ctype-simple.c 37.3 KB
Newer Older
1
/* Copyright (c) 2002-2007 MySQL AB, 2009 Sun Microsystems, Inc.
2
   Copyright (c) 2009-2011, Monty Program Ab
3
   Use is subject to license terms.
4 5 6

   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
unknown's avatar
unknown committed
7
   the Free Software Foundation; version 2 of the License.
8 9 10 11 12 13 14 15

   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
16
   Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301  USA */
17

18 19
#include "strings_def.h"
#include <m_ctype.h>
20
#include "my_sys.h"  /* Needed for MY_ERRNO_ERANGE */
unknown's avatar
unknown committed
21 22
#include <errno.h>

23
#include "stdarg.h"
24

25 26 27
/*
  Returns the number of bytes required for strnxfrm().
*/
28 29

size_t my_strnxfrmlen_simple(CHARSET_INFO *cs, size_t len)
30 31 32 33 34
{
  return len * (cs->strxfrm_multiply ? cs->strxfrm_multiply : 1);
}


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
/*
  Converts a string into its sort key.
  
  SYNOPSIS
     my_strnxfrm_xxx()
     
  IMPLEMENTATION
     
     The my_strxfrm_xxx() function transforms a string pointed to by
     'src' with length 'srclen' according to the charset+collation 
     pair 'cs' and copies the result key into 'dest'.
     
     Comparing two strings using memcmp() after my_strnxfrm_xxx()
     is equal to comparing two original strings with my_strnncollsp_xxx().
     
     Not more than 'dstlen' bytes are written into 'dst'.
     To garantee that the whole string is transformed, 'dstlen' must be
     at least srclen*cs->strnxfrm_multiply bytes long. Otherwise,
     consequent memcmp() may return a non-accurate result.
     
     If the source string is too short to fill whole 'dstlen' bytes,
     then the 'dest' string is padded up to 'dstlen', ensuring that:
     
       "a"  == "a "
       "a\0" < "a"
       "a\0" < "a "
     
     my_strnxfrm_simple() is implemented for 8bit charsets and
     simple collations with one-to-one string->key transformation.
     
     See also implementations for various charsets/collations in  
     other ctype-xxx.c files.
     
  RETURN
  
    Target len 'dstlen'.
  
*/

unknown's avatar
unknown committed
74

75 76 77
size_t my_strnxfrm_simple(CHARSET_INFO * cs, 
                          uchar *dest, size_t len,
                          const uchar *src, size_t srclen)
78
{
79
  const uchar *map= cs->sort_order;
80
  size_t dstlen= len;
81
  set_if_smaller(len, srclen);
unknown's avatar
unknown committed
82 83 84 85 86 87 88 89 90 91 92 93
  if (dest != src)
  {
    const uchar *end;
    for ( end=src+len; src < end ;  )
      *dest++= map[*src++];
  }
  else
  {
    const uchar *end;
    for ( end=dest+len; dest < end ; dest++)
      *dest= (char) map[(uchar) *dest];
  }
94 95 96
  if (dstlen > len)
    bfill(dest, dstlen - len, ' ');
  return dstlen;
97 98
}

99 100 101

int my_strnncoll_simple(CHARSET_INFO * cs, const uchar *s, size_t slen, 
                        const uchar *t, size_t tlen,
102
                        my_bool t_is_prefix)
103
{
104
  size_t len = ( slen > tlen ) ? tlen : slen;
105
  const uchar *map= cs->sort_order;
106 107
  if (t_is_prefix && slen > tlen)
    slen=tlen;
108 109
  while (len--)
  {
110 111
    if (map[*s++] != map[*t++])
      return ((int) map[s[-1]] - (int) map[t[-1]]);
112
  }
113 114 115 116 117
  /*
    We can't use (slen - tlen) here as the result may be outside of the
    precision of a signed int
  */
  return slen > tlen ? 1 : slen < tlen ? -1 : 0 ;
118 119
}

120

121 122 123 124 125 126 127 128 129 130
/*
  Compare strings, discarding end space

  SYNOPSIS
    my_strnncollsp_simple()
    cs			character set handler
    a			First string to compare
    a_length		Length of 'a'
    b			Second string to compare
    b_length		Length of 'b'
131 132 133
    diff_if_only_endspace_difference
		        Set to 1 if the strings should be regarded as different
                        if they only difference in end space
134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150

  IMPLEMENTATION
    If one string is shorter as the other, then we space extend the other
    so that the strings have equal length.

    This will ensure that the following things hold:

    "a"  == "a "
    "a\0" < "a"
    "a\0" < "a "

  RETURN
    < 0	 a <  b
    = 0	 a == b
    > 0	 a > b
*/

151 152
int my_strnncollsp_simple(CHARSET_INFO * cs, const uchar *a, size_t a_length, 
			  const uchar *b, size_t b_length,
153
                          my_bool diff_if_only_endspace_difference)
154
{
155
  const uchar *map= cs->sort_order, *end;
156
  size_t length;
157 158 159 160 161
  int res;

#ifndef VARCHAR_WITH_DIFF_ENDSPACE_ARE_DIFFERENT_FOR_UNIQUE
  diff_if_only_endspace_difference= 0;
#endif
162 163 164

  end= a + (length= min(a_length, b_length));
  while (a < end)
165
  {
166 167
    if (map[*a++] != map[*b++])
      return ((int) map[a[-1]] - (int) map[b[-1]]);
168
  }
169
  res= 0;
170 171
  if (a_length != b_length)
  {
172
    int swap= 1;
173 174
    if (diff_if_only_endspace_difference)
      res= 1;                                   /* Assume 'a' is bigger */
175 176 177 178 179 180 181 182 183
    /*
      Check the next not space character of the longer key. If it's < ' ',
      then it's smaller than the other key.
    */
    if (a_length < b_length)
    {
      /* put shorter key in s */
      a_length= b_length;
      a= b;
184
      swap= -1;                                 /* swap sign of result */
185
      res= -res;
186 187 188
    }
    for (end= a + a_length-length; a < end ; a++)
    {
189 190
      if (map[*a] != map[' '])
	return (map[*a] < map[' ']) ? -swap : swap;
191 192
    }
  }
193
  return res;
194 195
}

196

197
size_t my_caseup_str_8bit(CHARSET_INFO * cs,char *str)
198
{
199
  register const uchar *map= cs->to_upper;
200 201
  char *str_orig= str;
  while ((*str= (char) map[(uchar) *str]) != 0)
202
    str++;
203
  return (size_t) (str - str_orig);
204 205
}

206

207
size_t my_casedn_str_8bit(CHARSET_INFO * cs,char *str)
208
{
209
  register const uchar *map= cs->to_lower;
210 211
  char *str_orig= str;
  while ((*str= (char) map[(uchar) *str]) != 0)
212
    str++;
213
  return (size_t) (str - str_orig);
214 215
}

216

217 218 219
size_t my_caseup_8bit(CHARSET_INFO * cs, char *src, size_t srclen,
                      char *dst __attribute__((unused)),
                      size_t dstlen __attribute__((unused)))
220
{
221
  char *end= src + srclen;
222
  register const uchar *map= cs->to_upper;
223
  DBUG_ASSERT(src == dst && srclen == dstlen);
224
  for ( ; src != end ; src++)
225
    *src= (char) map[(uchar) *src];
226
  return srclen;
227 228
}

229 230 231 232

size_t my_casedn_8bit(CHARSET_INFO * cs, char *src, size_t srclen,
                      char *dst __attribute__((unused)),
                      size_t dstlen __attribute__((unused)))
233
{
234
  char *end= src + srclen;
235
  register const uchar *map=cs->to_lower;
236
  DBUG_ASSERT(src == dst && srclen == dstlen);
237
  for ( ; src != end ; src++)
238
    *src= (char) map[(uchar) *src];
239
  return srclen;
240 241 242 243
}

int my_strcasecmp_8bit(CHARSET_INFO * cs,const char *s, const char *t)
{
244
  register const uchar *map=cs->to_upper;
unknown's avatar
unknown committed
245
  while (map[(uchar) *s] == map[(uchar) *t++])
246
    if (!*s++) return 0;
unknown's avatar
unknown committed
247
  return ((int) map[(uchar) s[0]] - (int) map[(uchar) t[-1]]);
248 249 250
}


251
int my_mb_wc_8bit(CHARSET_INFO *cs,my_wc_t *wc,
252 253
		  const uchar *str,
		  const uchar *end __attribute__((unused)))
254
{
255
  if (str >= end)
256
    return MY_CS_TOOSMALL;
257
  
258
  *wc=cs->tab_to_uni[*str];
259
  return (!wc[0] && str[0]) ? -1 : 1;
260 261 262
}

int my_wc_mb_8bit(CHARSET_INFO *cs,my_wc_t wc,
263 264
		  uchar *str,
		  uchar *end)
265 266 267
{
  MY_UNI_IDX *idx;

268 269 270
  if (str >= end)
    return MY_CS_TOOSMALL;
  
271 272 273 274 275 276
  for (idx=cs->tab_from_uni; idx->tab ; idx++)
  {
    if (idx->from <= wc && idx->to >= wc)
    {
      str[0]= idx->tab[wc - idx->from];
      return (!str[0] && wc) ? MY_CS_ILUNI : 1;
277 278 279 280
    }
  }
  return MY_CS_ILUNI;
}
281 282


283 284 285 286 287 288
/* 
   We can't use vsprintf here as it's not guaranteed to return
   the length on all operating systems.
   This function is also not called in a safe environment, so the
   end buffer must be checked.
*/
289

290 291
size_t my_snprintf_8bit(CHARSET_INFO *cs  __attribute__((unused)),
                        char* to, size_t n  __attribute__((unused)),
292
		     const char* fmt, ...)
293 294
{
  va_list args;
295
  int result;
296
  va_start(args,fmt);
297 298 299
  result= my_vsnprintf(to, n, fmt, args);
  va_end(args);
  return result;
300 301 302
}


303
void my_hash_sort_simple(CHARSET_INFO *cs,
304
			 const uchar *key, size_t len,
305
			 ulong *nr1, ulong *nr2)
306
{
307
  register const uchar *sort_order=cs->sort_order;
308
  const uchar *end;
309
  ulong n1, n2;
Sergei Golubchik's avatar
Sergei Golubchik committed
310

311 312 313 314
  /*
    Remove end space. We have to do this to be able to compare
    'A ' and 'A' as identical
  */
315
  end= skip_trailing_space(key, len);
316 317 318

  n1= *nr1;
  n2= *nr2;
319
  for (; key < (uchar*) end ; key++)
320
  {
321 322 323
    n1^=(ulong) ((((uint) n1 & 63)+n2) *
	     ((uint) sort_order[(uint) *key])) + (n1 << 8);
    n2+=3;
324
  }
325 326
  *nr1= n1;
  *nr2= n2;
327
}
328

unknown's avatar
unknown committed
329

unknown's avatar
unknown committed
330
long my_strntol_8bit(CHARSET_INFO *cs,
331
		     const char *nptr, size_t l, int base,
unknown's avatar
unknown committed
332
		     char **endptr, int *err)
333
{
unknown's avatar
unknown committed
334
  int negative;
unknown's avatar
unknown committed
335
  register uint32 cutoff;
336
  register uint cutlim;
unknown's avatar
unknown committed
337
  register uint32 i;
unknown's avatar
unknown committed
338
  register const char *s;
339
  register uchar c;
unknown's avatar
unknown committed
340 341
  const char *save, *e;
  int overflow;
342

343
  *err= 0;				/* Initialize error indicator */
344

unknown's avatar
unknown committed
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
  s = nptr;
  e = nptr+l;
  
  for ( ; s<e && my_isspace(cs, *s) ; s++);
  
  if (s == e)
  {
    goto noconv;
  }
  
  /* Check for a sign.	*/
  if (*s == '-')
  {
    negative = 1;
    ++s;
  }
  else if (*s == '+')
  {
    negative = 0;
    ++s;
  }
  else
    negative = 0;

  save = s;
unknown's avatar
unknown committed
370 371
  cutoff = ((uint32)~0L) / (uint32) base;
  cutlim = (uint) (((uint32)~0L) % (uint32) base);
unknown's avatar
unknown committed
372 373 374 375 376 377 378

  overflow = 0;
  i = 0;
  for (c = *s; s != e; c = *++s)
  {
    if (c>='0' && c<='9')
      c -= '0';
379
    else if (c>='A' && c<='Z')
unknown's avatar
unknown committed
380
      c = c - 'A' + 10;
381
    else if (c>='a' && c<='z')
unknown's avatar
unknown committed
382 383 384 385 386 387 388 389 390
      c = c - 'a' + 10;
    else
      break;
    if (c >= base)
      break;
    if (i > cutoff || (i == cutoff && c > cutlim))
      overflow = 1;
    else
    {
unknown's avatar
unknown committed
391
      i *= (uint32) base;
unknown's avatar
unknown committed
392 393 394 395 396 397 398 399 400 401 402 403
      i += c;
    }
  }
  
  if (s == save)
    goto noconv;
  
  if (endptr != NULL)
    *endptr = (char *) s;
  
  if (negative)
  {
unknown's avatar
unknown committed
404
    if (i  > (uint32) INT_MIN32)
unknown's avatar
unknown committed
405 406
      overflow = 1;
  }
unknown's avatar
unknown committed
407
  else if (i > INT_MAX32)
unknown's avatar
unknown committed
408 409 410 411
    overflow = 1;
  
  if (overflow)
  {
412
    err[0]= ERANGE;
unknown's avatar
unknown committed
413
    return negative ? INT_MIN32 : INT_MAX32;
unknown's avatar
unknown committed
414 415 416 417 418
  }
  
  return (negative ? -((long) i) : (long) i);

noconv:
419
  err[0]= EDOM;
unknown's avatar
unknown committed
420 421 422
  if (endptr != NULL)
    *endptr = (char *) nptr;
  return 0L;
423 424
}

unknown's avatar
unknown committed
425

unknown's avatar
unknown committed
426
ulong my_strntoul_8bit(CHARSET_INFO *cs,
427
		       const char *nptr, size_t l, int base,
unknown's avatar
unknown committed
428
		       char **endptr, int *err)
429
{
unknown's avatar
unknown committed
430
  int negative;
unknown's avatar
unknown committed
431
  register uint32 cutoff;
432
  register uint cutlim;
unknown's avatar
unknown committed
433
  register uint32 i;
unknown's avatar
unknown committed
434
  register const char *s;
435
  register uchar c;
unknown's avatar
unknown committed
436 437 438
  const char *save, *e;
  int overflow;

439
  *err= 0;				/* Initialize error indicator */
440

unknown's avatar
unknown committed
441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464
  s = nptr;
  e = nptr+l;
  
  for( ; s<e && my_isspace(cs, *s); s++);
  
  if (s==e)
  {
    goto noconv;
  }

  if (*s == '-')
  {
    negative = 1;
    ++s;
  }
  else if (*s == '+')
  {
    negative = 0;
    ++s;
  }
  else
    negative = 0;

  save = s;
unknown's avatar
unknown committed
465 466
  cutoff = ((uint32)~0L) / (uint32) base;
  cutlim = (uint) (((uint32)~0L) % (uint32) base);
unknown's avatar
unknown committed
467 468 469 470 471 472 473
  overflow = 0;
  i = 0;
  
  for (c = *s; s != e; c = *++s)
  {
    if (c>='0' && c<='9')
      c -= '0';
474
    else if (c>='A' && c<='Z')
unknown's avatar
unknown committed
475
      c = c - 'A' + 10;
476
    else if (c>='a' && c<='z')
unknown's avatar
unknown committed
477 478 479 480 481 482 483 484 485
      c = c - 'a' + 10;
    else
      break;
    if (c >= base)
      break;
    if (i > cutoff || (i == cutoff && c > cutlim))
      overflow = 1;
    else
    {
unknown's avatar
unknown committed
486
      i *= (uint32) base;
unknown's avatar
unknown committed
487 488 489 490 491 492 493 494 495 496 497 498
      i += c;
    }
  }

  if (s == save)
    goto noconv;

  if (endptr != NULL)
    *endptr = (char *) s;

  if (overflow)
  {
499
    err[0]= ERANGE;
unknown's avatar
unknown committed
500
    return (~(uint32) 0);
unknown's avatar
unknown committed
501 502 503 504 505
  }
  
  return (negative ? -((long) i) : (long) i);
  
noconv:
506
  err[0]= EDOM;
unknown's avatar
unknown committed
507 508 509
  if (endptr != NULL)
    *endptr = (char *) nptr;
  return 0L;
510 511
}

unknown's avatar
unknown committed
512

unknown's avatar
unknown committed
513
longlong my_strntoll_8bit(CHARSET_INFO *cs __attribute__((unused)),
514
			  const char *nptr, size_t l, int base,
unknown's avatar
unknown committed
515
			  char **endptr,int *err)
516
{
unknown's avatar
unknown committed
517 518
  int negative;
  register ulonglong cutoff;
519
  register uint cutlim;
unknown's avatar
unknown committed
520 521 522 523 524
  register ulonglong i;
  register const char *s, *e;
  const char *save;
  int overflow;

525
  *err= 0;				/* Initialize error indicator */
unknown's avatar
unknown committed
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

  s = nptr;
  e = nptr+l;

  for(; s<e && my_isspace(cs,*s); s++);

  if (s == e)
  {
    goto noconv;
  }

  if (*s == '-')
  {
    negative = 1;
    ++s;
  }
  else if (*s == '+')
  {
    negative = 0;
    ++s;
  }
  else
    negative = 0;

  save = s;

  cutoff = (~(ulonglong) 0) / (unsigned long int) base;
  cutlim = (uint) ((~(ulonglong) 0) % (unsigned long int) base);

  overflow = 0;
  i = 0;
557
  for ( ; s != e; s++)
unknown's avatar
unknown committed
558
  {
559
    register uchar c= *s;
unknown's avatar
unknown committed
560 561
    if (c>='0' && c<='9')
      c -= '0';
562
    else if (c>='A' && c<='Z')
unknown's avatar
unknown committed
563
      c = c - 'A' + 10;
564
    else if (c>='a' && c<='z')
unknown's avatar
unknown committed
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
      c = c - 'a' + 10;
    else
      break;
    if (c >= base)
      break;
    if (i > cutoff || (i == cutoff && c > cutlim))
      overflow = 1;
    else
    {
      i *= (ulonglong) base;
      i += c;
    }
  }

  if (s == save)
    goto noconv;

  if (endptr != NULL)
    *endptr = (char *) s;

  if (negative)
  {
    if (i  > (ulonglong) LONGLONG_MIN)
      overflow = 1;
  }
  else if (i > (ulonglong) LONGLONG_MAX)
    overflow = 1;

  if (overflow)
  {
595
    err[0]= ERANGE;
unknown's avatar
unknown committed
596 597 598 599 600 601
    return negative ? LONGLONG_MIN : LONGLONG_MAX;
  }

  return (negative ? -((longlong) i) : (longlong) i);

noconv:
602
  err[0]= EDOM;
unknown's avatar
unknown committed
603 604 605
  if (endptr != NULL)
    *endptr = (char *) nptr;
  return 0L;
606 607
}

unknown's avatar
unknown committed
608 609

ulonglong my_strntoull_8bit(CHARSET_INFO *cs,
610
			   const char *nptr, size_t l, int base,
611
			   char **endptr, int *err)
612
{
unknown's avatar
unknown committed
613 614
  int negative;
  register ulonglong cutoff;
615
  register uint cutlim;
unknown's avatar
unknown committed
616 617 618 619 620
  register ulonglong i;
  register const char *s, *e;
  const char *save;
  int overflow;

621
  *err= 0;				/* Initialize error indicator */
unknown's avatar
unknown committed
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

  s = nptr;
  e = nptr+l;

  for(; s<e && my_isspace(cs,*s); s++);

  if (s == e)
  {
    goto noconv;
  }

  if (*s == '-')
  {
    negative = 1;
    ++s;
  }
  else if (*s == '+')
  {
    negative = 0;
    ++s;
  }
  else
    negative = 0;

  save = s;

  cutoff = (~(ulonglong) 0) / (unsigned long int) base;
  cutlim = (uint) ((~(ulonglong) 0) % (unsigned long int) base);

  overflow = 0;
  i = 0;
653
  for ( ; s != e; s++)
unknown's avatar
unknown committed
654
  {
655
    register uchar c= *s;
656

unknown's avatar
unknown committed
657 658
    if (c>='0' && c<='9')
      c -= '0';
659
    else if (c>='A' && c<='Z')
unknown's avatar
unknown committed
660
      c = c - 'A' + 10;
661
    else if (c>='a' && c<='z')
unknown's avatar
unknown committed
662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683
      c = c - 'a' + 10;
    else
      break;
    if (c >= base)
      break;
    if (i > cutoff || (i == cutoff && c > cutlim))
      overflow = 1;
    else
    {
      i *= (ulonglong) base;
      i += c;
    }
  }

  if (s == save)
    goto noconv;

  if (endptr != NULL)
    *endptr = (char *) s;

  if (overflow)
  {
684
    err[0]= ERANGE;
unknown's avatar
unknown committed
685 686 687 688 689 690
    return (~(ulonglong) 0);
  }

  return (negative ? -((longlong) i) : (longlong) i);

noconv:
691
  err[0]= EDOM;
unknown's avatar
unknown committed
692 693 694
  if (endptr != NULL)
    *endptr = (char *) nptr;
  return 0L;
695 696
}

697

698 699 700 701 702 703 704 705
/*
  Read double from string

  SYNOPSIS:
    my_strntod_8bit()
    cs		Character set information
    str		String to convert to double
    length	Optional length for string.
706 707
    end		result pointer to end of converted string
    err		Error number if failed conversion
708 709 710 711 712 713 714 715 716
    
  NOTES:
    If length is not INT_MAX32 or str[length] != 0 then the given str must
    be writeable
    If length == INT_MAX32 the str must be \0 terminated.

    It's implemented this way to save a buffer allocation and a memory copy.

  RETURN
717
    Value of number in string
718 719 720 721
*/


double my_strntod_8bit(CHARSET_INFO *cs __attribute__((unused)),
722
		       char *str, size_t length,
723
		       char **end, int *err)
724
{
725
  if (length == INT_MAX32)
726 727 728
    length= 65535;                          /* Should be big enough */
  *end= str + length;
  return my_strtod(str, end, err);
729
}
730 731


732 733
/*
  This is a fast version optimized for the case of radix 10 / -10
734 735

  Assume len >= 1
736 737
*/

738 739
size_t my_long10_to_str_8bit(CHARSET_INFO *cs __attribute__((unused)),
                             char *dst, size_t len, int radix, long int val)
740
{
unknown's avatar
unknown committed
741 742 743
  char buffer[66];
  register char *p, *e;
  long int new_val;
744
  uint sign=0;
745
  unsigned long int uval = (unsigned long int) val;
746

unknown's avatar
unknown committed
747
  e = p = &buffer[sizeof(buffer)-1];
748
  *p= 0;
unknown's avatar
unknown committed
749 750 751 752 753
  
  if (radix < 0)
  {
    if (val < 0)
    {
754 755
      /* Avoid integer overflow in (-val) for LONGLONG_MIN (BUG#31799). */
      uval= (unsigned long int)0 - uval;
756 757 758
      *dst++= '-';
      len--;
      sign= 1;
unknown's avatar
unknown committed
759 760 761
    }
  }
  
762 763
  new_val = (long) (uval / 10);
  *--p    = '0'+ (char) (uval - (unsigned long) new_val * 10);
unknown's avatar
unknown committed
764 765 766 767 768 769 770 771 772
  val     = new_val;
  
  while (val != 0)
  {
    new_val=val/10;
    *--p = '0' + (char) (val-new_val*10);
    val= new_val;
  }
  
773
  len= min(len, (size_t) (e-p));
774
  memcpy(dst, p, len);
775
  return len+sign;
unknown's avatar
unknown committed
776
}
777

778

779 780 781
size_t my_longlong10_to_str_8bit(CHARSET_INFO *cs __attribute__((unused)),
                                 char *dst, size_t len, int radix,
                                 longlong val)
782
{
unknown's avatar
unknown committed
783 784 785
  char buffer[65];
  register char *p, *e;
  long long_val;
786
  uint sign= 0;
787
  ulonglong uval = (ulonglong)val;
unknown's avatar
unknown committed
788 789 790 791 792
  
  if (radix < 0)
  {
    if (val < 0)
    {
793 794
      /* Avoid integer overflow in (-val) for LONGLONG_MIN (BUG#31799). */
      uval = (ulonglong)0 - uval;
795 796 797
      *dst++= '-';
      len--;
      sign= 1;
unknown's avatar
unknown committed
798 799 800 801
    }
  }
  
  e = p = &buffer[sizeof(buffer)-1];
802
  *p= 0;
unknown's avatar
unknown committed
803
  
804
  if (uval == 0)
unknown's avatar
unknown committed
805
  {
806 807
    *--p= '0';
    len= 1;
unknown's avatar
unknown committed
808 809 810
    goto cnv;
  }
  
811
  while (uval > (ulonglong) LONG_MAX)
unknown's avatar
unknown committed
812
  {
813 814
    ulonglong quo= uval/(uint) 10;
    uint rem= (uint) (uval- quo* (uint) 10);
unknown's avatar
unknown committed
815
    *--p = '0' + rem;
816
    uval= quo;
unknown's avatar
unknown committed
817 818
  }
  
819
  long_val= (long) uval;
unknown's avatar
unknown committed
820 821 822
  while (long_val != 0)
  {
    long quo= long_val/10;
823
    *--p = (char) ('0' + (long_val - quo*10));
unknown's avatar
unknown committed
824 825 826
    long_val= quo;
  }
  
827
  len= min(len, (size_t) (e-p));
unknown's avatar
unknown committed
828
cnv:
829 830
  memcpy(dst, p, len);
  return len+sign;
831 832 833
}


834 835 836 837 838 839 840 841 842 843 844 845 846
/*
** Compare string against string with wildcard
**	0 if matched
**	-1 if not matched with wildcard
**	 1 if matched with wildcard
*/

#ifdef LIKE_CMP_TOUPPER
#define likeconv(s,A) (uchar) my_toupper(s,A)
#else
#define likeconv(s,A) (uchar) (s)->sort_order[(uchar) (A)]
#endif

unknown's avatar
unknown committed
847
#define INC_PTR(cs,A,B) (A)++
848 849 850 851 852 853 854


int my_wildcmp_8bit(CHARSET_INFO *cs,
		    const char *str,const char *str_end,
		    const char *wildstr,const char *wildend,
		    int escape, int w_one, int w_many)
{
unknown's avatar
unknown committed
855
  int result= -1;			/* Not found, using wildcards */
856 857 858 859 860 861 862 863 864

  while (wildstr != wildend)
  {
    while (*wildstr != w_many && *wildstr != w_one)
    {
      if (*wildstr == escape && wildstr+1 != wildend)
	wildstr++;

      if (str == str_end || likeconv(cs,*wildstr++) != likeconv(cs,*str++))
unknown's avatar
unknown committed
865
	return(1);				/* No match */
866
      if (wildstr == wildend)
unknown's avatar
unknown committed
867
	return(str != str_end);		/* Match if both are at end */
unknown's avatar
unknown committed
868
      result=1;					/* Found an anchor char     */
869 870 871 872 873
    }
    if (*wildstr == w_one)
    {
      do
      {
unknown's avatar
unknown committed
874
	if (str == str_end)			/* Skip one char if possible */
unknown's avatar
unknown committed
875
	  return(result);
876 877 878 879 880 881
	INC_PTR(cs,str,str_end);
      } while (++wildstr < wildend && *wildstr == w_one);
      if (wildstr == wildend)
	break;
    }
    if (*wildstr == w_many)
unknown's avatar
unknown committed
882
    {						/* Found w_many */
883 884 885 886 887 888 889 890 891 892 893
      uchar cmp;
      
      wildstr++;
      /* Remove any '%' and '_' from the wild search string */
      for (; wildstr != wildend ; wildstr++)
      {
	if (*wildstr == w_many)
	  continue;
	if (*wildstr == w_one)
	{
	  if (str == str_end)
unknown's avatar
unknown committed
894
	    return(-1);
895 896 897
	  INC_PTR(cs,str,str_end);
	  continue;
	}
unknown's avatar
unknown committed
898
	break;					/* Not a wild character */
899 900
      }
      if (wildstr == wildend)
unknown's avatar
unknown committed
901
	return(0);				/* Ok if w_many is last */
902
      if (str == str_end)
unknown's avatar
unknown committed
903
	return(-1);
904 905 906 907
      
      if ((cmp= *wildstr) == escape && wildstr+1 != wildend)
	cmp= *++wildstr;

unknown's avatar
unknown committed
908 909
      INC_PTR(cs,wildstr,wildend);	/* This is compared trough cmp */
      cmp=likeconv(cs,cmp);
910 911
      do
      {
unknown's avatar
unknown committed
912 913 914
	while (str != str_end && (uchar) likeconv(cs,*str) != cmp)
	  str++;
	if (str++ == str_end) return(-1);
915
	{
unknown's avatar
unknown committed
916 917
	  int tmp=my_wildcmp_8bit(cs,str,str_end,wildstr,wildend,escape,w_one,
				  w_many);
918
	  if (tmp <= 0)
unknown's avatar
unknown committed
919
	    return(tmp);
920 921 922 923 924
	}
      } while (str != str_end && wildstr[0] != w_many);
      return(-1);
    }
  }
unknown's avatar
unknown committed
925
  return(str != str_end ? 1 : 0);
926
}
927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946


/*
** Calculate min_str and max_str that ranges a LIKE string.
** Arguments:
** ptr		Pointer to LIKE string.
** ptr_length	Length of LIKE string.
** escape	Escape character in LIKE.  (Normally '\').
**		All escape characters should be removed from min_str and max_str
** res_length	Length of min_str and max_str.
** min_str	Smallest case sensitive string that ranges LIKE.
**		Should be space padded to res_length.
** max_str	Largest case sensitive string that ranges LIKE.
**		Normally padded with the biggest character sort value.
**
** The function should return 0 if ok and 1 if the LIKE string can't be
** optimized !
*/

my_bool my_like_range_simple(CHARSET_INFO *cs,
947
			     const char *ptr, size_t ptr_length,
unknown's avatar
unknown committed
948
			     pbool escape, pbool w_one, pbool w_many,
949
			     size_t res_length,
unknown's avatar
unknown committed
950
			     char *min_str,char *max_str,
951
			     size_t *min_length, size_t *max_length)
952
{
953
  const char *end= ptr + ptr_length;
954 955
  char *min_org=min_str;
  char *min_end=min_str+res_length;
956
  size_t charlen= res_length / cs->mbmaxlen;
957

958
  for (; ptr != end && min_str != min_end && charlen > 0 ; ptr++, charlen--)
959 960 961
  {
    if (*ptr == escape && ptr+1 != end)
    {
unknown's avatar
unknown committed
962
      ptr++;					/* Skip escape */
963 964 965
      *min_str++= *max_str++ = *ptr;
      continue;
    }
unknown's avatar
unknown committed
966
    if (*ptr == w_one)				/* '_' in SQL */
967
    {
unknown's avatar
unknown committed
968
      *min_str++='\0';				/* This should be min char */
969
      *max_str++= (char) cs->max_sort_char;
970 971
      continue;
    }
unknown's avatar
unknown committed
972
    if (*ptr == w_many)				/* '%' in SQL */
973
    {
974
      /* Calculate length of keys */
975 976
      *min_length= ((cs->state & MY_CS_BINSORT) ?
                    (size_t) (min_str - min_org) :
977 978
                    res_length);
      *max_length= res_length;
979 980 981 982
      do
      {
	*min_str++= 0;
	*max_str++= (char) cs->max_sort_char;
983 984 985 986 987 988
      } while (min_str != min_end);
      return 0;
    }
    *min_str++= *max_str++ = *ptr;
  }

989
 *min_length= *max_length = (size_t) (min_str - min_org);
990
  while (min_str != min_end)
991
    *min_str++= *max_str++ = ' ';      /* Because if key compression */
992 993
  return 0;
}
994 995


996
size_t my_scan_8bit(CHARSET_INFO *cs, const char *str, const char *end, int sq)
997 998 999 1000 1001 1002 1003 1004
{
  const char *str0= str;
  switch (sq)
  {
  case MY_SEQ_INTTAIL:
    if (*str == '.')
    {
      for(str++ ; str != end && *str == '0' ; str++);
1005
      return (size_t) (str - str0);
1006 1007 1008 1009
    }
    return 0;

  case MY_SEQ_SPACES:
1010
    for ( ; str < end ; str++)
1011 1012 1013 1014
    {
      if (!my_isspace(cs,*str))
        break;
    }
1015
    return (size_t) (str - str0);
1016 1017 1018 1019
  default:
    return 0;
  }
}
1020

1021

1022
void my_fill_8bit(CHARSET_INFO *cs __attribute__((unused)),
1023
		   char *s, size_t l, int fill)
1024
{
1025
  bfill((uchar*) s,l,fill);
1026 1027
}

1028

1029
size_t my_numchars_8bit(CHARSET_INFO *cs __attribute__((unused)),
1030 1031
		      const char *b, const char *e)
{
1032
  return (size_t) (e - b);
1033 1034
}

1035

1036 1037
size_t my_numcells_8bit(CHARSET_INFO *cs __attribute__((unused)),
                        const char *b, const char *e)
1038
{
1039
  return (size_t) (e - b);
1040 1041 1042
}


1043 1044 1045 1046
size_t my_charpos_8bit(CHARSET_INFO *cs __attribute__((unused)),
                       const char *b  __attribute__((unused)),
                       const char *e  __attribute__((unused)),
                       size_t pos)
1047 1048 1049
{
  return pos;
}
1050

1051

1052 1053 1054
size_t my_well_formed_len_8bit(CHARSET_INFO *cs __attribute__((unused)),
                               const char *start, const char *end,
                               size_t nchars, int *error)
unknown's avatar
unknown committed
1055
{
1056
  size_t nbytes= (size_t) (end-start);
1057
  *error= 0;
1058
  return min(nbytes, nchars);
unknown's avatar
unknown committed
1059 1060
}

1061

1062 1063
size_t my_lengthsp_8bit(CHARSET_INFO *cs __attribute__((unused)),
                        const char *ptr, size_t length)
unknown's avatar
unknown committed
1064
{
1065 1066
  const char *end;
  end= (const char *) skip_trailing_space((const uchar *)ptr, length);
1067
  return (size_t) (end-ptr);
unknown's avatar
unknown committed
1068 1069 1070
}


1071
uint my_instr_simple(CHARSET_INFO *cs,
1072 1073 1074
                     const char *b, size_t b_length, 
                     const char *s, size_t s_length,
                     my_match_t *match, uint nmatch)
1075 1076 1077 1078 1079 1080
{
  register const uchar *str, *search, *end, *search_end;
  
  if (s_length <= b_length)
  {
    if (!s_length)
1081 1082 1083 1084 1085
    {
      if (nmatch)
      {
        match->beg= 0;
        match->end= 0;
1086
        match->mb_len= 0;
1087 1088 1089
      }
      return 1;		/* Empty string is always found */
    }
1090
    
unknown's avatar
unknown committed
1091 1092 1093 1094
    str= (const uchar*) b;
    search= (const uchar*) s;
    end= (const uchar*) b+b_length-s_length+1;
    search_end= (const uchar*) s + s_length;
1095
    
1096
skip:
1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
    while (str != end)
    {
      if (cs->sort_order[*str++] == cs->sort_order[*search])
      {
	register const uchar *i,*j;
	
	i= str; 
	j= search+1;
	
	while (j != search_end)
	  if (cs->sort_order[*i++] != cs->sort_order[*j++]) 
1108
            goto skip;
1109
        
1110 1111 1112
	if (nmatch > 0)
	{
	  match[0].beg= 0;
1113
	  match[0].end= (size_t) (str- (const uchar*)b-1);
1114
	  match[0].mb_len= match[0].end;
1115 1116 1117 1118 1119
	  
	  if (nmatch > 1)
	  {
	    match[1].beg= match[0].end;
	    match[1].end= match[0].end+s_length;
1120
	    match[1].mb_len= match[1].end-match[1].beg;
1121 1122 1123
	  }
	}
	return 2;
1124 1125 1126
      }
    }
  }
1127
  return 0;
1128 1129 1130
}


1131 1132 1133
typedef struct
{
  int		nchars;
1134
  struct my_uni_idx_st uidx;
1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151
} uni_idx;

#define PLANE_SIZE	0x100
#define PLANE_NUM	0x100
#define PLANE_NUMBER(x)	(((x)>>8) % PLANE_NUM)

static int pcmp(const void * f, const void * s)
{
  const uni_idx *F= (const uni_idx*) f;
  const uni_idx *S= (const uni_idx*) s;
  int res;

  if (!(res=((S->nchars)-(F->nchars))))
    res=((F->uidx.from)-(S->uidx.to));
  return res;
}

1152 1153
static my_bool create_fromuni(struct charset_info_st *cs,
                              void *(*alloc)(size_t))
1154 1155 1156
{
  uni_idx	idx[PLANE_NUM];
  int		i,n;
1157
  struct my_uni_idx_st *tab_from_uni;
1158
  
1159 1160 1161 1162 1163 1164 1165 1166 1167
  /*
    Check that Unicode map is loaded.
    It can be not loaded when the collation is
    listed in Index.xml but not specified
    in the character set specific XML file.
  */
  if (!cs->tab_to_uni)
    return TRUE;
  
1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197
  /* Clear plane statistics */
  bzero(idx,sizeof(idx));
  
  /* Count number of characters in each plane */
  for (i=0; i< 0x100; i++)
  {
    uint16 wc=cs->tab_to_uni[i];
    int pl= PLANE_NUMBER(wc);
    
    if (wc || !i)
    {
      if (!idx[pl].nchars)
      {
        idx[pl].uidx.from=wc;
        idx[pl].uidx.to=wc;
      }else
      {
        idx[pl].uidx.from=wc<idx[pl].uidx.from?wc:idx[pl].uidx.from;
        idx[pl].uidx.to=wc>idx[pl].uidx.to?wc:idx[pl].uidx.to;
      }
      idx[pl].nchars++;
    }
  }
  
  /* Sort planes in descending order */
  qsort(&idx,PLANE_NUM,sizeof(uni_idx),&pcmp);
  
  for (i=0; i < PLANE_NUM; i++)
  {
    int ch,numchars;
1198
    uchar *tab;
1199 1200 1201 1202 1203 1204
    
    /* Skip empty plane */
    if (!idx[i].nchars)
      break;
    
    numchars=idx[i].uidx.to-idx[i].uidx.from+1;
1205 1206
    if (!(idx[i].uidx.tab= tab= (uchar*)
          alloc(numchars * sizeof(*idx[i].uidx.tab))))
1207 1208
      return TRUE;
    
1209
    bzero(tab,numchars*sizeof(*tab));
1210 1211 1212 1213 1214 1215 1216
    
    for (ch=1; ch < PLANE_SIZE; ch++)
    {
      uint16 wc=cs->tab_to_uni[ch];
      if (wc >= idx[i].uidx.from && wc <= idx[i].uidx.to && wc)
      {
        int ofs= wc - idx[i].uidx.from;
1217
        tab[ofs]= ch;
1218 1219 1220 1221 1222 1223
      }
    }
  }
  
  /* Allocate and fill reverse table for each plane */
  n=i;
1224 1225
  if (!(cs->tab_from_uni= tab_from_uni= (struct my_uni_idx_st*)
        alloc(sizeof(MY_UNI_IDX)*(n+1))))
1226 1227 1228
    return TRUE;

  for (i=0; i< n; i++)
1229
    tab_from_uni[i]= idx[i].uidx;
1230 1231
  
  /* Set end-of-list marker */
1232
  bzero(&tab_from_uni[i],sizeof(MY_UNI_IDX));
1233 1234 1235
  return FALSE;
}

1236 1237
static my_bool my_cset_init_8bit(struct charset_info_st *cs,
                                 void *(*alloc)(size_t))
1238
{
1239 1240
  cs->caseup_multiply= 1;
  cs->casedn_multiply= 1;
1241
  cs->pad_char= ' ';
1242 1243 1244
  return create_fromuni(cs, alloc);
}

1245
static void set_max_sort_char(struct charset_info_st *cs)
1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263
{
  uchar max_char;
  uint  i;
  
  if (!cs->sort_order)
    return;
  
  max_char=cs->sort_order[(uchar) cs->max_sort_char];
  for (i= 0; i < 256; i++)
  {
    if ((uchar) cs->sort_order[i] > max_char)
    {
      max_char=(uchar) cs->sort_order[i];
      cs->max_sort_char= i;
    }
  }
}

1264
static my_bool my_coll_init_simple(struct charset_info_st *cs,
1265
                                   void *(*alloc)(size_t) __attribute__((unused)))
1266 1267 1268 1269 1270
{
  set_max_sort_char(cs);
  return FALSE;
}

1271

1272 1273 1274
longlong my_strtoll10_8bit(CHARSET_INFO *cs __attribute__((unused)),
                           const char *nptr, char **endptr, int *error)
{
1275
  return my_strtoll10(nptr, endptr, error);
1276 1277
}

1278

1279
int my_mb_ctype_8bit(CHARSET_INFO *cs, int *ctype,
1280
                   const uchar *s, const uchar *e)
1281 1282 1283 1284
{
  if (s >= e)
  {
    *ctype= 0;
unknown's avatar
unknown committed
1285
    return MY_CS_TOOSMALL;
1286
  }
1287
  *ctype= cs->ctype[*s + 1];
1288 1289 1290 1291
  return 1;
}


1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378
#define CUTOFF  (ULONGLONG_MAX / 10)
#define CUTLIM  (ULONGLONG_MAX % 10)
#define DIGITS_IN_ULONGLONG 20

static ulonglong d10[DIGITS_IN_ULONGLONG]=
{
  1,
  10,
  100,
  1000,
  10000,
  100000,
  1000000,
  10000000,
  100000000,
  1000000000,
  10000000000ULL,
  100000000000ULL,
  1000000000000ULL,
  10000000000000ULL,
  100000000000000ULL,
  1000000000000000ULL,
  10000000000000000ULL,
  100000000000000000ULL,
  1000000000000000000ULL,
  10000000000000000000ULL
};


/*

  Convert a string to unsigned long long integer value
  with rounding.
  
  SYNOPSYS
    my_strntoull10_8bit()
      cs              in      pointer to character set
      str             in      pointer to the string to be converted
      length          in      string length
      unsigned_flag   in      whether the number is unsigned
      endptr          out     pointer to the stop character
      error           out     returned error code

  DESCRIPTION
    This function takes the decimal representation of integer number
    from string str and converts it to an signed or unsigned
    long long integer value.
    Space characters and tab are ignored.
    A sign character might precede the digit characters.
    The number may have any number of pre-zero digits.
    The number may have decimal point and exponent.
    Rounding is always done in "away from zero" style:
      0.5  ->   1
     -0.5  ->  -1

    The function stops reading the string str after "length" bytes
    or at the first character that is not a part of correct number syntax:

    <signed numeric literal> ::=
      [ <sign> ] <exact numeric literal> [ E [ <sign> ] <unsigned integer> ]

    <exact numeric literal> ::=
                        <unsigned integer> [ <period> [ <unsigned integer> ] ]
                      | <period> <unsigned integer>
    <unsigned integer>   ::= <digit>...
     
  RETURN VALUES
    Value of string as a signed/unsigned longlong integer

    endptr cannot be NULL. The function will store the end pointer
    to the stop character here.

    The error parameter contains information how things went:
    0	     ok
    ERANGE   If the the value of the converted number is out of range
    In this case the return value is:
    - ULONGLONG_MAX if unsigned_flag and the number was too big
    - 0 if unsigned_flag and the number was negative
    - LONGLONG_MAX if no unsigned_flag and the number is too big
    - LONGLONG_MIN if no unsigned_flag and the number it too big negative
    
    EDOM If the string didn't contain any digits.
    In this case the return value is 0.
*/

ulonglong
my_strntoull10rnd_8bit(CHARSET_INFO *cs __attribute__((unused)),
1379
                       const char *str, size_t length, int unsigned_flag,
1380 1381 1382 1383 1384
                       char **endptr, int *error)
{
  const char *dot, *end9, *beg, *end= str + length;
  ulonglong ull;
  ulong ul;
1385
  uchar ch;
1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402
  int shift= 0, digits= 0, negative, addon;

  /* Skip leading spaces and tabs */
  for ( ; str < end && (*str == ' ' || *str == '\t') ; str++);

  if (str >= end)
    goto ret_edom;

  if ((negative= (*str == '-')) || *str=='+') /* optional sign */
  {
    if (++str == end)
      goto ret_edom;
  }

  beg= str;
  end9= (str + 9) > end ? end : (str + 9);
  /* Accumulate small number into ulong, for performance purposes */
1403
  for (ul= 0 ; str < end9 && (ch= (uchar) (*str - '0')) < 10; str++)
1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420
  {
    ul= ul * 10 + ch;
  }
  
  if (str >= end) /* Small number without dots and expanents */
  {
    *endptr= (char*) str;
    if (negative)
    {
      if (unsigned_flag)
      {
        *error= ul ? MY_ERRNO_ERANGE : 0;
        return 0;
      }
      else
      {
        *error= 0;
1421
        return (ulonglong) (longlong) -(long) ul;
1422 1423 1424 1425 1426 1427 1428 1429 1430
      }
    }
    else
    {
      *error=0;
      return (ulonglong) ul;
    }
  }
  
1431
  digits= (int) (str - beg);
1432 1433 1434 1435

  /* Continue to accumulate into ulonglong */
  for (dot= NULL, ull= ul; str < end; str++)
  {
1436
    if ((ch= (uchar) (*str - '0')) < 10)
1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456
    {
      if (ull < CUTOFF || (ull == CUTOFF && ch <= CUTLIM))
      {
        ull= ull * 10 + ch;
        digits++;
        continue;
      }
      /*
        Adding the next digit would overflow.
        Remember the next digit in "addon", for rounding.
        Scan all digits with an optional single dot.
      */
      if (ull == CUTOFF)
      {
        ull= ULONGLONG_MAX;
        addon= 1;
        str++;
      }
      else
        addon= (*str >= '5');
1457
      if (!dot)
1458
      {
1459
        for ( ; str < end && (ch= (uchar) (*str - '0')) < 10; shift++, str++);
1460 1461 1462
        if (str < end && *str == '.')
        {
          str++;
1463
          for ( ; str < end && (ch= (uchar) (*str - '0')) < 10; str++);
1464
        }
1465
      }
1466
      else
1467
      {
1468
        shift= (int) (dot - str);
1469
        for ( ; str < end && (ch= (uchar) (*str - '0')) < 10; str++);
1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507
      }
      goto exp;
    }
    
    if (*str == '.')
    {
      if (dot)
      {
        /* The second dot character */
        addon= 0;
        goto exp;
      }
      else
      {
        dot= str + 1;
      }
      continue;
    }
    
    /* Unknown character, exit the loop */
    break; 
  }
  shift= dot ? dot - str : 0; /* Right shift */
  addon= 0;

exp:    /* [ E [ <sign> ] <unsigned integer> ] */

  if (!digits)
  {
    str= beg;
    goto ret_edom;
  }
  
  if (str < end && (*str == 'e' || *str == 'E'))
  {
    str++;
    if (str < end)
    {
1508
      int negative_exp, exponent;
1509 1510 1511 1512 1513
      if ((negative_exp= (*str == '-')) || *str=='+')
      {
        if (++str == end)
          goto ret_sign;
      }
1514
      for (exponent= 0 ;
1515
           str < end && (ch= (uchar) (*str - '0')) < 10;
1516 1517
           str++)
      {
1518
        exponent= exponent * 10 + ch;
1519
      }
1520
      shift+= negative_exp ? -exponent : exponent;
1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575
    }
  }
  
  if (shift == 0) /* No shift, check addon digit */
  {
    if (addon)
    {
      if (ull == ULONGLONG_MAX)
        goto ret_too_big;
      ull++;
    }
    goto ret_sign;
  }

  if (shift < 0) /* Right shift */
  {
    ulonglong d, r;
    
    if (-shift >= DIGITS_IN_ULONGLONG)
      goto ret_zero; /* Exponent is a big negative number, return 0 */
    
    d= d10[-shift];
    r= (ull % d) * 2;
    ull /= d;
    if (r >= d)
      ull++;
    goto ret_sign;
  }

  if (shift > DIGITS_IN_ULONGLONG) /* Huge left shift */
  {
    if (!ull)
      goto ret_sign;
    goto ret_too_big;
  }

  for ( ; shift > 0; shift--, ull*= 10) /* Left shift */
  {
    if (ull > CUTOFF)
      goto ret_too_big; /* Overflow, number too big */
  }

ret_sign:
  *endptr= (char*) str;

  if (!unsigned_flag)
  {
    if (negative)
    {
      if (ull > (ulonglong) LONGLONG_MIN)
      {
        *error= MY_ERRNO_ERANGE;
        return (ulonglong) LONGLONG_MIN;
      }
      *error= 0;
1576
      return (ulonglong) -(longlong) ull;
1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617
    }
    else
    {
      if (ull > (ulonglong) LONGLONG_MAX)
      {
        *error= MY_ERRNO_ERANGE;
        return (ulonglong) LONGLONG_MAX;
      }
      *error= 0;
      return ull;
    }
  }

  /* Unsigned number */
  if (negative && ull)
  {
    *error= MY_ERRNO_ERANGE;
    return 0;
  }
  *error= 0;
  return ull;

ret_zero:
  *endptr= (char*) str;
  *error= 0;
  return 0;

ret_edom:
  *endptr= (char*) str;
  *error= MY_ERRNO_EDOM;
  return 0;
  
ret_too_big:
  *endptr= (char*) str;
  *error= MY_ERRNO_ERANGE;
  return unsigned_flag ?
         ULONGLONG_MAX :
         negative ? (ulonglong) LONGLONG_MIN : (ulonglong) LONGLONG_MAX;
}


1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657
/*
  Check if a constant can be propagated

  SYNOPSIS:
    my_propagate_simple()
    cs		Character set information
    str		String to convert to double
    length	Optional length for string.
    
  NOTES:
   Takes the string in the given charset and check
   if it can be safely propagated in the optimizer.
   
   create table t1 (
     s char(5) character set latin1 collate latin1_german2_ci);
   insert into t1 values (0xf6); -- o-umlaut
   select * from t1 where length(s)=1 and s='oe';

   The above query should return one row.
   We cannot convert this query into:
   select * from t1 where length('oe')=1 and s='oe';
   
   Currently we don't check the constant itself,
   and decide not to propagate a constant
   just if the collation itself allows tricky things
   like expansions and contractions. In the future
   we can write a more sophisticated functions to
   check the constants. For example, 'oa' can always
   be safety propagated in German2 because unlike 
   'oe' it does not have any special meaning.

  RETURN
    1 if constant can be safely propagated
    0 if it is not safe to propagate the constant
*/



my_bool my_propagate_simple(CHARSET_INFO *cs __attribute__((unused)),
                            const uchar *str __attribute__((unused)),
1658
                            size_t length __attribute__((unused)))
1659 1660 1661 1662 1663 1664 1665
{
  return 1;
}


my_bool my_propagate_complex(CHARSET_INFO *cs __attribute__((unused)),
                             const uchar *str __attribute__((unused)),
1666
                             size_t length __attribute__((unused)))
1667 1668 1669 1670 1671
{
  return 0;
}


1672 1673
MY_CHARSET_HANDLER my_charset_8bit_handler=
{
1674
    my_cset_init_8bit,
1675
    NULL,			/* ismbchar      */
1676
    my_mbcharlen_8bit,		/* mbcharlen     */
1677 1678
    my_numchars_8bit,
    my_charpos_8bit,
1679
    my_well_formed_len_8bit,
unknown's avatar
unknown committed
1680
    my_lengthsp_8bit,
1681
    my_numcells_8bit,
1682 1683
    my_mb_wc_8bit,
    my_wc_mb_8bit,
1684
    my_mb_ctype_8bit,
1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697
    my_caseup_str_8bit,
    my_casedn_str_8bit,
    my_caseup_8bit,
    my_casedn_8bit,
    my_snprintf_8bit,
    my_long10_to_str_8bit,
    my_longlong10_to_str_8bit,
    my_fill_8bit,
    my_strntol_8bit,
    my_strntoul_8bit,
    my_strntoll_8bit,
    my_strntoull_8bit,
    my_strntod_8bit,
1698
    my_strtoll10_8bit,
1699
    my_strntoull10rnd_8bit,
1700 1701 1702 1703 1704
    my_scan_8bit
};

MY_COLLATION_HANDLER my_collation_8bit_simple_ci_handler =
{
1705
    my_coll_init_simple,	/* init */
1706 1707 1708
    my_strnncoll_simple,
    my_strnncollsp_simple,
    my_strnxfrm_simple,
1709
    my_strnxfrmlen_simple,
1710 1711 1712
    my_like_range_simple,
    my_wildcmp_8bit,
    my_strcasecmp_8bit,
1713
    my_instr_simple,
1714 1715
    my_hash_sort_simple,
    my_propagate_simple
1716
};