summaryrefslogtreecommitdiffstats
path: root/strings/ctype-uca.inl
blob: 6aa72847c81c77530dfd67e9dfa37a338042209d (plain)
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
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
/*
  Copyright (c) 2018, 2020, MariaDB Corporation

  This program is free software; you can redistribute it and/or modify
  it under the terms of the GNU General Public License as published by
  the Free Software Foundation; version 2 of the License.

  This program 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 General Public License for more details.

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

#include "ctype-mb.h"

#ifndef MY_FUNCTION_NAME
#error MY_FUNCTION_NAME is not defined
#endif
#ifndef MY_MB_WC
#error MY_MB_WC is not defined
#endif
#ifndef MY_LIKE_RANGE
#error MY_LIKE_RANGE is not defined
#endif
#ifndef MY_UCA_ASCII_OPTIMIZE
#error MY_ASCII_OPTIMIZE is not defined
#endif
#ifndef MY_UCA_COMPILE_CONTRACTIONS
#error MY_UCA_COMPILE_CONTRACTIONS is not defined
#endif
#ifndef MY_UCA_COLL_INIT
#error MY_UCA_COLL_INIT is not defined
#endif


#include "ctype-uca-scanner_next.inl"
#define SCANNER_NEXT_NCHARS
#include "ctype-uca-scanner_next.inl"

/*
  Compares two strings according to the collation

  SYNOPSIS:
    strnncoll_onelevel()
    cs		Character set information
    level       Weight level (0 primary, 1 secondary, 2 tertiary, etc)
    s		First string
    slen	First string length
    t		Second string
    tlen	Seconf string length
    level	DUCETweight level
  
  NOTES:
    Initializes two weight scanners and gets weights
    corresponding to two strings in a loop. If weights are not
    the same at some step then returns their difference.
    
    In the while() comparison these situations are possible:
    1. (s_res>0) and (t_res>0) and (s_res == t_res)
       Weights are the same so far, continue comparison
    2. (s_res>0) and (t_res>0) and (s_res!=t_res)
       A difference has been found, return.
    3. (s_res>0) and (t_res<0)
       We have reached the end of the second string, or found
       an illegal multibyte sequence in the second string.
       Return a positive number, i.e. the first string is bigger.
    4. (s_res<0) and (t_res>0)   
       We have reached the end of the first string, or found
       an illegal multibyte sequence in the first string.
       Return a negative number, i.e. the second string is bigger.
    5. (s_res<0) and (t_res<0)
       Both scanners returned -1. It means we have riched
       the end-of-string of illegal-sequence in both strings
       at the same time. Return 0, strings are equal.
    
  RETURN
    Difference between two strings, according to the collation:
    0               - means strings are equal
    negative number - means the first string is smaller
    positive number - means the first string is bigger
*/

static int
MY_FUNCTION_NAME(strnncoll_onelevel)(CHARSET_INFO *cs, 
                                     const MY_UCA_WEIGHT_LEVEL *level,
                                     const uchar *s, size_t slen,
                                     const uchar *t, size_t tlen,
                                     my_bool t_is_prefix)
{
  my_uca_scanner sscanner;
  my_uca_scanner tscanner;
  my_uca_scanner_param param;
  int s_res;
  int t_res;

#if MY_UCA_ASCII_OPTIMIZE
{
  size_t prefix= my_uca_level_booster_equal_prefix_length(level->booster,
                                                          s, slen, t, tlen);
  s+= prefix, slen-= prefix;
  t+= prefix, tlen-= prefix;
}
#endif

  my_uca_scanner_param_init(&param, cs, level);
  my_uca_scanner_init_any(&sscanner, s, slen);
  my_uca_scanner_init_any(&tscanner, t, tlen);
  
  do
  {
    s_res= MY_FUNCTION_NAME(scanner_next)(&sscanner, &param);
    t_res= MY_FUNCTION_NAME(scanner_next)(&tscanner, &param);
  } while ( s_res == t_res && s_res >0);
  
  return  (t_is_prefix && t_res < 0) ? 0 : (s_res - t_res);
}


/*
  One-level, PAD SPACE.
*/
static int
MY_FUNCTION_NAME(strnncoll)(CHARSET_INFO *cs,
                            const uchar *s, size_t slen,
                            const uchar *t, size_t tlen,
                            my_bool t_is_prefix)
{
  return MY_FUNCTION_NAME(strnncoll_onelevel)(cs, &cs->uca->level[0],
                                              s, slen, t, tlen, t_is_prefix);
}


/*
  Multi-level, PAD SPACE.
*/
static int
MY_FUNCTION_NAME(strnncoll_multilevel)(CHARSET_INFO *cs,
                                       const uchar *s, size_t slen,
                                       const uchar *t, size_t tlen,
                                       my_bool t_is_prefix)
{
  uint i, level_flags= cs->levels_for_order;
  for (i= 0; level_flags; i++, level_flags>>= 1)
  {
    int ret;
    if (!(level_flags & 1))
      continue;
    ret= MY_FUNCTION_NAME(strnncoll_onelevel)(cs, &cs->uca->level[i],
                                              s, slen, t, tlen,
                                              t_is_prefix);
    if (ret)
       return ret;
  }
  return 0;
}


/*
  Compares two strings according to the collation,
  ignoring trailing spaces.

  SYNOPSIS:
    strnncollsp_onelevel()
    cs		Character set information
    level       UCA weight level
    s		First string
    slen	First string length
    t		Second string
    tlen	Seconf string length
    level	DUCETweight level

  NOTES:
    Works exactly the same with my_strnncoll_uca(),
    but ignores trailing spaces.

    In the while() comparison these situations are possible:
    1. (s_res>0) and (t_res>0) and (s_res == t_res)
       Weights are the same so far, continue comparison
    2. (s_res>0) and (t_res>0) and (s_res!=t_res)
       A difference has been found, return.
    3. (s_res>0) and (t_res<0)
       We have reached the end of the second string, or found
       an illegal multibyte sequence in the second string.
       Compare the first string to an infinite array of
       space characters until difference is found, or until
       the end of the first string.
    4. (s_res<0) and (t_res>0)
       We have reached the end of the first string, or found
       an illegal multibyte sequence in the first string.
       Compare the second string to an infinite array of
       space characters until difference is found or until
       the end of the second steing.
    5. (s_res<0) and (t_res<0)
       Both scanners returned -1. It means we have riched
       the end-of-string of illegal-sequence in both strings
       at the same time. Return 0, strings are equal.

  RETURN
    Difference between two strings, according to the collation:
    0               - means strings are equal
    negative number - means the first string is smaller
    positive number - means the first string is bigger
*/

static int
MY_FUNCTION_NAME(strnncollsp_onelevel)(CHARSET_INFO *cs,
                                       const MY_UCA_WEIGHT_LEVEL *level,
                                       const uchar *s, size_t slen,
                                       const uchar *t, size_t tlen)
{
  my_uca_scanner sscanner, tscanner;
  my_uca_scanner_param param;
  int s_res, t_res;

#if MY_UCA_ASCII_OPTIMIZE
{
  size_t prefix= my_uca_level_booster_equal_prefix_length(level->booster,
                                                          s, slen, t, tlen);
  s+= prefix, slen-= prefix;
  t+= prefix, tlen-= prefix;
}
#endif

  my_uca_scanner_param_init(&param, cs, level);
  my_uca_scanner_init_any(&sscanner, s, slen);
  my_uca_scanner_init_any(&tscanner, t, tlen);

  do
  {
    s_res= MY_FUNCTION_NAME(scanner_next)(&sscanner, &param);
    t_res= MY_FUNCTION_NAME(scanner_next)(&tscanner, &param);
  } while ( s_res == t_res && s_res >0);

  if (s_res > 0 && t_res < 0)
  {
    /* Calculate weight for SPACE character */
    t_res= my_space_weight(level);

    /* compare the first string to spaces */
    do
    {
      if (s_res != t_res)
        return (s_res - t_res);
      s_res= MY_FUNCTION_NAME(scanner_next)(&sscanner, &param);
    } while (s_res > 0);
    return 0;
  }

  if (s_res < 0 && t_res > 0)
  {
    /* Calculate weight for SPACE character */
    s_res= my_space_weight(level);

    /* compare the second string to spaces */
    do
    {
      if (s_res != t_res)
        return (s_res - t_res);
      t_res= MY_FUNCTION_NAME(scanner_next)(&tscanner, &param);
    } while (t_res > 0);
    return 0;
  }

  return ( s_res - t_res );
}


/*
  One-level, PAD SPACE
*/
static int
MY_FUNCTION_NAME(strnncollsp)(CHARSET_INFO *cs,
                              const uchar *s, size_t slen,
                              const uchar *t, size_t tlen)
{
  return MY_FUNCTION_NAME(strnncollsp_onelevel)(cs, &cs->uca->level[0],
                                                s, slen, t, tlen);
}


/*
  One-level, NO PAD
*/
static int
MY_FUNCTION_NAME(strnncollsp_nopad)(CHARSET_INFO *cs,
                                    const uchar *s, size_t slen,
                                    const uchar *t, size_t tlen)
{
  return MY_FUNCTION_NAME(strnncoll_onelevel)(cs, &cs->uca->level[0],
                                              s, slen, t, tlen, FALSE);
}


/*
  Multi-level, PAD SPACE
*/
static int
MY_FUNCTION_NAME(strnncollsp_multilevel)(CHARSET_INFO *cs,
                                         const uchar *s, size_t slen,
                                         const uchar *t, size_t tlen)
{
  uint i, level_flags= cs->levels_for_order;
  for (i= 0; level_flags; i++, level_flags>>= 1)
  {
    int ret;
    if (!(level_flags & 1))
      continue;
    ret= MY_FUNCTION_NAME(strnncollsp_onelevel)(cs, &cs->uca->level[i],
                                                s, slen, t, tlen);
    if (ret)
      return ret;
  }
  return 0;
}


/*
  Multi-level, NO PAD
*/
static int
MY_FUNCTION_NAME(strnncollsp_nopad_multilevel)(CHARSET_INFO *cs,
                                               const uchar *s, size_t slen,
                                               const uchar *t, size_t tlen)
{
  uint i, level_flags;
  int ret;

  /* Compare only the primary level using NO PAD */
  if ((ret= MY_FUNCTION_NAME(strnncoll_onelevel)(cs, &cs->uca->level[0],
                                                 s, slen, t, tlen, FALSE)))
    return ret;

  /*
    Compare the other levels using PAD SPACE.
    These are Unicode-14.0.0 DUCTET weights:

0020  ; [*0209.0020.0002] # SPACE

0035  ; [.2070.0020.0002] # DIGIT FIVE
248C  ; [.2070.0020.0004][*0281.0020.0004] # DIGIT FIVE FULL STOP

0041  ; [.2075.0020.0008] # LATIN CAPITAL LETTER A
0061  ; [.2075.0020.0002] # LATIN SMALL LETTER A
00C1  ; [.2075.0020.0008][.0000.0024.0002] # LATIN CAPITAL LETTER A WITH ACUTE
00E1  ; [.2075.0020.0002][.0000.0024.0002] # LATIN SMALL LETTER A WITH ACUTE

    Examples demonstrating that it's important to use PAD SPACE
    on the tertiary level:

    The third level weights for "SMALL LETTER A"
    - U+0061 produces one weight 0002
    - U+00E1 produces two weights 0002+0002
      For _ai_cs collations these two letters must be equal.
      Therefore, the difference in trailing 0002 should be ignored.

    The third level weights for "CAPITAL LETTER A"
    - U+0041 produces one weight 0008
    - U+00C1 produces two weights 0008+0002
      For _ai_cs collations these two letters must be equal.
      Therefore, the difference in trailing 0002 should be ignored.

    Examples demonstrating that it's important to use PAD SPACE
    on the secondary level:

    When we implement variable shifted alternative weighting collations,
    U+0035 will be equal to U+248C on the primary level in these collations.
    The second level weights for "DIGIT FIVE" are:
    - U+0035 produces one weight 0020
    - U+248C produces two weights 0020+0020.
    The difference for these two characters must be found only
    on the tertiary level. Therefore, the trailing 0020 should be ignored.
  */

  for (i= 1, level_flags= cs->levels_for_order >> 1;
       level_flags;
       i++, level_flags>>= 1)
  {
    if (!(level_flags & 1))
      continue;
    ret= MY_FUNCTION_NAME(strnncollsp_onelevel)(cs, &cs->uca->level[i],
                                                s, slen, t, tlen);
    if (ret)
       return ret;
  }
  return 0;
}


/*
  Scan the next weight and perform space padding
  or trimming according to "nchars".
*/
static inline weight_and_nchars_t
MY_FUNCTION_NAME(scanner_next_pad_trim)(my_uca_scanner *scanner,
                                        my_uca_scanner_param *param,
                                        size_t nchars,
                                        uint flags,
                                        uint *generated)
{
  weight_and_nchars_t res;
  if (nchars > 0 ||
      scanner->wbeg[0] /* Some weights from a previous expansion left */)
  {
    if ((res= MY_FUNCTION_NAME(scanner_next_with_nchars)(scanner, param,
                                                         nchars)).weight < 0)
    {
      /*
        We reached the end of the string, but the caller wants more weights.
        Perform space padding.
      */
      res.weight=
        flags & MY_STRNNCOLLSP_NCHARS_EMULATE_TRIMMED_TRAILING_SPACES ?
        my_space_weight(param->level) : 0;

      res.nchars= 1;
      (*generated)++;
    }
    else if (res.nchars > nchars)
    {
      /*
        We scanned the next collation element, but it does not fit into
        the "nchars" limit. This is possible in case of:
        - A contraction, e.g. Czech 'ch' with nchars=1
        - A sequence of ignorable characters followed by non-ignorable ones,
          e.g. CONCAT(x'00','a') with nchars=1.
        Perform trimming.
      */
      res.weight= param->cs->state & MY_CS_NOPAD ?
                  0 : my_space_weight(param->level);
      res.nchars= (uint) nchars;
      (*generated)++;
    }
  }
  else
  {
    /* The caller wants nchars==0. Perform trimming. */
    res.weight= param->cs->state & MY_CS_NOPAD ?
                0 : my_space_weight(param->level);
    res.nchars= 0;
    (*generated)++;
  }
  return res;
}


static int
MY_FUNCTION_NAME(strnncollsp_nchars_onelevel)(CHARSET_INFO *cs,
                                              const MY_UCA_WEIGHT_LEVEL *level,
                                              const uchar *s, size_t slen,
                                              const uchar *t, size_t tlen,
                                              size_t nchars,
                                              uint flags)
{
  my_uca_scanner sscanner;
  my_uca_scanner tscanner;
  my_uca_scanner_param param;
  size_t s_nchars_left= nchars;
  size_t t_nchars_left= nchars;

/*
TODO: strnncollsp_nchars_onelevel
#if MY_UCA_ASCII_OPTIMIZE
{
  size_t prefix= my_uca_level_booster_equal_prefix_length(level->booster,
                                                          s, slen, t, tlen);
  s+= prefix, slen-= prefix;
  t+= prefix, tlen-= prefix;
}
#endif
*/

  my_uca_scanner_param_init(&param, cs, level);
  my_uca_scanner_init_any(&sscanner, s, slen);
  my_uca_scanner_init_any(&tscanner, t, tlen);

  for ( ; ; )
  {
    weight_and_nchars_t s_res;
    weight_and_nchars_t t_res;
    uint generated= 0;
    int diff;

    s_res= MY_FUNCTION_NAME(scanner_next_pad_trim)(&sscanner, &param,
                                                   s_nchars_left,
                                                   flags, &generated);
    t_res= MY_FUNCTION_NAME(scanner_next_pad_trim)(&tscanner, &param,
                                                   t_nchars_left,
                                                   flags, &generated);
    if ((diff= (s_res.weight - t_res.weight)))
      return diff;

    if (generated == 2)
    {
      if ((cs->state & MY_CS_NOPAD) &&
          (flags & MY_STRNNCOLLSP_NCHARS_EMULATE_TRIMMED_TRAILING_SPACES))
      {
        /*
          Both values are auto-generated. There's no real data any more.
          We need to handle the remaining virtual trailing spaces.
          The two strings still have s_nchars_left and t_nchars_left imaginary
          trailing spaces at the end. If s_nchars_left != t_nchars_left,
          the strings will be not equal in case of a NOPAD collation.

          Example:
          "B" is German "U+00DF LATIN SMALL LETTER SHARP S"
          When we have these values in a
          CHAR(3) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_nopad_ci
          column:
          'B  '        (one character, two trailing spaces)
          'ss '        (two characters, one trailing space)
          The 'B  ' is greater than the 'ss '.
          They are compared in the following steps:
            1. 'B' == 'ss'
            2. ' ' == ' '
            3. ' ' >   ''

          We need to emulate the same behavior in this function even if
          it's called with strings 'B' and 'ss' (with space trimmed).
          The side which has more remaining virtual spaces at the end
          is greater.
        */
        if (s_nchars_left < t_nchars_left)
          return -1;
        if (s_nchars_left > t_nchars_left)
          return +1;
      }
      return 0;
    }

    DBUG_ASSERT(s_nchars_left >= s_res.nchars);
    DBUG_ASSERT(t_nchars_left >= t_res.nchars);
    s_nchars_left-= s_res.nchars;
    t_nchars_left-= t_res.nchars;
  }

  return 0;
}


/*
  One-level collations.
*/
static int
MY_FUNCTION_NAME(strnncollsp_nchars)(CHARSET_INFO *cs,
                                     const uchar *s, size_t slen,
                                     const uchar *t, size_t tlen,
                                     size_t nchars,
                                     uint flags)
{
  return MY_FUNCTION_NAME(strnncollsp_nchars_onelevel)(cs, &cs->uca->level[0],
                                                       s, slen, t, tlen,
                                                       nchars, flags);
}


/*
  Multi-level collations.
*/
static int
MY_FUNCTION_NAME(strnncollsp_nchars_multilevel)(CHARSET_INFO *cs,
                                                const uchar *s, size_t slen,
                                                const uchar *t, size_t tlen,
                                                size_t nchars,
                                                uint flags)
{
  uint i, level_flags= cs->levels_for_order;
  for (i= 0; level_flags; i++, level_flags>>= 1)
  {
    int ret;
    if (!(level_flags & 1))
      continue;
    ret= MY_FUNCTION_NAME(strnncollsp_nchars_onelevel)(cs,
                                                       &cs->uca->level[i],
                                                       s, slen,
                                                       t, tlen,
                                                       nchars, flags);
    if (ret)
       return ret;
  }
  return 0;
}


/*
  Calculates hash value for the given string,
  according to the collation, and ignoring trailing spaces.

  SYNOPSIS:
    hash_sort()
    cs		Character set information
    s		String
    slen	String's length
    n1		First hash parameter
    n2		Second hash parameter

  NOTES:
    Scans consequently weights and updates
    hash parameters n1 and n2. In a case insensitive collation,
    upper and lower case of the same letter will return the same
    weight sequence, and thus will produce the same hash values
    in n1 and n2.

    This functions is used for one-level and for multi-level collations.
    We intentionally use only primary level in multi-level collations.
    This helps to have PARTITION BY KEY put primarily equal records
    into the same partition. E.g. in utf8mb3_thai_520_ci records that differ
    only in tone marks go into the same partition.

  RETURN
    N/A
*/

static void
MY_FUNCTION_NAME(hash_sort)(CHARSET_INFO *cs,
                            const uchar *s, size_t slen,
                            ulong *nr1, ulong *nr2)
{
  int   s_res;
  my_uca_scanner scanner;
  my_uca_scanner_param param;
  int space_weight= my_space_weight(&cs->uca->level[0]);
  register ulong m1= *nr1, m2= *nr2;

  my_uca_scanner_param_init(&param, cs, &cs->uca->level[0]);
  my_uca_scanner_init_any(&scanner, s, slen);

  while ((s_res= MY_FUNCTION_NAME(scanner_next)(&scanner, &param)) >0)
  {
    if (s_res == space_weight)
    {
      /* Combine all spaces to be able to skip end spaces */
      uint count= 0;
      do
      {
        count++;
        if ((s_res= MY_FUNCTION_NAME(scanner_next)(&scanner, &param)) <= 0)
        {
          /* Skip strings at end of string */
          goto end;
        }
      }
      while (s_res == space_weight);

      /* Add back that has for the space characters */
      do
      {
        /*
          We can't use MY_HASH_ADD_16() here as we, because of a misstake
          in the original code, where we added the 16 byte variable the
          opposite way.  Changing this would cause old partitioned tables
          to fail.
        */
        MY_HASH_ADD(m1, m2, space_weight >> 8);
        MY_HASH_ADD(m1, m2, space_weight & 0xFF);
      }
      while (--count != 0);

    }
    /* See comment above why we can't use MY_HASH_ADD_16() */
    MY_HASH_ADD(m1, m2, s_res >> 8);
    MY_HASH_ADD(m1, m2, s_res & 0xFF);
  }
end:
  *nr1= m1;
  *nr2= m2;
}


static void
MY_FUNCTION_NAME(hash_sort_nopad)(CHARSET_INFO *cs,
                                  const uchar *s, size_t slen,
                                  ulong *nr1, ulong *nr2)
{
  int   s_res;
  my_uca_scanner scanner;
  my_uca_scanner_param param;
  register ulong m1= *nr1, m2= *nr2;

  my_uca_scanner_param_init(&param, cs, &cs->uca->level[0]);
  my_uca_scanner_init_any(&scanner, s, slen);

  while ((s_res= MY_FUNCTION_NAME(scanner_next)(&scanner, &param)) >0)
  {
    /* See comment above why we can't use MY_HASH_ADD_16() */
    MY_HASH_ADD(m1, m2, s_res >> 8);
    MY_HASH_ADD(m1, m2, s_res & 0xFF);
  }
  *nr1= m1;
  *nr2= m2;
}



/*
  For the given string creates its "binary image", suitable
  to be used in binary comparison, i.e. in memcmp(). 
  
  SYNOPSIS:
    my_strnxfrm_uca()
    cs		Character set information
    dst		Where to write the image
    dstlen	Space available for the image, in bytes
    src		The source string
    srclen	Length of the source string, in bytes
  
  NOTES:
    In a loop, scans weights from the source string and writes
    them into the binary image. In a case insensitive collation,
    upper and lower cases of the same letter will produce the
    same image subsequences. When we have reached the end-of-string
    or found an illegal multibyte sequence, the loop stops.

    It is impossible to restore the original string using its
    binary image. 
    
    Binary images are used for bulk comparison purposes,
    e.g. in ORDER BY, when it is more efficient to create
    a binary image and use it instead of weight scanner
    for the original strings for every comparison.
  
  RETURN
    Number of bytes that have been written into the binary image.
*/

static uchar *
MY_FUNCTION_NAME(strnxfrm_onelevel_internal)(CHARSET_INFO *cs,
                                             MY_UCA_WEIGHT_LEVEL *level,
                                             uchar *dst, uchar *de,
                                             uint *nweights,
                                             const uchar *src, size_t srclen)
{
  my_uca_scanner scanner;
  my_uca_scanner_param param;
  int s_res;

  DBUG_ASSERT(src || !srclen);

#if MY_UCA_ASCII_OPTIMIZE && !MY_UCA_COMPILE_CONTRACTIONS
 /*
    Fast path for the ASCII range with no contractions.
  */
  {
    const uchar *de2= de - 1; /* Last position where 2 bytes fit */
    const uint16 *weights0= level->weights[0];
    uint lengths0= level->lengths[0];
    for ( ; ; src++, srclen--)
    {
      const uint16 *weight;
      if (!srclen || !*nweights)
        return dst;         /* Done */
      if (*src > 0x7F)
        break;              /* Non-ASCII */

      weight= weights0 + (((uint) *src) * lengths0);
      if (!(s_res= *weight))
        continue;           /* Ignorable */
      if (weight[1])        /* Expansion (e.g. in a user defined collation */
        break;

      /* Here we have a character with extactly one 2-byte UCA weight */
      if (dst < de2)        /* Most typical case is when both bytes fit */
      {
        *dst++= s_res >> 8;
        *dst++= s_res & 0xFF;
        (*nweights)--;
        continue;
      }
      if (dst >= de)        /* No space left in "dst" */
        return dst;
      *dst++= s_res >> 8;   /* There is space only for one byte */
      (*nweights)--;
      return dst;
    }
  }
#endif

  my_uca_scanner_param_init(&param, cs, level);
  my_uca_scanner_init_any(&scanner, src, srclen);

  for (; dst < de && *nweights &&
         (s_res= MY_FUNCTION_NAME(scanner_next)(&scanner, &param)) > 0 ;
       (*nweights)--)
  {
    *dst++= s_res >> 8;
    if (dst < de)
      *dst++= s_res & 0xFF;
  }
  return dst;
}


static uchar *
MY_FUNCTION_NAME(strnxfrm_onelevel)(CHARSET_INFO *cs,
                                    MY_UCA_WEIGHT_LEVEL *level,
                                    uchar *dst, uchar *de, uint nweights,
                                    const uchar *src, size_t srclen, uint flags)
{
  uchar *d0= dst;
  dst= MY_FUNCTION_NAME(strnxfrm_onelevel_internal)(cs, level,
                                                    dst, de, &nweights,
                                                    src, srclen);
  DBUG_ASSERT(dst <= de);
  if (dst < de && nweights && (flags & MY_STRXFRM_PAD_WITH_SPACE))
    dst= my_strnxfrm_uca_padn(dst, de, nweights, my_space_weight(level));
  DBUG_ASSERT(dst <= de);
  my_strxfrm_desc_and_reverse(d0, dst, flags, 0);
  return dst;
}



static uchar *
MY_FUNCTION_NAME(strnxfrm_nopad_onelevel)(CHARSET_INFO *cs,
                                          MY_UCA_WEIGHT_LEVEL *level,
                                          uchar *dst, uchar *de, uint nweights,
                                          const uchar *src, size_t srclen,
                                          uint flags)
{
  uchar *d0= dst;
  dst= MY_FUNCTION_NAME(strnxfrm_onelevel_internal)(cs, level,
                                                    dst, de, &nweights,
                                                    src, srclen);
  DBUG_ASSERT(dst <= de);
  /*  Pad with the minimum possible weight on this level */
  if (dst < de && nweights && (flags & MY_STRXFRM_PAD_WITH_SPACE))
    dst= my_strnxfrm_uca_padn(dst, de, nweights, min_weight_on_level(level));
  DBUG_ASSERT(dst <= de);
  my_strxfrm_desc_and_reverse(d0, dst, flags, 0);
  return dst;
}


static size_t
MY_FUNCTION_NAME(strnxfrm)(CHARSET_INFO *cs,
                           uchar *dst, size_t dstlen, uint nweights,
                           const uchar *src, size_t srclen, uint flags)
{
  uchar *d0= dst;
  uchar *de= dst + dstlen;

  /*
    There are two ways to handle trailing spaces for PAD SPACE collations:
    1. Keep trailing spaces as they are, so have strnxfrm_onelevel() scan
       spaces as normal characters. This will call scanner_next() for every
       trailing space and calculate its weight using UCA weights.
    2. Strip trailing spaces before calling strnxfrm_onelevel(), as it will
       append weights for implicit spaces anyway, up to the desired key size.
       This will effectively generate exactly the same sortable key result.
    The latter is much faster.
  */

  if (flags & MY_STRXFRM_PAD_WITH_SPACE)
    srclen= my_ci_lengthsp(cs, (const char*) src, srclen);
  dst= MY_FUNCTION_NAME(strnxfrm_onelevel)(cs, &cs->uca->level[0],
                                           dst, de, nweights,
                                           src, srclen, flags);
  /*
    This can probably be changed to memset(dst, 0, de - dst),
    like my_strnxfrm_uca_multilevel() does.
  */
  if ((flags & MY_STRXFRM_PAD_TO_MAXLEN) && dst < de)
    dst= my_strnxfrm_uca_pad(dst, de, my_space_weight(&cs->uca->level[0]));
  return dst - d0;
}


static size_t
MY_FUNCTION_NAME(strnxfrm_nopad)(CHARSET_INFO *cs,
                                 uchar *dst, size_t dstlen,
                                 uint nweights,
                                 const uchar *src, size_t srclen,
                                 uint flags)
{
  uchar *d0= dst;
  uchar *de= dst + dstlen;

  dst= MY_FUNCTION_NAME(strnxfrm_nopad_onelevel)(cs, &cs->uca->level[0],
                                                 dst, de, nweights,
                                                 src, srclen, flags);
  if ((flags & MY_STRXFRM_PAD_TO_MAXLEN) && dst < de)
  {
    memset(dst, 0, de - dst);
    dst= de;
  }
  return dst - d0;
}


static size_t
MY_FUNCTION_NAME(strnxfrm_multilevel)(CHARSET_INFO *cs, 
                                      uchar *dst, size_t dstlen,
                                      uint nweights,
                                      const uchar *src, size_t srclen,
                                      uint flags)
{
  uint level_flags= cs->levels_for_order;
  uchar *d0= dst;
  uchar *de= dst + dstlen;
  uint current_level;

  for (current_level= 0; level_flags; current_level++, level_flags>>= 1)
  {
    if (!(level_flags & 1))
      continue;
    if (!(flags & MY_STRXFRM_LEVEL_ALL) ||
        (flags & (MY_STRXFRM_LEVEL1 << current_level)))
      dst= cs->state & MY_CS_NOPAD ?
           MY_FUNCTION_NAME(strnxfrm_nopad_onelevel)(cs,
                                          &cs->uca->level[current_level],
                                          dst, de, nweights,
                                          src, srclen, flags) :
           MY_FUNCTION_NAME(strnxfrm_onelevel)(cs,
                                    &cs->uca->level[current_level],
                                    dst, de, nweights,
                                    src, srclen, flags);
  }

  if (dst < de && (flags & MY_STRXFRM_PAD_TO_MAXLEN))
  {
    memset(dst, 0, de - dst);
    dst= de;
  }

  return dst - d0;
}


/*
  One-level, PAD SPACE
*/
MY_COLLATION_HANDLER MY_FUNCTION_NAME(collation_handler)=
{
  MY_UCA_COLL_INIT,
  MY_FUNCTION_NAME(strnncoll),
  MY_FUNCTION_NAME(strnncollsp),
  MY_FUNCTION_NAME(strnncollsp_nchars),
  MY_FUNCTION_NAME(strnxfrm),
  my_strnxfrmlen_any_uca,
  MY_LIKE_RANGE,
  my_wildcmp_uca,
  NULL,                                /* strcasecmp() */
  my_instr_mb,
  MY_FUNCTION_NAME(hash_sort),
  my_propagate_complex,
  my_min_str_mb_simple,
  my_max_str_mb_simple,
  my_ci_get_id_uca,
  my_ci_get_collation_name_uca
};


/*
  One-level, NO PAD
  For character sets with mbminlen==1 use MY_LIKE_RANGE=my_like_range_mb
  For character sets with mbminlen>=2 use MY_LIKE_RANGE=my_like_range_generic
*/
MY_COLLATION_HANDLER MY_FUNCTION_NAME(collation_handler_nopad)=
{
  MY_UCA_COLL_INIT,
  MY_FUNCTION_NAME(strnncoll),
  MY_FUNCTION_NAME(strnncollsp_nopad),
  MY_FUNCTION_NAME(strnncollsp_nchars),
  MY_FUNCTION_NAME(strnxfrm_nopad),
  my_strnxfrmlen_any_uca,
  MY_LIKE_RANGE,    /* my_like_range_mb or my_like_range_generic */
  my_wildcmp_uca,
  NULL,                                /* strcasecmp() */
  my_instr_mb,
  MY_FUNCTION_NAME(hash_sort_nopad),
  my_propagate_complex,
  my_min_str_mb_simple_nopad,
  my_max_str_mb_simple,
  my_ci_get_id_uca,
  my_ci_get_collation_name_uca
};


/*
  Multi-level, PAD SPACE
*/
MY_COLLATION_HANDLER MY_FUNCTION_NAME(collation_handler_multilevel)=
{
  MY_UCA_COLL_INIT,
  MY_FUNCTION_NAME(strnncoll_multilevel),
  MY_FUNCTION_NAME(strnncollsp_multilevel),
  MY_FUNCTION_NAME(strnncollsp_nchars_multilevel),
  MY_FUNCTION_NAME(strnxfrm_multilevel),
  my_strnxfrmlen_any_uca_multilevel,
  MY_LIKE_RANGE,
  my_wildcmp_uca,
  NULL,                                /* strcasecmp() */
  my_instr_mb,
  MY_FUNCTION_NAME(hash_sort),
  my_propagate_complex,
  my_min_str_mb_simple,
  my_max_str_mb_simple,
  my_ci_get_id_uca,
  my_ci_get_collation_name_uca
};


/*
  Multi-level, NO PAD
*/
MY_COLLATION_HANDLER MY_FUNCTION_NAME(collation_handler_nopad_multilevel)=
{
  MY_UCA_COLL_INIT,
  MY_FUNCTION_NAME(strnncoll_multilevel),
  MY_FUNCTION_NAME(strnncollsp_nopad_multilevel),
  MY_FUNCTION_NAME(strnncollsp_nchars_multilevel),
  MY_FUNCTION_NAME(strnxfrm_multilevel),
  my_strnxfrmlen_any_uca_multilevel,
  MY_LIKE_RANGE,
  my_wildcmp_uca,
  NULL,                                /* strcasecmp() */
  my_instr_mb,
  MY_FUNCTION_NAME(hash_sort),
  my_propagate_complex,
  my_min_str_mb_simple_nopad,
  my_max_str_mb_simple,
  my_ci_get_id_uca,
  my_ci_get_collation_name_uca
};


MY_COLLATION_HANDLER_PACKAGE MY_FUNCTION_NAME(package)=
{
  &MY_FUNCTION_NAME(collation_handler),
  &MY_FUNCTION_NAME(collation_handler_nopad),
  &MY_FUNCTION_NAME(collation_handler_multilevel),
  &MY_FUNCTION_NAME(collation_handler_nopad_multilevel)
};


#undef MY_FUNCTION_NAME
#undef MY_MB_WC
#undef MY_LIKE_RANGE
#undef MY_UCA_ASCII_OPTIMIZE
#undef MY_UCA_COMPILE_CONTRACTIONS
#undef MY_UCA_COLL_INIT