mariadb/strings/strcoll.inl
Alexander Barkov fd247cc21f MDEV-31340 Remove MY_COLLATION_HANDLER::strcasecmp()
This patch also fixes:
  MDEV-33050 Build-in schemas like oracle_schema are accent insensitive
  MDEV-33084 LASTVAL(t1) and LASTVAL(T1) do not work well with lower-case-table-names=0
  MDEV-33085 Tables T1 and t1 do not work well with ENGINE=CSV and lower-case-table-names=0
  MDEV-33086 SHOW OPEN TABLES IN DB1 -- is case insensitive with lower-case-table-names=0
  MDEV-33088 Cannot create triggers in the database `MYSQL`
  MDEV-33103 LOCK TABLE t1 AS t2 -- alias is not case sensitive with lower-case-table-names=0
  MDEV-33109 DROP DATABASE MYSQL -- does not drop SP with lower-case-table-names=0
  MDEV-33110 HANDLER commands are case insensitive with lower-case-table-names=0
  MDEV-33119 User is case insensitive in INFORMATION_SCHEMA.VIEWS
  MDEV-33120 System log table names are case insensitive with lower-cast-table-names=0

- Removing the virtual function strnncoll() from MY_COLLATION_HANDLER

- Adding a wrapper function CHARSET_INFO::streq(), to compare
  two strings for equality. For now it calls strnncoll() internally.
  In the future it will turn into a virtual function.

- Adding new accent sensitive case insensitive collations:
    - utf8mb4_general1400_as_ci
    - utf8mb3_general1400_as_ci
  They implement accent sensitive case insensitive comparison.
  The weight of a character is equal to the code point of its
  upper case variant. These collations use Unicode-14.0.0 casefolding data.

  The result of
     my_charset_utf8mb3_general1400_as_ci.strcoll()
  is very close to the former
     my_charset_utf8mb3_general_ci.strcasecmp()

  There is only a difference in a couple dozen rare characters, because:
    - the switch from "tolower" to "toupper" comparison, to make
      utf8mb3_general1400_as_ci closer to utf8mb3_general_ci
    - the switch from Unicode-3.0.0 to Unicode-14.0.0
  This difference should be tolarable. See the list of affected
  characters in the MDEV description.

  Note, utf8mb4_general1400_as_ci correctly handles non-BMP characters!
  Unlike utf8mb4_general_ci, it does not treat all BMP characters
  as equal.

- Adding classes representing names of the file based database objects:

    Lex_ident_db
    Lex_ident_table
    Lex_ident_trigger

  Their comparison collation depends on the underlying
  file system case sensitivity and on --lower-case-table-names
  and can be either my_charset_bin or my_charset_utf8mb3_general1400_as_ci.

- Adding classes representing names of other database objects,
  whose names have case insensitive comparison style,
  using my_charset_utf8mb3_general1400_as_ci:

  Lex_ident_column
  Lex_ident_sys_var
  Lex_ident_user_var
  Lex_ident_sp_var
  Lex_ident_ps
  Lex_ident_i_s_table
  Lex_ident_window
  Lex_ident_func
  Lex_ident_partition
  Lex_ident_with_element
  Lex_ident_rpl_filter
  Lex_ident_master_info
  Lex_ident_host
  Lex_ident_locale
  Lex_ident_plugin
  Lex_ident_engine
  Lex_ident_server
  Lex_ident_savepoint
  Lex_ident_charset
  engine_option_value::Name

- All the mentioned Lex_ident_xxx classes implement a method streq():

  if (ident1.streq(ident2))
     do_equal();

  This method works as a wrapper for CHARSET_INFO::streq().

- Changing a lot of "LEX_CSTRING name" to "Lex_ident_xxx name"
  in class members and in function/method parameters.

- Replacing all calls like
    system_charset_info->coll->strcasecmp(ident1, ident2)
  to
    ident1.streq(ident2)

- Taking advantage of the c++11 user defined literal operator
  for LEX_CSTRING (see m_strings.h) and Lex_ident_xxx (see lex_ident.h)
  data types. Use example:

  const Lex_ident_column primary_key_name= "PRIMARY"_Lex_ident_column;

  is now a shorter version of:

  const Lex_ident_column primary_key_name=
    Lex_ident_column({STRING_WITH_LEN("PRIMARY")});
2024-04-18 15:22:10 +04:00

745 lines
22 KiB
C++

/*
Copyright (c) 2015, MariaDB Foundation
Copyright (c) 2015, 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 02110-1335 USA
*/
#include "ctype-ascii.h"
#ifndef MY_FUNCTION_NAME
#error MY_FUNCTION_NAME is not defined
#endif
/*
Define strnncoll() and strnncollsp() by default,
unless "#define DEFINE_STRNNCOLL 0" is specified.
*/
#ifndef DEFINE_STRNNCOLL
#define DEFINE_STRNNCOLL 1
#endif
/*
The weight for automatically padded spaces when comparing strings with
the PAD SPACE property.
Should normally be equal to the weight of a regular space.
*/
#ifndef WEIGHT_PAD_SPACE
#define WEIGHT_PAD_SPACE (' ')
#endif
/*
For binary collations:
- on 32bit platforms perform only 4 byte optimization
- on 64bit platforms perform both 4 byte and 8 byte optimization
*/
#if defined(STRCOLL_MB7_BIN)
#define MY_STRCOLL_MB7_4BYTES(a,b) my_strcoll_mb7_bin_4bytes((a),(b))
#if SIZEOF_VOIDP == 8
#define STRCOLL_MB7_8BYTES
#define MY_STRCOLL_MB7_8BYTES(a,b) my_strcoll_mb7_bin_8bytes((a),(b))
#endif /* Architecture test */
#endif /* STRCOLL_MB7_BIN */
/*
For case insensitive collations with trivial mapping from [a-z] to [A-Z]
perform optimization only on 64 bit platforms.
There is no sense to perform my_ascii_to_upper_magic_uint64() based
optimization on 32bit platforms. The idea of this optimization
is that it handles 8bytes at a time, using 64bit CPU registers.
Enabling this optimization on 32bit platform may only slow things down.
*/
#if defined(STRCOLL_MB7_TOUPPER)
#if SIZEOF_VOIDP == 8
#define MY_STRCOLL_MB7_4BYTES(a,b) my_strcoll_ascii_toupper_4bytes((a),(b))
#define MY_STRCOLL_MB7_8BYTES(a,b) my_strcoll_ascii_toupper_8bytes((a),(b))
#endif /* Architecture test */
#endif /* STRCOLL_MB7_TOUPPER */
/*
A helper macro to shift two pointers forward, to the given amount.
*/
#define MY_STRING_SHIFT_PTR_PTR(a,b,len) do { a+= len; b+= len; } while(0)
/*
Weight of an illegal byte, must follow these rules:
1. Must be greater than weight of any normal character in the collation.
2. Two different bad bytes must have different weights and must be
compared in their binary order.
Depends on mbmaxlen of the character set, as well as how the collation
sorts various single-byte and multi-byte character blocks.
The macro below is the default definition, it is suitable for mbmaxlen=2
character sets that sort all multi-byte characters after all single-byte
characters: big5, euckr, gb2312, gbk.
All mbmaxlen>2 character sets must provide their own definitions.
All collations that have a more complex order (than just MB1 followed by MB2)
must also provide their own definitions (see definitions for
cp932_japanese_ci and sjis_japanese_ci as examples of a more complex order).
*/
#ifndef WEIGHT_ILSEQ
#define WEIGHT_ILSEQ(x) (0xFF00 + (x))
#endif
#if defined(WEIGHT_SIZE) && WEIGHT_SIZE == 3
#define PUT_WC_BE_HAVE_1BYTE(dst, de, wc) PUT_WC_BE3_HAVE_1BYTE((dst), (de), (wc))
#define PAD_NWEIGHTS_UNICODE_BE(str, end, n) my_strxfrm_pad_nweights_unicode_be3(str, end, n)
#define PAD_UNICODE_BE(str, end) my_strxfrm_pad_unicode_be3(str, end)
#else
#define PUT_WC_BE_HAVE_1BYTE(dst, de, wc) PUT_WC_BE2_HAVE_1BYTE((dst), (de), (wc))
#define PAD_NWEIGHTS_UNICODE_BE(str, end, n) my_strxfrm_pad_nweights_unicode_be2(str, end, n)
#define PAD_UNICODE_BE(str, end) my_strxfrm_pad_unicode_be2(str, end)
#endif
#if DEFINE_STRNNCOLL
/**
Scan a valid character, or a bad byte, or an auto-padded space
from a string and calculate the weight of the scanned sequence.
@param [OUT] weight - the weight is returned here
@param str - the string
@param end - the end of the string
@return - the number of bytes scanned
The including source file must define the following macros:
IS_MB1_CHAR(b0) - for character sets that have MB1 characters
IS_MB1_MB2HEAD_GAP(b0) - optional, for better performance
IS_MB2_CHAR(b0,b1) - for character sets that have MB2 characters
IS_MB3_CHAR(b0,b1,b2) - for character sets that have MB3 characters
IS_MB4_CHAR(b0,b1,b2,b3) - for character sets with have MB4 characters
WEIGHT_PAD_SPACE
WEIGHT_MB1(b0) - for character sets that have MB1 characters
WEIGHT_MB2(b0,b1) - for character sets that have MB2 characters
WEIGHT_MB3(b0,b1,b2) - for character sets that have MB3 characters
WEIGHT_MB4(b0,b1,b2,b3) - for character sets that have MB4 characters
WEIGHT_ILSEQ(x)
*/
static inline uint
MY_FUNCTION_NAME(scan_weight)(int *weight, const uchar *str, const uchar *end)
{
if (str >= end)
{
*weight= WEIGHT_PAD_SPACE;
return 0;
}
#ifdef IS_MB1_CHAR
if (IS_MB1_CHAR(*str))
{
*weight= WEIGHT_MB1(*str); /* A valid single byte character*/
return 1;
}
#endif
#ifdef IS_MB1_MBHEAD_UNUSED_GAP
/*
Quickly filter out unused bytes that are neither MB1 nor MBHEAD.
E.g. [0x80..0xC1] in utf8mb(3|4). This allows using simplified conditions
in IS_MB2_CHAR(), IS_MB3_CHAR(), etc.
*/
if (IS_MB1_MBHEAD_UNUSED_GAP(*str))
goto bad;
#endif
#ifdef IS_MB2_CHAR
if (str + 2 > end) /* The string ended unexpectedly */
goto bad; /* Treat as a bad byte */
if (IS_MB2_CHAR(str[0], str[1]))
{
*weight= WEIGHT_MB2(str[0], str[1]);
return 2; /* A valid two-byte character */
}
#endif
#ifdef IS_MB3_CHAR
if (str + 3 > end) /* Incomplete three-byte character */
goto bad;
if (IS_MB3_CHAR(str[0], str[1], str[2]))
{
*weight= WEIGHT_MB3(str[0], str[1], str[2]);
return 3; /* A valid three-byte character */
}
#endif
#ifdef IS_MB4_CHAR
if (str + 4 > end) /* Incomplete four-byte character */
goto bad;
if (IS_MB4_CHAR(str[0], str[1], str[2], str[3]))
{
*weight= WEIGHT_MB4(str[0], str[1], str[2], str[3]);
return 4; /* A valid four-byte character */
}
#endif
bad:
*weight= WEIGHT_ILSEQ(str[0]); /* Bad byte */
return 1;
}
/**
Compare two strings according to the collation,
without handling the PAD SPACE property.
Note, strnncoll() is usually used to compare identifiers.
Perhaps we should eventually (in 10.2?) create a new collation
my_charset_utf8mb3_general_ci_no_pad and have only one comparison function
in MY_COLLATION_HANDLER.
@param cs - the character set and collation
@param a - the left string
@param a_length - the length of the left string
@param b - the right string
@param b_length - the length of the right string
@param b_is_prefix - if the caller wants to check if "b" is a prefix of "a"
@return - the comparison result
*/
static int
MY_FUNCTION_NAME(strnncoll)(CHARSET_INFO *cs __attribute__((unused)),
const uchar *a, size_t a_length,
const uchar *b, size_t b_length,
my_bool b_is_prefix)
{
const uchar *a_end= a + a_length;
const uchar *b_end= b + b_length;
for ( ; ; )
{
int a_weight, b_weight, res;
uint a_wlen= MY_FUNCTION_NAME(scan_weight)(&a_weight, a, a_end);
uint b_wlen;
#ifdef MY_STRCOLL_MB7_4BYTES
if (a_wlen == 1 && my_strcoll_ascii_4bytes_found(a, a_end, b, b_end))
{
int res;
#ifdef MY_STRCOLL_MB7_8BYTES
/*TODO: a a loop here >='a' <='z' here, for automatic vectorization*/
if (my_strcoll_ascii_4bytes_found(a + 4, a_end, b + 4, b_end))
{
if ((res= MY_STRCOLL_MB7_8BYTES(a, b)))
return res;
MY_STRING_SHIFT_PTR_PTR(a, b, 8);
continue;
}
#endif
if ((res= MY_STRCOLL_MB7_4BYTES(a, b)))
return res;
MY_STRING_SHIFT_PTR_PTR(a, b, 4);
continue;
}
#endif /* MY_STRCOLL_MB7_4BYTES */
b_wlen= MY_FUNCTION_NAME(scan_weight)(&b_weight, b, b_end);
/*
a_wlen b_wlen Comment
------ ------ -------
0 0 Strings ended simultaneously, "a" and "b" are equal.
0 >0 "a" is a prefix of "b", so "a" is smaller.
>0 0 "b" is a prefix of "a", check b_is_prefix.
>0 >0 Two weights were scanned, check weight difference.
*/
if (!a_wlen)
return b_wlen ? -1 : 0;
if (!b_wlen)
return b_is_prefix ? 0 : +1;
if ((res= (a_weight - b_weight)))
return res;
/*
None of the strings has ended yet.
*/
DBUG_ASSERT(a < a_end);
DBUG_ASSERT(b < b_end);
a+= a_wlen;
b+= b_wlen;
}
DBUG_ASSERT(0);
return 0;
}
#ifdef DEFINE_STRNNCOLLSP_NOPAD
/**
Compare two strings according to the collation, with NO PAD handling.
@param cs - the character set and collation
@param a - the left string
@param a_length - the length of the left string
@param b - the right string
@param b_length - the length of the right string
@return - the comparison result
*/
static int
MY_FUNCTION_NAME(strnncollsp)(CHARSET_INFO *cs __attribute__((unused)),
const uchar *a, size_t a_length,
const uchar *b, size_t b_length)
{
return MY_FUNCTION_NAME(strnncoll)(cs, a, a_length, b, b_length, FALSE);
}
#else
/**
Compare two strings according to the collation, with PAD SPACE handling.
@param cs - the character set and collation
@param a - the left string
@param a_length - the length of the left string
@param b - the right string
@param b_length - the length of the right string
@return - the comparison result
*/
static int
MY_FUNCTION_NAME(strnncollsp)(CHARSET_INFO *cs __attribute__((unused)),
const uchar *a, size_t a_length,
const uchar *b, size_t b_length)
{
const uchar *a_end= a + a_length;
const uchar *b_end= b + b_length;
for ( ; ; )
{
int a_weight, b_weight, res;
uint a_wlen= MY_FUNCTION_NAME(scan_weight)(&a_weight, a, a_end);
uint b_wlen;
#ifdef MY_STRCOLL_MB7_4BYTES
if (a_wlen == 1 && my_strcoll_ascii_4bytes_found(a, a_end, b, b_end))
{
int res;
#ifdef MY_STRCOLL_MB7_8BYTES
if (my_strcoll_ascii_4bytes_found(a + 4, a_end, b + 4, b_end))
{
if ((res= MY_STRCOLL_MB7_8BYTES(a, b)))
return res;
MY_STRING_SHIFT_PTR_PTR(a, b, 8);
continue;
}
#endif
if ((res= MY_STRCOLL_MB7_4BYTES(a, b)))
return res;
MY_STRING_SHIFT_PTR_PTR(a, b, 4);
continue;
}
#endif /* MY_STRCOLL_MB7_4BYTES */
b_wlen= MY_FUNCTION_NAME(scan_weight)(&b_weight, b, b_end);
if ((res= (a_weight - b_weight)))
{
/*
Got two different weights. Each weight can be generated by either of:
- a real character
- a bad byte sequence or an incomplete byte sequence
- an auto-generated trailing space (PAD SPACE)
It does not matter how exactly each weight was generated.
Just return the weight difference.
*/
return res;
}
if (!a_wlen && !b_wlen)
{
/*
Got two auto-generated trailing spaces, i.e.
both strings have now ended, so they are equal.
*/
DBUG_ASSERT(a == a_end);
DBUG_ASSERT(b == b_end);
return 0;
}
/*
At least one of the strings has not ended yet, continue comparison.
*/
DBUG_ASSERT(a < a_end || b < b_end);
a+= a_wlen;
b+= b_wlen;
}
DBUG_ASSERT(0);
return 0;
}
#endif /* DEFINE_STRNNCOLLSP_NOPAD */
/**
Compare two strings according to the collation,
with trailing space padding or trimming, according to "nchars".
@param cs - the character set and collation
@param a - the left string
@param a_length - the length of the left string
@param b - the right string
@param b_length - the length of the right string
@param nchars - compare this amount of characters only
@return - the comparison result
*/
static int
MY_FUNCTION_NAME(strnncollsp_nchars)(CHARSET_INFO *cs __attribute__((unused)),
const uchar *a, size_t a_length,
const uchar *b, size_t b_length,
size_t nchars,
uint flags)
{
const uchar *a_end= a + a_length;
const uchar *b_end= b + b_length;
for ( ; nchars ; nchars--)
{
int a_weight, b_weight, res;
uint a_wlen= MY_FUNCTION_NAME(scan_weight)(&a_weight, a, a_end);
uint b_wlen= MY_FUNCTION_NAME(scan_weight)(&b_weight, b, b_end);
if ((res= (a_weight - b_weight)))
{
/* Got two different weights. See comments in strnncollsp above. */
return res;
}
if (!a_wlen && !b_wlen)
{
/* Got two auto-generated trailing spaces. */
DBUG_ASSERT(a == a_end);
DBUG_ASSERT(b == b_end);
return 0;
}
/*
At least one of the strings has not ended yet, continue comparison.
*/
DBUG_ASSERT(a < a_end || b < b_end);
a+= a_wlen;
b+= b_wlen;
}
return 0;
}
#endif /* DEFINE_STRNNCOLL */
#ifdef DEFINE_STRNXFRM
#ifndef WEIGHT_MB2_FRM
#define WEIGHT_MB2_FRM(x,y) WEIGHT_MB2(x,y)
#endif
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;
const uchar *se= src + srclen;
const uchar *sort_order= cs->sort_order;
for (; dst < de && src < se && nweights; nweights--)
{
if (my_ci_charlen(cs, src, se) > 1)
{
/*
Note, it is safe not to check (src < se)
in the code below, because my_ci_charlen() would
not return 2 if src was too short
*/
uint16 e= WEIGHT_MB2_FRM(src[0], src[1]);
*dst++= (uchar) (e >> 8);
if (dst < de)
*dst++= (uchar) (e & 0xFF);
src+= 2;
}
else
*dst++= sort_order ? sort_order[*src++] : *src++;
}
#ifdef DEFINE_STRNNCOLLSP_NOPAD
return my_strxfrm_pad_desc_and_reverse_nopad(cs, d0, dst, de,
nweights, flags, 0);
#else
return my_strxfrm_pad_desc_and_reverse(cs, d0, dst, de, nweights, flags, 0);
#endif
}
#endif /* DEFINE_STRNXFRM */
#if defined(DEFINE_STRNXFRM_UNICODE) || defined(DEFINE_STRNXFRM_UNICODE_NOPAD)
/*
Store sorting weights using 2 bytes per character.
This function is shared between
- utf8mb3_general_ci, utf8mb3_bin, ucs2_general_ci, ucs2_bin
which support BMP only (U+0000..U+FFFF).
- utf8mb4_general_ci, utf16_general_ci, utf32_general_ci,
which map all supplementary characters to weight 0xFFFD.
*/
#ifndef MY_MB_WC
#error MY_MB_WC must be defined for DEFINE_STRNXFRM_UNICODE
#endif
#ifndef OPTIMIZE_ASCII
#error OPTIMIZE_ASCII must be defined for DEFINE_STRNXFRM_UNICODE
#endif
#if OPTIMIZE_ASCII && !defined(WEIGHT_MB1)
#error WEIGHT_MB1 must be defined for DEFINE_STRNXFRM_UNICODE
#endif
#ifndef MY_WC_WEIGHT
#error MY_WC_WEIGHT must be defined for DEFINE_STRNXFRM_UNICODE
#endif
static size_t
MY_FUNCTION_NAME(strnxfrm_internal)(CHARSET_INFO *cs __attribute__((unused)),
uchar *dst, uchar *de,
uint *nweights,
const uchar *src, const uchar *se)
{
my_wc_t UNINIT_VAR(wc);
uchar *dst0= dst;
DBUG_ASSERT(src || !se);
DBUG_ASSERT((cs->state & MY_CS_LOWER_SORT) == 0);
for (; dst < de && *nweights; (*nweights)--)
{
int res;
#if OPTIMIZE_ASCII
if (src >= se)
break;
if (src[0] <= 0x7F)
{
wc= WEIGHT_MB1(*src++);
PUT_WC_BE_HAVE_1BYTE(dst, de, wc);
continue;
}
#endif
if ((res= MY_MB_WC(cs, &wc, src, se)) <= 0)
break;
src+= res;
wc= MY_WC_WEIGHT(wc);
PUT_WC_BE_HAVE_1BYTE(dst, de, wc);
}
return dst - dst0;
}
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 *dst0= dst;
uchar *de= dst + dstlen;
dst+= MY_FUNCTION_NAME(strnxfrm_internal)(cs, dst, de, &nweights,
src, src + srclen);
DBUG_ASSERT(dst <= de); /* Safety */
if (dst < de && nweights && (flags & MY_STRXFRM_PAD_WITH_SPACE))
dst+= PAD_NWEIGHTS_UNICODE_BE(dst, de, nweights);
my_strxfrm_desc_and_reverse(dst0, dst, flags, 0);
if ((flags & MY_STRXFRM_PAD_TO_MAXLEN) && dst < de)
dst+= PAD_UNICODE_BE(dst, de);
return dst - dst0;
}
#ifdef DEFINE_STRNXFRM_UNICODE_NOPAD
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 *dst0= dst;
uchar *de= dst + dstlen;
dst+= MY_FUNCTION_NAME(strnxfrm_internal)(cs, dst, de, &nweights,
src, src + srclen);
DBUG_ASSERT(dst <= de); /* Safety */
if (dst < de && nweights && (flags & MY_STRXFRM_PAD_WITH_SPACE))
{
size_t len= de - dst;
set_if_smaller(len, nweights * 2);
memset(dst, 0x00, len);
dst+= len;
}
my_strxfrm_desc_and_reverse(dst0, dst, flags, 0);
if ((flags & MY_STRXFRM_PAD_TO_MAXLEN) && dst < de)
{
memset(dst, 0x00, de - dst);
dst= de;
}
return dst - dst0;
}
#endif
#endif /* DEFINE_STRNXFRM_UNICODE || DEFINE_STRNXFRM_UNICODE_NOPAD */
#ifdef DEFINE_STRNXFRM_UNICODE_BIN2
/*
Store sorting weights using 2 bytes per character.
These functions are shared between
- utf8mb3_general_ci, utf8mb3_bin, ucs2_general_ci, ucs2_bin
which support BMP only (U+0000..U+FFFF).
- utf8mb4_general_ci, utf16_general_ci, utf32_general_ci,
which map all supplementary characters to weight 0xFFFD.
*/
#ifndef MY_MB_WC
#error MY_MB_WC must be defined for DEFINE_STRNXFRM_UNICODE_BIN2
#endif
#ifndef OPTIMIZE_ASCII
#error OPTIMIZE_ASCII must be defined for DEFINE_STRNXFRM_UNICODE_BIN2
#endif
static size_t
MY_FUNCTION_NAME(strnxfrm_internal)(CHARSET_INFO *cs __attribute__((unused)),
uchar *dst, uchar *de,
uint *nweights,
const uchar *src,
const uchar *se)
{
my_wc_t UNINIT_VAR(wc);
uchar *dst0= dst;
DBUG_ASSERT(src || !se);
for (; dst < de && *nweights; (*nweights)--)
{
int res;
#if OPTIMIZE_ASCII
if (src >= se)
break;
if (src[0] <= 0x7F)
{
wc= *src++;
PUT_WC_BE2_HAVE_1BYTE(dst, de, wc);
continue;
}
#endif
if ((res= MY_MB_WC(cs, &wc, src, se)) <= 0)
break;
src+= res;
if (wc > 0xFFFF)
wc= MY_CS_REPLACEMENT_CHARACTER;
PUT_WC_BE2_HAVE_1BYTE(dst, de, wc);
}
return dst - dst0;
}
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 *dst0= dst;
uchar *de= dst + dstlen;
dst+= MY_FUNCTION_NAME(strnxfrm_internal)(cs, dst, de, &nweights,
src, src + srclen);
DBUG_ASSERT(dst <= de); /* Safety */
if (dst < de && nweights && (flags & MY_STRXFRM_PAD_WITH_SPACE))
dst+= PAD_NWEIGHTS_UNICODE_BE(dst, de, nweights);
my_strxfrm_desc_and_reverse(dst0, dst, flags, 0);
if ((flags & MY_STRXFRM_PAD_TO_MAXLEN) && dst < de)
dst+= PAD_UNICODE_BE(dst, de);
return dst - dst0;
}
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 *dst0= dst;
uchar *de= dst + dstlen;
dst+= MY_FUNCTION_NAME(strnxfrm_internal)(cs, dst, de, &nweights,
src, src + srclen);
DBUG_ASSERT(dst <= de); /* Safety */
if (dst < de && nweights && (flags & MY_STRXFRM_PAD_WITH_SPACE))
{
size_t len= de - dst;
set_if_smaller(len, nweights * 2);
memset(dst, 0x00, len);
dst+= len;
}
my_strxfrm_desc_and_reverse(dst0, dst, flags, 0);
if ((flags & MY_STRXFRM_PAD_TO_MAXLEN) && dst < de)
{
memset(dst, 0x00, de - dst);
dst= de;
}
return dst - dst0;
}
#endif /* DEFINE_STRNXFRM_UNICODE_BIN2 */
/*
We usually include this file at least two times from the same source file,
for the _ci and the _bin collations. Prepare for the second inclusion.
*/
#undef MY_FUNCTION_NAME
#undef MY_MB_WC
#undef OPTIMIZE_ASCII
#undef MY_WC_WEIGHT
#undef WEIGHT_ILSEQ
#undef WEIGHT_MB1
#undef WEIGHT_MB2
#undef WEIGHT_MB3
#undef WEIGHT_MB4
#undef WEIGHT_PAD_SPACE
#undef WEIGHT_MB2_FRM
#undef WEIGHT_SIZE
#undef DEFINE_STRNXFRM
#undef DEFINE_STRNXFRM_UNICODE
#undef DEFINE_STRNXFRM_UNICODE_NOPAD
#undef DEFINE_STRNXFRM_UNICODE_BIN2
#undef DEFINE_STRNNCOLL
#undef DEFINE_STRNNCOLLSP_NOPAD
#undef STRCOLL_MB7_TOUPPER
#undef STRCOLL_MB7_BIN
#undef MY_STRCOLL_MB7_4BYTES
#undef MY_STRCOLL_MB7_8BYTES
#undef PUT_WC_BE_HAVE_1BYTE
#undef PAD_NWEIGHTS_UNICODE_BE
#undef PAD_UNICODE_BE