File: pl-list.c

package info (click to toggle)
swi-prolog 6.6.6-1~bpo70+1
  • links: PTS, VCS
  • area: main
  • in suites: wheezy-backports
  • size: 82,312 kB
  • sloc: ansic: 322,250; perl: 245,822; sh: 6,651; java: 5,254; makefile: 4,423; cpp: 4,153; ruby: 1,594; yacc: 843; xml: 82; sed: 12; sql: 6
file content (474 lines) | stat: -rw-r--r-- 11,355 bytes parent folder | download | duplicates (2)
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
/*  $Id$

    Part of SWI-Prolog

    Author:        Jan Wielemaker
    E-mail:        jan@swi.psy.uva.nl
    WWW:           http://www.swi-prolog.org
    Copyright (C): 1985-2002, University of Amsterdam

    This library is free software; you can redistribute it and/or
    modify it under the terms of the GNU Lesser General Public
    License as published by the Free Software Foundation; either
    version 2.1 of the License, or (at your option) any later version.

    This library is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
    Lesser General Public License for more details.

    You should have received a copy of the GNU Lesser General Public
    License along with this library; if not, write to the Free Software
    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
*/

#include "pl-incl.h"

#undef LD
#define LD LOCAL_LD

static
PRED_IMPL("is_list", 1, is_list, 0)
{ if ( lengthList(A1, FALSE) >= 0 )
    succeed;

  fail;
}


/** $length(-List, +Len) is semidet.

Implements `known-length' generation path of length/2. Fails if Len < 0.
*/

static
PRED_IMPL("$length", 2, dlength, 0)
{ PRED_LD
  intptr_t len;

  if ( PL_get_intptr(A2, &len) )
  { if ( len > 0 )
    { Word p;
      term_t list = PL_new_term_ref();

      if ( !hasGlobalSpace(len*3) )
      { int rc;

	if ( (rc=ensureGlobalSpace(len*3, ALLOW_GC)) != TRUE )
	  return raiseStackOverflow(rc);
      }

      p = gTop;
      *valTermRef(list) = consPtr(p, TAG_COMPOUND|STG_GLOBAL);
      while(len-- > 0)
      { p[0] = FUNCTOR_dot2;
	setVar(p[1]);
	p[2] = consPtr(&p[3], TAG_COMPOUND|STG_GLOBAL);
	p += 3;
      }
      p[-1] = ATOM_nil;
      gTop = p;

      return PL_unify(A1, list);
    } else if ( len == 0 )
    { return PL_unify_nil(A1);
    } else
    { return FALSE;
    }
  } else if ( PL_is_integer(A2) )
  { number i;
    Word p =  valTermRef(A2);

    deRef(p);
    get_integer(*p, &i);
    if ( ar_sign_i(&i) < 0 )
      return FALSE;

    return outOfStack((Stack)&LD->stacks.global, STACK_OVERFLOW_RAISE);
  }

  return PL_error("length", 2, NULL, ERR_TYPE, ATOM_integer, A2);
}


static
PRED_IMPL("memberchk", 2, memberchk, 0)
{ GET_LD
  term_t h = PL_new_term_ref();
  term_t l = PL_copy_term_ref(A2);
  fid_t fid;

  if ( !(fid=PL_open_foreign_frame()) )
    return FALSE;

  for(;;)
  { if ( !PL_unify_list(l, h, l) )
    { PL_close_foreign_frame(fid);
      PL_unify_nil_ex(l);
      return FALSE;
    }

    if ( PL_unify(A1, h) )
    { term_t ex = 0;

      if ( foreignWakeup(&ex PASS_LD) )
      { PL_close_foreign_frame(fid);
	succeed;
      } else
      { if ( ex )
	  return PL_raise_exception(ex);
	PL_rewind_foreign_frame(fid);
      }
    } else
    { PL_rewind_foreign_frame(fid);
    }
  }
}


		 /*******************************
		 *	      SORTING		*
		 *******************************/

/* - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Natural merge sort. Code contributed by   Richard O'Keefe and integrated
into SWI-Prolog by Jan Wielemaker. The  nice   point  about this code is
that it uses no  extra  space  and   is  pretty  stable  in performance.
Richards claim it that many  qsort()   implementations  in libc are very
slow. This isn't the case for glibc   2.2, where this performs about the
same  as  the  previous  qsort()    based  implementation.  However,  it
integrated keysort/2 in the set and here the difference is huge.

Here is C code implementing a bottom-up  natural merge sort on lists; it
has remove_dups and compare_keys options.   (Actually  I wouldn't handle
the compare_keys option quite like this.)   The  difference between this
and sam-sort is the way runs are built:

    natural merge:
        add new node r after last node q of run if item(q) <= item(r)
        otherwise end this run.

    sam-sort:
        add new node r after last node q of run if item(q) <= item(r)
        otherwise
            add new new r before first node p of run if item(r) < item(p)
            otherwise end this run.

The natural merge has the nice  property   that  if  the list is already
sorted it takes O(N) time. In  general  if   you  have  a list made of M
already sorted pieces S1++S2++...++SM it will  take no more than O(N.log
M). Sam-sort (for "Smooth Applicative Merge sort") has the nice property
that it likes the reverse order almost as much as forward order, so \ /\
and \/ patterns are sorted (nearly) as  fast   as  /  // and // patterns
respectively.

I've been using a variant of this code  in a sorting utility since about
1988. It leaves the UNIX sort(1) program in   the dust. As you may know,
sort(1) breaks the input into  blocks  that   fit  in  memory, sorts the
blocks using qsort(), and writes the blocks out to disc, then merges the
blocks. For files that fit into memory,   the  variant of this code runs
about twice as fast as sort(1). Part of  that is better I/O, but part is
just plain not using qsort().
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - */


/*  Things in capital letters should be replaced for different applications  */

/*  ITEM	The type of an individual item.
    COMPARE	Compares two items given their addresses (allows ITEM to be
		large and avoids pass by copy).  Return <0, =0, or >0.
    COMPARE_KEY	Compares the keys of two items given the addresses of the
		entire items.
    FREE	Frees a List_Record including its ITEM.
*/

typedef struct
{ Word term;
  Word key;
} ITEM;

					/* TBD: handle CMP_ERROR */
#ifndef COMPARE
#define COMPARE(x,y) compareStandard((x)->term, (y)->term, FALSE PASS_LD)
#endif
#ifndef COMPARE_KEY
#define COMPARE_KEY(x,y) compareStandard((x)->key, (y)->key, FALSE PASS_LD)
#endif
#ifndef FREE
#define FREE(x) \
	{ x->next = NULL; \
	  x->item.term = NULL; \
	  x->item.key = NULL; \
	}
#endif

typedef struct List_Record *list;
struct List_Record {
    list next;
    ITEM item;
};

#define NIL (list)0

#define compare(c, x, y) \
    int c = compare_keys ? COMPARE_KEY(&(x)->item, &(y)->item) \
                         : COMPARE(    &(x)->item, &(y)->item)

static list
nat_sort(list data, int remove_dups, int compare_keys)
{ GET_LD
  list stack[64];			/* enough for biggest machine */
  list *sp = stack;
  int runs = 0;				/* total number of runs processed */
  list p, q, r, s;
  struct List_Record header;
  int k;

  remove_dups = !remove_dups;		/* 0 -> do, 1 -> don't */
  while ((p = data) != NIL)
  { /* pick up a run from the front of data, setting */
    /* p = (pointer to beginning of run), data = (rest of data) */
    if ((q = p->next) != NIL)
    { compare(c, p, q);

      data = q->next;
      if (c > 0)
      { r = q, q = p, p = r;
	p->next = q;
      } else if (c == remove_dups)
      {	/* c < 0 or = 0, so c = 1 impossible */
	p->next = q->next;
	FREE(q);
	q = p;
      }

      for (r = data; r != NIL; )
      { compare(c, q, r);

	if (c > 0)
	  break;
	if (c == remove_dups)
	{ s = r->next;
	  FREE(r);
	  r = s;
	} else
	{ q->next = r, q = r, r = r->next;
	}
      }

      q->next = NIL;
      data = r;
    } else
    { data = NIL;
    }

    runs++;
    /* merge this run with 0 or more runs off the top of the stack */
    for (k = runs; 1 &~ k; k >>= 1)
    { q = *--sp;
      r = &header;
      while (q && p)
      {	/* q precedes p */
	compare(c, q, p);

	if (c <= 0)
	{ r->next = q, r = q, q = q->next;
	  if (c == remove_dups)
	  { s = p->next;
	    FREE(p);
	    p = s;
	  }
	} else
	{ r->next = p, r = p, p = p->next;
	}
      }
      r->next = q ? q : p;
      p = header.next;
    }

	 /* push the merged run onto the stack */
    *sp++ = p;
  }

  if (sp == stack)
    return NIL;

  /* merge all the runs on the stack */
  p = *--sp;
  while (sp != stack)
  { q = *--sp;
    r = &header;
    while (q && p)
    {	/* q precedes p */
      compare(c, q, p);

      if (c <= 0)
      { r->next = q, r = q, q = q->next;
	if (c == remove_dups)
	{ s = p->next;
	  FREE(p);
	  p = s;
	}
      } else
      { r->next = p, r = p, p = p->next;
      }
    }
    r->next = q ? q : p;
    p = header.next;
  }

  return p;
}


/* - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Create a list on the global stack, just   at  the place the final result
will be.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - */

static int
prolog_list_to_sort_list(term_t t, int remove_dups, int key,
			 list *lp, Word *end)
{ GET_LD
  Word l, tail;
  list p;
  intptr_t len;
  int rc;

  l = valTermRef(t);
  len = skip_list(l, &tail PASS_LD);
  if ( !(isNil(*tail) ||			/* proper list */
	 (isList(*tail) && remove_dups)) )	/* sort/2 on cyclic list */
  {
    if ( isVar(*tail) )
      return PL_error(NULL, 0, NULL, ERR_INSTANTIATION);
    else
      return PL_error(NULL, 0, NULL, ERR_TYPE, ATOM_list, t);
  }

  if ( !hasGlobalSpace(len*3) )
  { if ( (rc=ensureGlobalSpace(len*3, ALLOW_GC)) != TRUE )
      return raiseStackOverflow(rc);
    l = valTermRef(t);			/* may be shifted */
    deRef(l);
  }

  p = (list)gTop;
  *lp = p;

  while(len-- > 0)
  { p->item.term = HeadList(l);
    deRef(p->item.term);
    if ( key )
    { word w = *p->item.term;

      if ( hasFunctor(w, FUNCTOR_minus2) )
      { p->item.key = argTermP(w, 0);
	deRef(p->item.key);
      } else
      { PL_error("keysort", 2, NULL, ERR_TYPE,
		 ATOM_pair, pushWordAsTermRef(p->item.term));
	popTermRef();
	return FALSE;
      }
    }
    l = TailList(l);
    deRef(l);
    if ( len > 0 )
    { assert(isList(*l));
      p->next = p+1;
      p++;
    }
  }

  p->next = NULL;
  *end = (Word)(p+1);

  succeed;
}


static void
put_sort_list(term_t l, list sl)
{ GET_LD

  *valTermRef(l) = consPtr(sl, TAG_COMPOUND|STG_GLOBAL);

  for(;;)
  { list n = sl->next;
    Word p = (Word)sl;

    n = sl->next;
					/* see also linkVal() */
    p[1] = (needsRef(*sl->item.term) ? makeRef(sl->item.term)
				     : *sl->item.term);
    p[0] = FUNCTOR_dot2;
    if ( n )
    { p[2] = consPtr(n, TAG_COMPOUND|STG_GLOBAL);
      sl = n;
    } else
    { p[2] = ATOM_nil;
      return;
    }
  }
}


static int
pl_nat_sort(term_t in, term_t out, int remove_dups, int compare_keys ARG_LD)
{ if ( PL_get_nil(in) )
    return PL_unify_atom(out, ATOM_nil);
  else
  { list l = 0;
    term_t tmp = PL_new_term_ref();
    Word top = NULL;

    if ( prolog_list_to_sort_list(in, remove_dups, compare_keys, &l, &top) )
    { l = nat_sort(l, remove_dups, compare_keys);
      put_sort_list(tmp, l);
      gTop = top;

      return PL_unify(out, tmp);
    }

    fail;
  }
}


static
PRED_IMPL("sort", 2, sort, PL_FA_ISO)
{ PRED_LD

  return pl_nat_sort(A1, A2, TRUE, FALSE PASS_LD);
}


static
PRED_IMPL("msort", 2, msort, 0)
{ PRED_LD

  return pl_nat_sort(A1, A2, FALSE, FALSE PASS_LD);
}


static
PRED_IMPL("keysort", 2, keysort, PL_FA_ISO)
{ PRED_LD

  return pl_nat_sort(A1, A2, FALSE, TRUE PASS_LD);
}


		 /*******************************
		 *      PUBLISH PREDICATES	*
		 *******************************/

BeginPredDefs(list)
  PRED_DEF("is_list", 1, is_list, 0)
  PRED_DEF("$length", 2, dlength, 0)
  PRED_DEF("memberchk", 2, memberchk, 0)
  PRED_DEF("sort", 2, sort, PL_FA_ISO)
  PRED_DEF("msort", 2, msort, 0)
  PRED_DEF("keysort", 2, keysort, PL_FA_ISO)
EndPredDefs