3 elements to sort so they can insure the ends of the |
849 | partition will contain values which will stop the scan in the |
850 | compare loop, but when you have to call an arbitrarily complex |
851 | routine to do a compare, its really better to just keep track of |
852 | array index values to know when you hit the edge of the |
853 | partition and avoid the extra compare. An even better reason to |
854 | avoid using a compare call is the fact that you can drop off the |
855 | edge of the array if someone foolishly provides you with an |
856 | unstable compare function that doesn't always provide consistent |
857 | results. |
858 | |
859 | So, since it is simpler for us to compare the three adjacent |
860 | elements in the middle of the partition, those are the ones we |
861 | pick here (conveniently pointed at by u_right, pc_left, and |
862 | u_left). The values of the left, center, and right elements |
863 | are refered to as l c and r in the following comments. |
864 | */ |
865 | |
866 | #ifdef QSORT_ORDER_GUESS |
867 | swapped = 0; |
868 | #endif |
869 | s = qsort_cmp(u_right, pc_left); |
870 | if (s < 0) { |
871 | /* l < c */ |
872 | s = qsort_cmp(pc_left, u_left); |
873 | /* if l < c, c < r - already in order - nothing to do */ |
874 | if (s == 0) { |
875 | /* l < c, c == r - already in order, pc grows */ |
876 | ++pc_right; |
877 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
878 | } else if (s > 0) { |
879 | /* l < c, c > r - need to know more */ |
880 | s = qsort_cmp(u_right, u_left); |
881 | if (s < 0) { |
882 | /* l < c, c > r, l < r - swap c & r to get ordered */ |
883 | qsort_swap(pc_left, u_left); |
884 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
885 | } else if (s == 0) { |
886 | /* l < c, c > r, l == r - swap c&r, grow pc */ |
887 | qsort_swap(pc_left, u_left); |
888 | --pc_left; |
889 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
890 | } else { |
891 | /* l < c, c > r, l > r - make lcr into rlc to get ordered */ |
892 | qsort_rotate(pc_left, u_right, u_left); |
893 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
894 | } |
895 | } |
896 | } else if (s == 0) { |
897 | /* l == c */ |
898 | s = qsort_cmp(pc_left, u_left); |
899 | if (s < 0) { |
900 | /* l == c, c < r - already in order, grow pc */ |
901 | --pc_left; |
902 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
903 | } else if (s == 0) { |
904 | /* l == c, c == r - already in order, grow pc both ways */ |
905 | --pc_left; |
906 | ++pc_right; |
907 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
908 | } else { |
909 | /* l == c, c > r - swap l & r, grow pc */ |
910 | qsort_swap(u_right, u_left); |
911 | ++pc_right; |
912 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
913 | } |
914 | } else { |
915 | /* l > c */ |
916 | s = qsort_cmp(pc_left, u_left); |
917 | if (s < 0) { |
918 | /* l > c, c < r - need to know more */ |
919 | s = qsort_cmp(u_right, u_left); |
920 | if (s < 0) { |
921 | /* l > c, c < r, l < r - swap l & c to get ordered */ |
922 | qsort_swap(u_right, pc_left); |
923 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
924 | } else if (s == 0) { |
925 | /* l > c, c < r, l == r - swap l & c, grow pc */ |
926 | qsort_swap(u_right, pc_left); |
927 | ++pc_right; |
928 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
929 | } else { |
930 | /* l > c, c < r, l > r - rotate lcr into crl to order */ |
931 | qsort_rotate(u_right, pc_left, u_left); |
932 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
933 | } |
934 | } else if (s == 0) { |
935 | /* l > c, c == r - swap ends, grow pc */ |
936 | qsort_swap(u_right, u_left); |
937 | --pc_left; |
938 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
939 | } else { |
940 | /* l > c, c > r - swap ends to get in order */ |
941 | qsort_swap(u_right, u_left); |
942 | qsort_all_asserts(pc_left, pc_right, u_left + 1, u_right - 1); |
943 | } |
944 | } |
945 | /* We now know the 3 middle elements have been compared and |
946 | arranged in the desired order, so we can shrink the uncompared |
947 | sets on both sides |
948 | */ |
949 | --u_right; |
950 | ++u_left; |
951 | qsort_all_asserts(pc_left, pc_right, u_left, u_right); |
952 | |
953 | /* The above massive nested if was the simple part :-). We now have |
954 | the middle 3 elements ordered and we need to scan through the |
955 | uncompared sets on either side, swapping elements that are on |
956 | the wrong side or simply shuffling equal elements around to get |
957 | all equal elements into the pivot chunk. |
958 | */ |
959 | |
960 | for ( ; ; ) { |
961 | int still_work_on_left; |
962 | int still_work_on_right; |
963 | |
964 | /* Scan the uncompared values on the left. If I find a value |
965 | equal to the pivot value, move it over so it is adjacent to |
966 | the pivot chunk and expand the pivot chunk. If I find a value |
967 | less than the pivot value, then just leave it - its already |
968 | on the correct side of the partition. If I find a greater |
969 | value, then stop the scan. |
970 | */ |
971 | while ((still_work_on_left = (u_right >= part_left))) { |
972 | s = qsort_cmp(u_right, pc_left); |
973 | if (s < 0) { |
974 | --u_right; |
975 | } else if (s == 0) { |
976 | --pc_left; |
977 | if (pc_left != u_right) { |
978 | qsort_swap(u_right, pc_left); |
979 | } |
980 | --u_right; |
981 | } else { |
982 | break; |
983 | } |
984 | qsort_assert(u_right < pc_left); |
985 | qsort_assert(pc_left <= pc_right); |
986 | qsort_assert(qsort_cmp(u_right + 1, pc_left) <= 0); |
987 | qsort_assert(qsort_cmp(pc_left, pc_right) == 0); |
988 | } |
989 | |
990 | /* Do a mirror image scan of uncompared values on the right |
991 | */ |
992 | while ((still_work_on_right = (u_left <= part_right))) { |
993 | s = qsort_cmp(pc_right, u_left); |
994 | if (s < 0) { |
995 | ++u_left; |
996 | } else if (s == 0) { |
997 | ++pc_right; |
998 | if (pc_right != u_left) { |
999 | qsort_swap(pc_right, u_left); |
1000 | } |
1001 | ++u_left; |
1002 | } else { |
1003 | break; |
1004 | } |
1005 | qsort_assert(u_left > pc_right); |
1006 | qsort_assert(pc_left <= pc_right); |
1007 | qsort_assert(qsort_cmp(pc_right, u_left - 1) <= 0); |
1008 | qsort_assert(qsort_cmp(pc_left, pc_right) == 0); |
1009 | } |
1010 | |
1011 | if (still_work_on_left) { |
1012 | /* I know I have a value on the left side which needs to be |
1013 | on the right side, but I need to know more to decide |
1014 | exactly the best thing to do with it. |
1015 | */ |
1016 | if (still_work_on_right) { |
1017 | /* I know I have values on both side which are out of |
1018 | position. This is a big win because I kill two birds |
1019 | with one swap (so to speak). I can advance the |
1020 | uncompared pointers on both sides after swapping both |
1021 | of them into the right place. |
1022 | */ |
1023 | qsort_swap(u_right, u_left); |
1024 | --u_right; |
1025 | ++u_left; |
1026 | qsort_all_asserts(pc_left, pc_right, u_left, u_right); |
1027 | } else { |
1028 | /* I have an out of position value on the left, but the |
1029 | right is fully scanned, so I "slide" the pivot chunk |
1030 | and any less-than values left one to make room for the |
1031 | greater value over on the right. If the out of position |
1032 | value is immediately adjacent to the pivot chunk (there |
1033 | are no less-than values), I can do that with a swap, |
1034 | otherwise, I have to rotate one of the less than values |
1035 | into the former position of the out of position value |
1036 | and the right end of the pivot chunk into the left end |
1037 | (got all that?). |
1038 | */ |
1039 | --pc_left; |
1040 | if (pc_left == u_right) { |
1041 | qsort_swap(u_right, pc_right); |
1042 | qsort_all_asserts(pc_left, pc_right-1, u_left, u_right-1); |
1043 | } else { |
1044 | qsort_rotate(u_right, pc_left, pc_right); |
1045 | qsort_all_asserts(pc_left, pc_right-1, u_left, u_right-1); |
1046 | } |
1047 | --pc_right; |
1048 | --u_right; |
1049 | } |
1050 | } else if (still_work_on_right) { |
1051 | /* Mirror image of complex case above: I have an out of |
1052 | position value on the right, but the left is fully |
1053 | scanned, so I need to shuffle things around to make room |
1054 | for the right value on the left. |
1055 | */ |
1056 | ++pc_right; |
1057 | if (pc_right == u_left) { |
1058 | qsort_swap(u_left, pc_left); |
1059 | qsort_all_asserts(pc_left+1, pc_right, u_left+1, u_right); |
1060 | } else { |
1061 | qsort_rotate(pc_right, pc_left, u_left); |
1062 | qsort_all_asserts(pc_left+1, pc_right, u_left+1, u_right); |
1063 | } |
1064 | ++pc_left; |
1065 | ++u_left; |
1066 | } else { |
1067 | /* No more scanning required on either side of partition, |
1068 | break out of loop and figure out next set of partitions |
1069 | */ |
1070 | break; |
1071 | } |
1072 | } |
1073 | |
1074 | /* The elements in the pivot chunk are now in the right place. They |
1075 | will never move or be compared again. All I have to do is decide |
1076 | what to do with the stuff to the left and right of the pivot |
1077 | chunk. |
1078 | |
1079 | Notes on the QSORT_ORDER_GUESS ifdef code: |
1080 | |
1081 | 1. If I just built these partitions without swapping any (or |
1082 | very many) elements, there is a chance that the elements are |
1083 | already ordered properly (being properly ordered will |
1084 | certainly result in no swapping, but the converse can't be |
1085 | proved :-). |
1086 | |
1087 | 2. A (properly written) insertion sort will run faster on |
1088 | already ordered data than qsort will. |
1089 | |
1090 | 3. Perhaps there is some way to make a good guess about |
1091 | switching to an insertion sort earlier than partition size 6 |
1092 | (for instance - we could save the partition size on the stack |
1093 | and increase the size each time we find we didn't swap, thus |
1094 | switching to insertion sort earlier for partitions with a |
1095 | history of not swapping). |
1096 | |
1097 | 4. Naturally, if I just switch right away, it will make |
1098 | artificial benchmarks with pure ascending (or descending) |
1099 | data look really good, but is that a good reason in general? |
1100 | Hard to say... |
1101 | */ |
1102 | |
1103 | #ifdef QSORT_ORDER_GUESS |
1104 | if (swapped < 3) { |
1105 | #if QSORT_ORDER_GUESS == 1 |
1106 | qsort_break_even = (part_right - part_left) + 1; |
1107 | #endif |
1108 | #if QSORT_ORDER_GUESS == 2 |
1109 | qsort_break_even *= 2; |
1110 | #endif |
1111 | #if QSORT_ORDER_GUESS == 3 |
1112 | int prev_break = qsort_break_even; |
1113 | qsort_break_even *= qsort_break_even; |
1114 | if (qsort_break_even < prev_break) { |
1115 | qsort_break_even = (part_right - part_left) + 1; |
1116 | } |
1117 | #endif |
1118 | } else { |
1119 | qsort_break_even = QSORT_BREAK_EVEN; |
1120 | } |
1121 | #endif |
1122 | |
1123 | if (part_left < pc_left) { |
1124 | /* There are elements on the left which need more processing. |
1125 | Check the right as well before deciding what to do. |
1126 | */ |
1127 | if (pc_right < part_right) { |
1128 | /* We have two partitions to be sorted. Stack the biggest one |
1129 | and process the smallest one on the next iteration. This |
1130 | minimizes the stack height by insuring that any additional |
1131 | stack entries must come from the smallest partition which |
1132 | (because it is smallest) will have the fewest |
1133 | opportunities to generate additional stack entries. |
1134 | */ |
1135 | if ((part_right - pc_right) > (pc_left - part_left)) { |
1136 | /* stack the right partition, process the left */ |
1137 | partition_stack[next_stack_entry].left = pc_right + 1; |
1138 | partition_stack[next_stack_entry].right = part_right; |
1139 | #ifdef QSORT_ORDER_GUESS |
1140 | partition_stack[next_stack_entry].qsort_break_even = qsort_break_even; |
1141 | #endif |
1142 | part_right = pc_left - 1; |
1143 | } else { |
1144 | /* stack the left partition, process the right */ |
1145 | partition_stack[next_stack_entry].left = part_left; |
1146 | partition_stack[next_stack_entry].right = pc_left - 1; |
1147 | #ifdef QSORT_ORDER_GUESS |
1148 | partition_stack[next_stack_entry].qsort_break_even = qsort_break_even; |
1149 | #endif |
1150 | part_left = pc_right + 1; |
1151 | } |
1152 | qsort_assert(next_stack_entry < QSORT_MAX_STACK); |
1153 | ++next_stack_entry; |
1154 | } else { |
1155 | /* The elements on the left are the only remaining elements |
1156 | that need sorting, arrange for them to be processed as the |
1157 | next partition. |
1158 | */ |
1159 | part_right = pc_left - 1; |
1160 | } |
1161 | } else if (pc_right < part_right) { |
1162 | /* There is only one chunk on the right to be sorted, make it |
1163 | the new partition and loop back around. |
1164 | */ |
1165 | part_left = pc_right + 1; |
1166 | } else { |
1167 | /* This whole partition wound up in the pivot chunk, so |
1168 | we need to get a new partition off the stack. |
1169 | */ |
1170 | if (next_stack_entry == 0) { |
1171 | /* the stack is empty - we are done */ |
1172 | break; |
1173 | } |
1174 | --next_stack_entry; |
1175 | part_left = partition_stack[next_stack_entry].left; |
1176 | part_right = partition_stack[next_stack_entry].right; |
1177 | #ifdef QSORT_ORDER_GUESS |
1178 | qsort_break_even = partition_stack[next_stack_entry].qsort_break_even; |
1179 | #endif |
1180 | } |
1181 | } else { |
1182 | /* This partition is too small to fool with qsort complexity, just |
1183 | do an ordinary insertion sort to minimize overhead. |
1184 | */ |
1185 | int i; |
1186 | /* Assume 1st element is in right place already, and start checking |
1187 | at 2nd element to see where it should be inserted. |
1188 | */ |
1189 | for (i = part_left + 1; i <= part_right; ++i) { |
1190 | int j; |
1191 | /* Scan (backwards - just in case 'i' is already in right place) |
1192 | through the elements already sorted to see if the ith element |
1193 | belongs ahead of one of them. |
1194 | */ |
1195 | for (j = i - 1; j >= part_left; --j) { |
1196 | if (qsort_cmp(i, j) >= 0) { |
1197 | /* i belongs right after j |
1198 | */ |
1199 | break; |
1200 | } |
1201 | } |
1202 | ++j; |
1203 | if (j != i) { |
1204 | /* Looks like we really need to move some things |
1205 | */ |
1206 | int k; |
1207 | temp = array[i]; |
1208 | for (k = i - 1; k >= j; --k) |
1209 | array[k + 1] = array[k]; |
1210 | array[j] = temp; |
1211 | } |
1212 | } |
1213 | |
1214 | /* That partition is now sorted, grab the next one, or get out |
1215 | of the loop if there aren't any more. |
1216 | */ |
1217 | |
1218 | if (next_stack_entry == 0) { |
1219 | /* the stack is empty - we are done */ |
1220 | break; |
1221 | } |
1222 | --next_stack_entry; |
1223 | part_left = partition_stack[next_stack_entry].left; |
1224 | part_right = partition_stack[next_stack_entry].right; |
1225 | #ifdef QSORT_ORDER_GUESS |
1226 | qsort_break_even = partition_stack[next_stack_entry].qsort_break_even; |
1227 | #endif |
1228 | } |
1229 | } |
1230 | |
1231 | /* Believe it or not, the array is sorted at this point! */ |
1232 | } |
1233 | |
84d4ea48 |
1234 | /* Stabilize what is, presumably, an otherwise unstable sort method. |
1235 | * We do that by allocating (or having on hand) an array of pointers |
1236 | * that is the same size as the original array of elements to be sorted. |
1237 | * We initialize this parallel array with the addresses of the original |
1238 | * array elements. This indirection can make you crazy. |
1239 | * Some pictures can help. After initializing, we have |
1240 | * |
1241 | * indir list1 |
1242 | * +----+ +----+ |
1243 | * | | --------------> | | ------> first element to be sorted |
1244 | * +----+ +----+ |
1245 | * | | --------------> | | ------> second element to be sorted |
1246 | * +----+ +----+ |
1247 | * | | --------------> | | ------> third element to be sorted |
1248 | * +----+ +----+ |
1249 | * ... |
1250 | * +----+ +----+ |
1251 | * | | --------------> | | ------> n-1st element to be sorted |
1252 | * +----+ +----+ |
1253 | * | | --------------> | | ------> n-th element to be sorted |
1254 | * +----+ +----+ |
1255 | * |
1256 | * During the sort phase, we leave the elements of list1 where they are, |
1257 | * and sort the pointers in the indirect array in the same order determined |
1258 | * by the original comparison routine on the elements pointed to. |
1259 | * Because we don't move the elements of list1 around through |
1260 | * this phase, we can break ties on elements that compare equal |
1261 | * using their address in the list1 array, ensuring stabilty. |
1262 | * This leaves us with something looking like |
1263 | * |
1264 | * indir list1 |
1265 | * +----+ +----+ |
1266 | * | | --+ +---> | | ------> first element to be sorted |
1267 | * +----+ | | +----+ |
1268 | * | | --|-------|---> | | ------> second element to be sorted |
1269 | * +----+ | | +----+ |
1270 | * | | --|-------+ +-> | | ------> third element to be sorted |
1271 | * +----+ | | +----+ |
1272 | * ... |
1273 | * +----+ | | | | +----+ |
1274 | * | | ---|-+ | +--> | | ------> n-1st element to be sorted |
1275 | * +----+ | | +----+ |
1276 | * | | ---+ +----> | | ------> n-th element to be sorted |
1277 | * +----+ +----+ |
1278 | * |
1279 | * where the i-th element of the indirect array points to the element |
1280 | * that should be i-th in the sorted array. After the sort phase, |
1281 | * we have to put the elements of list1 into the places |
1282 | * dictated by the indirect array. |
1283 | */ |
1284 | |
1285 | static SVCOMPARE_t RealCmp; |
1286 | |
1287 | static I32 |
1288 | cmpindir(pTHX_ gptr a, gptr b) |
1289 | { |
1290 | I32 sense; |
1291 | gptr *ap = (gptr *)a; |
1292 | gptr *bp = (gptr *)b; |
1293 | |
1294 | if ((sense = RealCmp(aTHX_ *ap, *bp)) == 0) |
1295 | sense = (ap > bp) ? 1 : ((ap < bp) ? -1 : 0); |
1296 | return sense; |
1297 | } |
1298 | |
1299 | STATIC void |
1300 | S_qsortsv(pTHX_ gptr *list1, size_t nmemb, SVCOMPARE_t cmp) |
1301 | { |
1302 | SV **hintsvp; |
1303 | |
c53fc8a6 |
1304 | if (SORTHINTS(hintsvp) & HINT_SORT_STABLE) { |
84d4ea48 |
1305 | register gptr **pp, *q; |
1306 | register size_t n, j, i; |
1307 | gptr *small[SMALLSORT], **indir, tmp; |
1308 | SVCOMPARE_t savecmp; |
1309 | if (nmemb <= 1) return; /* sorted trivially */ |
4eb872f6 |
1310 | |
84d4ea48 |
1311 | /* Small arrays can use the stack, big ones must be allocated */ |
1312 | if (nmemb <= SMALLSORT) indir = small; |
1313 | else { New(1799, indir, nmemb, gptr *); } |
4eb872f6 |
1314 | |
84d4ea48 |
1315 | /* Copy pointers to original array elements into indirect array */ |
1316 | for (n = nmemb, pp = indir, q = list1; n--; ) *pp++ = q++; |
4eb872f6 |
1317 | |
84d4ea48 |
1318 | savecmp = RealCmp; /* Save current comparison routine, if any */ |
1319 | RealCmp = cmp; /* Put comparison routine where cmpindir can find it */ |
4eb872f6 |
1320 | |
84d4ea48 |
1321 | /* sort, with indirection */ |
1322 | S_qsortsvu(aTHX_ (gptr *)indir, nmemb, cmpindir); |
4eb872f6 |
1323 | |
84d4ea48 |
1324 | pp = indir; |
1325 | q = list1; |
1326 | for (n = nmemb; n--; ) { |
1327 | /* Assert A: all elements of q with index > n are already |
1328 | * in place. This is vacuosly true at the start, and we |
1329 | * put element n where it belongs below (if it wasn't |
1330 | * already where it belonged). Assert B: we only move |
1331 | * elements that aren't where they belong, |
1332 | * so, by A, we never tamper with elements above n. |
1333 | */ |
1334 | j = pp[n] - q; /* This sets j so that q[j] is |
1335 | * at pp[n]. *pp[j] belongs in |
1336 | * q[j], by construction. |
1337 | */ |
1338 | if (n != j) { /* all's well if n == j */ |
1339 | tmp = q[j]; /* save what's in q[j] */ |
1340 | do { |
1341 | q[j] = *pp[j]; /* put *pp[j] where it belongs */ |
1342 | i = pp[j] - q; /* the index in q of the element |
1343 | * just moved */ |
1344 | pp[j] = q + j; /* this is ok now */ |
1345 | } while ((j = i) != n); |
1346 | /* There are only finitely many (nmemb) addresses |
1347 | * in the pp array. |
1348 | * So we must eventually revisit an index we saw before. |
1349 | * Suppose the first revisited index is k != n. |
1350 | * An index is visited because something else belongs there. |
1351 | * If we visit k twice, then two different elements must |
1352 | * belong in the same place, which cannot be. |
1353 | * So j must get back to n, the loop terminates, |
1354 | * and we put the saved element where it belongs. |
1355 | */ |
1356 | q[n] = tmp; /* put what belongs into |
1357 | * the n-th element */ |
1358 | } |
1359 | } |
1360 | |
1361 | /* free iff allocated */ |
1362 | if (indir != small) { Safefree(indir); } |
1363 | /* restore prevailing comparison routine */ |
1364 | RealCmp = savecmp; |
c53fc8a6 |
1365 | } else { |
1366 | S_qsortsvu(aTHX_ list1, nmemb, cmp); |
84d4ea48 |
1367 | } |
1368 | } |
4eb872f6 |
1369 | |
1370 | /* |
ccfc67b7 |
1371 | =head1 Array Manipulation Functions |
1372 | |
84d4ea48 |
1373 | =for apidoc sortsv |
1374 | |
1375 | Sort an array. Here is an example: |
1376 | |
4eb872f6 |
1377 | sortsv(AvARRAY(av), av_len(av)+1, Perl_sv_cmp_locale); |
84d4ea48 |
1378 | |
1379 | =cut |
1380 | */ |
4eb872f6 |
1381 | |
84d4ea48 |
1382 | void |
1383 | Perl_sortsv(pTHX_ SV **array, size_t nmemb, SVCOMPARE_t cmp) |
1384 | { |
1385 | void (*sortsvp)(pTHX_ SV **array, size_t nmemb, SVCOMPARE_t cmp) = |
1386 | S_mergesortsv; |
1387 | SV **hintsvp; |
1388 | I32 hints; |
4eb872f6 |
1389 | |
84d4ea48 |
1390 | if ((hints = SORTHINTS(hintsvp))) { |
1391 | if (hints & HINT_SORT_QUICKSORT) |
1392 | sortsvp = S_qsortsv; |
1393 | else { |
1394 | if (hints & HINT_SORT_MERGESORT) |
3fe0b9a9 |
1395 | sortsvp = S_mergesortsv; |
84d4ea48 |
1396 | else |
1397 | sortsvp = S_mergesortsv; |
1398 | } |
1399 | } |
4eb872f6 |
1400 | |
84d4ea48 |
1401 | sortsvp(aTHX_ array, nmemb, cmp); |
1402 | } |
1403 | |
1404 | PP(pp_sort) |
1405 | { |
1406 | dSP; dMARK; dORIGMARK; |
1407 | register SV **up; |
1408 | SV **myorigmark = ORIGMARK; |
1409 | register I32 max; |
1410 | HV *stash; |
1411 | GV *gv; |
1412 | CV *cv = 0; |
1413 | I32 gimme = GIMME; |
1414 | OP* nextop = PL_op->op_next; |
1415 | I32 overloading = 0; |
1416 | bool hasargs = FALSE; |
1417 | I32 is_xsub = 0; |
1418 | |
1419 | if (gimme != G_ARRAY) { |
1420 | SP = MARK; |
1421 | RETPUSHUNDEF; |
1422 | } |
1423 | |
1424 | ENTER; |
1425 | SAVEVPTR(PL_sortcop); |
1426 | if (PL_op->op_flags & OPf_STACKED) { |
1427 | if (PL_op->op_flags & OPf_SPECIAL) { |
1428 | OP *kid = cLISTOP->op_first->op_sibling; /* pass pushmark */ |
1429 | kid = kUNOP->op_first; /* pass rv2gv */ |
1430 | kid = kUNOP->op_first; /* pass leave */ |
1431 | PL_sortcop = kid->op_next; |
1432 | stash = CopSTASH(PL_curcop); |
1433 | } |
1434 | else { |
1435 | cv = sv_2cv(*++MARK, &stash, &gv, 0); |
1436 | if (cv && SvPOK(cv)) { |
1437 | STRLEN n_a; |
1438 | char *proto = SvPV((SV*)cv, n_a); |
1439 | if (proto && strEQ(proto, "$$")) { |
1440 | hasargs = TRUE; |
1441 | } |
1442 | } |
1443 | if (!(cv && CvROOT(cv))) { |
1444 | if (cv && CvXSUB(cv)) { |
1445 | is_xsub = 1; |
1446 | } |
1447 | else if (gv) { |
1448 | SV *tmpstr = sv_newmortal(); |
1449 | gv_efullname3(tmpstr, gv, Nullch); |
1450 | DIE(aTHX_ "Undefined sort subroutine \"%s\" called", |
1451 | SvPVX(tmpstr)); |
1452 | } |
1453 | else { |
1454 | DIE(aTHX_ "Undefined subroutine in sort"); |
1455 | } |
1456 | } |
1457 | |
1458 | if (is_xsub) |
1459 | PL_sortcop = (OP*)cv; |
1460 | else { |
1461 | PL_sortcop = CvSTART(cv); |
1462 | SAVEVPTR(CvROOT(cv)->op_ppaddr); |
1463 | CvROOT(cv)->op_ppaddr = PL_ppaddr[OP_NULL]; |
1464 | |
1465 | SAVEVPTR(PL_curpad); |
1466 | PL_curpad = AvARRAY((AV*)AvARRAY(CvPADLIST(cv))[1]); |
1467 | } |
1468 | } |
1469 | } |
1470 | else { |
1471 | PL_sortcop = Nullop; |
1472 | stash = CopSTASH(PL_curcop); |
1473 | } |
1474 | |
1475 | up = myorigmark + 1; |
1476 | while (MARK < SP) { /* This may or may not shift down one here. */ |
1477 | /*SUPPRESS 560*/ |
1478 | if ((*up = *++MARK)) { /* Weed out nulls. */ |
1479 | SvTEMP_off(*up); |
1480 | if (!PL_sortcop && !SvPOK(*up)) { |
1481 | STRLEN n_a; |
1482 | if (SvAMAGIC(*up)) |
1483 | overloading = 1; |
1484 | else |
1485 | (void)sv_2pv(*up, &n_a); |
1486 | } |
1487 | up++; |
1488 | } |
1489 | } |
1490 | max = --up - myorigmark; |
1491 | if (PL_sortcop) { |
1492 | if (max > 1) { |
1493 | PERL_CONTEXT *cx; |
1494 | SV** newsp; |
1495 | bool oldcatch = CATCH_GET; |
1496 | |
1497 | SAVETMPS; |
1498 | SAVEOP(); |
1499 | |
1500 | CATCH_SET(TRUE); |
1501 | PUSHSTACKi(PERLSI_SORT); |
1502 | if (!hasargs && !is_xsub) { |
1503 | if (PL_sortstash != stash || !PL_firstgv || !PL_secondgv) { |
1504 | SAVESPTR(PL_firstgv); |
1505 | SAVESPTR(PL_secondgv); |
1506 | PL_firstgv = gv_fetchpv("a", TRUE, SVt_PV); |
1507 | PL_secondgv = gv_fetchpv("b", TRUE, SVt_PV); |
1508 | PL_sortstash = stash; |
1509 | } |
1510 | #ifdef USE_5005THREADS |
1511 | sv_lock((SV *)PL_firstgv); |
1512 | sv_lock((SV *)PL_secondgv); |
1513 | #endif |
1514 | SAVESPTR(GvSV(PL_firstgv)); |
1515 | SAVESPTR(GvSV(PL_secondgv)); |
1516 | } |
1517 | |
1518 | PUSHBLOCK(cx, CXt_NULL, PL_stack_base); |
1519 | if (!(PL_op->op_flags & OPf_SPECIAL)) { |
1520 | cx->cx_type = CXt_SUB; |
1521 | cx->blk_gimme = G_SCALAR; |
1522 | PUSHSUB(cx); |
1523 | if (!CvDEPTH(cv)) |
1524 | (void)SvREFCNT_inc(cv); /* in preparation for POPSUB */ |
1525 | } |
1526 | PL_sortcxix = cxstack_ix; |
1527 | |
1528 | if (hasargs && !is_xsub) { |
1529 | /* This is mostly copied from pp_entersub */ |
1530 | AV *av = (AV*)PL_curpad[0]; |
1531 | |
1532 | #ifndef USE_5005THREADS |
1533 | cx->blk_sub.savearray = GvAV(PL_defgv); |
1534 | GvAV(PL_defgv) = (AV*)SvREFCNT_inc(av); |
1535 | #endif /* USE_5005THREADS */ |
1536 | cx->blk_sub.oldcurpad = PL_curpad; |
1537 | cx->blk_sub.argarray = av; |
1538 | } |
1539 | sortsv((myorigmark+1), max, |
1540 | is_xsub ? sortcv_xsub : hasargs ? sortcv_stacked : sortcv); |
1541 | |
1542 | POPBLOCK(cx,PL_curpm); |
1543 | PL_stack_sp = newsp; |
1544 | POPSTACK; |
1545 | CATCH_SET(oldcatch); |
1546 | } |
1547 | } |
1548 | else { |
1549 | if (max > 1) { |
1550 | MEXTEND(SP, 20); /* Can't afford stack realloc on signal. */ |
1551 | sortsv(ORIGMARK+1, max, |
1552 | (PL_op->op_private & OPpSORT_NUMERIC) |
1553 | ? ( (PL_op->op_private & OPpSORT_INTEGER) |
1554 | ? ( overloading ? amagic_i_ncmp : sv_i_ncmp) |
1555 | : ( overloading ? amagic_ncmp : sv_ncmp)) |
1556 | : ( IN_LOCALE_RUNTIME |
1557 | ? ( overloading |
1558 | ? amagic_cmp_locale |
1559 | : sv_cmp_locale_static) |
1560 | : ( overloading ? amagic_cmp : sv_cmp_static))); |
1561 | if (PL_op->op_private & OPpSORT_REVERSE) { |
1562 | SV **p = ORIGMARK+1; |
1563 | SV **q = ORIGMARK+max; |
1564 | while (p < q) { |
1565 | SV *tmp = *p; |
1566 | *p++ = *q; |
1567 | *q-- = tmp; |
1568 | } |
1569 | } |
1570 | } |
1571 | } |
1572 | LEAVE; |
1573 | PL_stack_sp = ORIGMARK + max; |
1574 | return nextop; |
1575 | } |
1576 | |
1577 | static I32 |
1578 | sortcv(pTHX_ SV *a, SV *b) |
1579 | { |
1580 | I32 oldsaveix = PL_savestack_ix; |
1581 | I32 oldscopeix = PL_scopestack_ix; |
1582 | I32 result; |
1583 | GvSV(PL_firstgv) = a; |
1584 | GvSV(PL_secondgv) = b; |
1585 | PL_stack_sp = PL_stack_base; |
1586 | PL_op = PL_sortcop; |
1587 | CALLRUNOPS(aTHX); |
1588 | if (PL_stack_sp != PL_stack_base + 1) |
1589 | Perl_croak(aTHX_ "Sort subroutine didn't return single value"); |
1590 | if (!SvNIOKp(*PL_stack_sp)) |
1591 | Perl_croak(aTHX_ "Sort subroutine didn't return a numeric value"); |
1592 | result = SvIV(*PL_stack_sp); |
1593 | while (PL_scopestack_ix > oldscopeix) { |
1594 | LEAVE; |
1595 | } |
1596 | leave_scope(oldsaveix); |
1597 | return result; |
1598 | } |
1599 | |
1600 | static I32 |
1601 | sortcv_stacked(pTHX_ SV *a, SV *b) |
1602 | { |
1603 | I32 oldsaveix = PL_savestack_ix; |
1604 | I32 oldscopeix = PL_scopestack_ix; |
1605 | I32 result; |
1606 | AV *av; |
1607 | |
1608 | #ifdef USE_5005THREADS |
1609 | av = (AV*)PL_curpad[0]; |
1610 | #else |
1611 | av = GvAV(PL_defgv); |
1612 | #endif |
1613 | |
1614 | if (AvMAX(av) < 1) { |
1615 | SV** ary = AvALLOC(av); |
1616 | if (AvARRAY(av) != ary) { |
1617 | AvMAX(av) += AvARRAY(av) - AvALLOC(av); |
1618 | SvPVX(av) = (char*)ary; |
1619 | } |
1620 | if (AvMAX(av) < 1) { |
1621 | AvMAX(av) = 1; |
1622 | Renew(ary,2,SV*); |
1623 | SvPVX(av) = (char*)ary; |
1624 | } |
1625 | } |
1626 | AvFILLp(av) = 1; |
1627 | |
1628 | AvARRAY(av)[0] = a; |
1629 | AvARRAY(av)[1] = b; |
1630 | PL_stack_sp = PL_stack_base; |
1631 | PL_op = PL_sortcop; |
1632 | CALLRUNOPS(aTHX); |
1633 | if (PL_stack_sp != PL_stack_base + 1) |
1634 | Perl_croak(aTHX_ "Sort subroutine didn't return single value"); |
1635 | if (!SvNIOKp(*PL_stack_sp)) |
1636 | Perl_croak(aTHX_ "Sort subroutine didn't return a numeric value"); |
1637 | result = SvIV(*PL_stack_sp); |
1638 | while (PL_scopestack_ix > oldscopeix) { |
1639 | LEAVE; |
1640 | } |
1641 | leave_scope(oldsaveix); |
1642 | return result; |
1643 | } |
1644 | |
1645 | static I32 |
1646 | sortcv_xsub(pTHX_ SV *a, SV *b) |
1647 | { |
1648 | dSP; |
1649 | I32 oldsaveix = PL_savestack_ix; |
1650 | I32 oldscopeix = PL_scopestack_ix; |
1651 | I32 result; |
1652 | CV *cv=(CV*)PL_sortcop; |
1653 | |
1654 | SP = PL_stack_base; |
1655 | PUSHMARK(SP); |
1656 | EXTEND(SP, 2); |
1657 | *++SP = a; |
1658 | *++SP = b; |
1659 | PUTBACK; |
1660 | (void)(*CvXSUB(cv))(aTHX_ cv); |
1661 | if (PL_stack_sp != PL_stack_base + 1) |
1662 | Perl_croak(aTHX_ "Sort subroutine didn't return single value"); |
1663 | if (!SvNIOKp(*PL_stack_sp)) |
1664 | Perl_croak(aTHX_ "Sort subroutine didn't return a numeric value"); |
1665 | result = SvIV(*PL_stack_sp); |
1666 | while (PL_scopestack_ix > oldscopeix) { |
1667 | LEAVE; |
1668 | } |
1669 | leave_scope(oldsaveix); |
1670 | return result; |
1671 | } |
1672 | |
1673 | |
1674 | static I32 |
1675 | sv_ncmp(pTHX_ SV *a, SV *b) |
1676 | { |
1677 | NV nv1 = SvNV(a); |
1678 | NV nv2 = SvNV(b); |
1679 | return nv1 < nv2 ? -1 : nv1 > nv2 ? 1 : 0; |
1680 | } |
1681 | |
1682 | static I32 |
1683 | sv_i_ncmp(pTHX_ SV *a, SV *b) |
1684 | { |
1685 | IV iv1 = SvIV(a); |
1686 | IV iv2 = SvIV(b); |
1687 | return iv1 < iv2 ? -1 : iv1 > iv2 ? 1 : 0; |
1688 | } |
1689 | #define tryCALL_AMAGICbin(left,right,meth,svp) STMT_START { \ |
1690 | *svp = Nullsv; \ |
1691 | if (PL_amagic_generation) { \ |
1692 | if (SvAMAGIC(left)||SvAMAGIC(right))\ |
1693 | *svp = amagic_call(left, \ |
1694 | right, \ |
1695 | CAT2(meth,_amg), \ |
1696 | 0); \ |
1697 | } \ |
1698 | } STMT_END |
1699 | |
1700 | static I32 |
1701 | amagic_ncmp(pTHX_ register SV *a, register SV *b) |
1702 | { |
1703 | SV *tmpsv; |
1704 | tryCALL_AMAGICbin(a,b,ncmp,&tmpsv); |
1705 | if (tmpsv) { |
1706 | NV d; |
4eb872f6 |
1707 | |
84d4ea48 |
1708 | if (SvIOK(tmpsv)) { |
1709 | I32 i = SvIVX(tmpsv); |
1710 | if (i > 0) |
1711 | return 1; |
1712 | return i? -1 : 0; |
1713 | } |
1714 | d = SvNV(tmpsv); |
1715 | if (d > 0) |
1716 | return 1; |
1717 | return d? -1 : 0; |
1718 | } |
1719 | return sv_ncmp(aTHX_ a, b); |
1720 | } |
1721 | |
1722 | static I32 |
1723 | amagic_i_ncmp(pTHX_ register SV *a, register SV *b) |
1724 | { |
1725 | SV *tmpsv; |
1726 | tryCALL_AMAGICbin(a,b,ncmp,&tmpsv); |
1727 | if (tmpsv) { |
1728 | NV d; |
4eb872f6 |
1729 | |
84d4ea48 |
1730 | if (SvIOK(tmpsv)) { |
1731 | I32 i = SvIVX(tmpsv); |
1732 | if (i > 0) |
1733 | return 1; |
1734 | return i? -1 : 0; |
1735 | } |
1736 | d = SvNV(tmpsv); |
1737 | if (d > 0) |
1738 | return 1; |
1739 | return d? -1 : 0; |
1740 | } |
1741 | return sv_i_ncmp(aTHX_ a, b); |
1742 | } |
1743 | |
1744 | static I32 |
1745 | amagic_cmp(pTHX_ register SV *str1, register SV *str2) |
1746 | { |
1747 | SV *tmpsv; |
1748 | tryCALL_AMAGICbin(str1,str2,scmp,&tmpsv); |
1749 | if (tmpsv) { |
1750 | NV d; |
4eb872f6 |
1751 | |
84d4ea48 |
1752 | if (SvIOK(tmpsv)) { |
1753 | I32 i = SvIVX(tmpsv); |
1754 | if (i > 0) |
1755 | return 1; |
1756 | return i? -1 : 0; |
1757 | } |
1758 | d = SvNV(tmpsv); |
1759 | if (d > 0) |
1760 | return 1; |
1761 | return d? -1 : 0; |
1762 | } |
1763 | return sv_cmp(str1, str2); |
1764 | } |
1765 | |
1766 | static I32 |
1767 | amagic_cmp_locale(pTHX_ register SV *str1, register SV *str2) |
1768 | { |
1769 | SV *tmpsv; |
1770 | tryCALL_AMAGICbin(str1,str2,scmp,&tmpsv); |
1771 | if (tmpsv) { |
1772 | NV d; |
4eb872f6 |
1773 | |
84d4ea48 |
1774 | if (SvIOK(tmpsv)) { |
1775 | I32 i = SvIVX(tmpsv); |
1776 | if (i > 0) |
1777 | return 1; |
1778 | return i? -1 : 0; |
1779 | } |
1780 | d = SvNV(tmpsv); |
1781 | if (d > 0) |
1782 | return 1; |
1783 | return d? -1 : 0; |
1784 | } |
1785 | return sv_cmp_locale(str1, str2); |
1786 | } |