Skip to content

M
MariaDB
Commit 6eae037c authored by Alexander Barkov's avatar Alexander Barkov
Browse files
Options

MDEV-17474 Change Unicode collation implementation from "handler" to "inline" style

parent fee24b12
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 1323 additions and 1268 deletions
include/m_ctype.h
View file @ 6eae037c
......@@ -362,7 +362,6 @@ extern MY_COLLATION_HANDLER my_collation_8bit_bin_handler;
extern MY_COLLATION_HANDLER my_collation_8bit_simple_ci_handler;
extern MY_COLLATION_HANDLER my_collation_8bit_nopad_bin_handler;
extern MY_COLLATION_HANDLER my_collation_8bit_simple_nopad_ci_handler;
extern MY_COLLATION_HANDLER my_collation_ucs2_uca_handler;
/* Some typedef to make it easy for C++ to make function pointers */
typedef int (*my_charset_conv_mb_wc)(CHARSET_INFO *, my_wc_t *,
......
strings/ctype-uca.c
View file @ 6eae037c
This diff is collapsed.
strings/ctype-uca.ic 0 → 100644
View file @ 6eae037c
This diff is collapsed.
strings/ctype-ucs2.c
View file @ 6eae037c
......@@ -1184,35 +1184,7 @@ my_lengthsp_mb2(CHARSET_INFO *cs __attribute__((unused)),
but the JSON functions needed my_utf16_uni()
so the #ifdef was moved lower.
*/
/*
D800..DB7F - Non-provate surrogate high (896 pages)
DB80..DBFF - Private surrogate high (128 pages)
DC00..DFFF - Surrogate low (1024 codes in a page)
*/
#define MY_UTF16_SURROGATE_HIGH_FIRST 0xD800
#define MY_UTF16_SURROGATE_HIGH_LAST 0xDBFF
#define MY_UTF16_SURROGATE_LOW_FIRST 0xDC00
#define MY_UTF16_SURROGATE_LOW_LAST 0xDFFF
#define MY_UTF16_HIGH_HEAD(x) ((((uchar) (x)) & 0xFC) == 0xD8)
#define MY_UTF16_LOW_HEAD(x) ((((uchar) (x)) & 0xFC) == 0xDC)
/* Test if a byte is a leading byte of a high or low surrogate head: */
#define MY_UTF16_SURROGATE_HEAD(x) ((((uchar) (x)) & 0xF8) == 0xD8)
/* Test if a Unicode code point is a high or low surrogate head */
#define MY_UTF16_SURROGATE(x) (((x) & 0xF800) == 0xD800)
#define MY_UTF16_WC2(a, b) ((a << 8) + b)
/*
a= 110110?? (<< 18)
b= ???????? (<< 10)
c= 110111?? (<< 8)
d= ???????? (<< 0)
*/
#define MY_UTF16_WC4(a, b, c, d) (((a & 3) << 18) + (b << 10) + \
((c & 3) << 8) + d + 0x10000)
#include "ctype-utf16.h"
#define IS_MB2_CHAR(b0,b1) (!MY_UTF16_SURROGATE_HEAD(b0))
#define IS_MB4_CHAR(b0,b1,b2,b3) (MY_UTF16_HIGH_HEAD(b0) && MY_UTF16_LOW_HEAD(b2))
......@@ -1261,32 +1233,7 @@ static inline int my_weight_mb2_utf16mb2_general_ci(uchar b0, uchar b1)
my_utf16_uni(CHARSET_INFO *cs __attribute__((unused)),
my_wc_t *pwc, const uchar *s, const uchar *e)
{
if (s + 2 > e)
return MY_CS_TOOSMALL2;
/*
High bytes: 0xD[89AB] = B'110110??'
Low bytes: 0xD[CDEF] = B'110111??'
Surrogate mask: 0xFC = B'11111100'
*/
if (MY_UTF16_HIGH_HEAD(*s)) /* Surrogate head */
{
if (s + 4 > e)
return MY_CS_TOOSMALL4;
if (!MY_UTF16_LOW_HEAD(s[2])) /* Broken surrigate pair */
return MY_CS_ILSEQ;
*pwc= MY_UTF16_WC4(s[0], s[1], s[2], s[3]);
return 4;
}
if (MY_UTF16_LOW_HEAD(*s)) /* Low surrogate part without high part */
return MY_CS_ILSEQ;
*pwc= MY_UTF16_WC2(s[0], s[1]);
return 2;
return my_mb_wc_utf16_quick(pwc, s, e);
}
......@@ -2109,6 +2056,8 @@ struct charset_info_st my_charset_utf16le_nopad_bin=
#ifdef HAVE_CHARSET_utf32
#include "ctype-utf32.h"
/*
Check is b0 and b1 start a valid UTF32 four-byte sequence.
Don't accept characters greater than U+10FFFF.
......@@ -2117,8 +2066,6 @@ struct charset_info_st my_charset_utf16le_nopad_bin=
#define IS_MB4_CHAR(b0,b1,b2,b3) (IS_UTF32_MBHEAD4(b0,b1))
#define MY_UTF32_WC4(b0,b1,b2,b3) ((((my_wc_t)b0) << 24) + (b1 << 16) + \
(b2 << 8) + (b3))
static inline int my_weight_utf32_general_ci(uchar b0, uchar b1,
uchar b2, uchar b3)
......@@ -2161,10 +2108,7 @@ static int
my_utf32_uni(CHARSET_INFO *cs __attribute__((unused)),
my_wc_t *pwc, const uchar *s, const uchar *e)
{
if (s + 4 > e)
return MY_CS_TOOSMALL4;
*pwc= MY_UTF32_WC4(s[0], s[1], s[2], s[3]);
return *pwc > 0x10FFFF ? MY_CS_ILSEQ : 4;
return my_mb_wc_utf32_quick(pwc, s, e);
}
......@@ -2928,6 +2872,8 @@ struct charset_info_st my_charset_utf32_nopad_bin=
#ifdef HAVE_CHARSET_ucs2
#include "ctype-ucs2.h"
static const uchar ctype_ucs2[] = {
0,
32, 32, 32, 32, 32, 32, 32, 32, 32, 40, 40, 40, 40, 40, 32, 32,
......@@ -3037,11 +2983,7 @@ my_charlen_ucs2(CHARSET_INFO *cs __attribute__((unused)),
static int my_ucs2_uni(CHARSET_INFO *cs __attribute__((unused)),
my_wc_t * pwc, const uchar *s, const uchar *e)
{
if (s+2 > e) /* Need 2 characters */
return MY_CS_TOOSMALL2;
*pwc= ((uchar)s[0]) * 256 + ((uchar)s[1]);
return 2;
return my_mb_wc_ucs2_quick(pwc, s, e);
}
static int my_uni_ucs2(CHARSET_INFO *cs __attribute__((unused)) ,
......
strings/ctype-ucs2.h 0 → 100644
View file @ 6eae037c
/*
Copyright (c) 2018 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
*/
#ifndef _CTYPE_UCS2_H
#define _CTYPE_UCS2_H
static inline int
my_mb_wc_ucs2_quick(my_wc_t * pwc, const uchar *s, const uchar *e)
{
if (s+2 > e) /* Need 2 characters */
return MY_CS_TOOSMALL2;
*pwc= ((uchar)s[0]) * 256 + ((uchar)s[1]);
return 2;
}
#endif /* _CTYPE_UCS2_H */
strings/ctype-utf16.h 0 → 100644
View file @ 6eae037c
/*
Copyright (c) 2018 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
*/
#ifndef _CTYPE_UTF16_H
#define _CTYPE_UTF16_H
/*
D800..DB7F - Non-provate surrogate high (896 pages)
DB80..DBFF - Private surrogate high (128 pages)
DC00..DFFF - Surrogate low (1024 codes in a page)
*/
#define MY_UTF16_SURROGATE_HIGH_FIRST 0xD800
#define MY_UTF16_SURROGATE_HIGH_LAST 0xDBFF
#define MY_UTF16_SURROGATE_LOW_FIRST 0xDC00
#define MY_UTF16_SURROGATE_LOW_LAST 0xDFFF
#define MY_UTF16_HIGH_HEAD(x) ((((uchar) (x)) & 0xFC) == 0xD8)
#define MY_UTF16_LOW_HEAD(x) ((((uchar) (x)) & 0xFC) == 0xDC)
/* Test if a byte is a leading byte of a high or low surrogate head: */
#define MY_UTF16_SURROGATE_HEAD(x) ((((uchar) (x)) & 0xF8) == 0xD8)
/* Test if a Unicode code point is a high or low surrogate head */
#define MY_UTF16_SURROGATE(x) (((x) & 0xF800) == 0xD800)
#define MY_UTF16_WC2(a, b) ((a << 8) + b)
/*
a= 110110?? (<< 18)
b= ???????? (<< 10)
c= 110111?? (<< 8)
d= ???????? (<< 0)
*/
#define MY_UTF16_WC4(a, b, c, d) (((a & 3) << 18) + (b << 10) + \
((c & 3) << 8) + d + 0x10000)
static inline int
my_mb_wc_utf16_quick(my_wc_t *pwc, const uchar *s, const uchar *e)
{
if (s + 2 > e)
return MY_CS_TOOSMALL2;
/*
High bytes: 0xD[89AB] = B'110110??'
Low bytes: 0xD[CDEF] = B'110111??'
Surrogate mask: 0xFC = B'11111100'
*/
if (MY_UTF16_HIGH_HEAD(*s)) /* Surrogate head */
{
if (s + 4 > e)
return MY_CS_TOOSMALL4;
if (!MY_UTF16_LOW_HEAD(s[2])) /* Broken surrigate pair */
return MY_CS_ILSEQ;
*pwc= MY_UTF16_WC4(s[0], s[1], s[2], s[3]);
return 4;
}
if (MY_UTF16_LOW_HEAD(*s)) /* Low surrogate part without high part */
return MY_CS_ILSEQ;
*pwc= MY_UTF16_WC2(s[0], s[1]);
return 2;
}
#endif /* _CTYPE_UTF16_H */
strings/ctype-utf32.h 0 → 100644
View file @ 6eae037c
/*
Copyright (c) 2018 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
*/
#ifndef _CTYPE_UTF32_H
#define _CTYPE_UTF32_H
#define MY_UTF32_WC4(b0,b1,b2,b3) ((((my_wc_t)b0) << 24) + (b1 << 16) + \
(b2 << 8) + (b3))
static inline int
my_mb_wc_utf32_quick(my_wc_t *pwc, const uchar *s, const uchar *e)
{
if (s + 4 > e)
return MY_CS_TOOSMALL4;
*pwc= MY_UTF32_WC4(s[0], s[1], s[2], s[3]);
return *pwc > 0x10FFFF ? MY_CS_ILSEQ : 4;
}
#endif /* _CTYPE_UTF32_H */
strings/ctype-utf8.c
View file @ 6eae037c
......@@ -26,78 +26,9 @@
#define EILSEQ ENOENT
#endif
/* Detect special bytes and sequences */
#define IS_CONTINUATION_BYTE(c) (((uchar) (c) ^ 0x80) < 0x40)
/*
Check MB2 character assuming that b0 is alredy known to be >= 0xC2.
Use this macro if the caller already checked b0 for:
- an MB1 character
- an unused gap between MB1 and MB2HEAD
*/
#define IS_UTF8MB2_STEP2(b0,b1) (((uchar) (b0) < 0xE0) && \
IS_CONTINUATION_BYTE((uchar) b1))
#include "ctype-utf8.h"
/*
Check MB3 character assuming that b0 is already known to be
in the valid MB3HEAD range [0xE0..0xEF].
*/
#define IS_UTF8MB3_STEP2(b0,b1,b2) (IS_CONTINUATION_BYTE(b1) && \
IS_CONTINUATION_BYTE(b2) && \
((uchar) b0 >= 0xe1 || (uchar) b1 >= 0xa0))
/*
Check MB3 character assuming that b0 is already known to be >= 0xE0,
but is not checked for the high end 0xF0 yet.
Use this macro if the caller already checked b0 for:
- an MB1 character
- an unused gap between MB1 and MB2HEAD
- an MB2HEAD
*/
#define IS_UTF8MB3_STEP3(b0,b1,b2) (((uchar) (b0) < 0xF0) && \
IS_UTF8MB3_STEP2(b0,b1,b2))
/*
UTF-8 quick four-byte mask:
11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
Encoding allows to encode U+00010000..U+001FFFFF
The maximum character defined in the Unicode standard is U+0010FFFF.
Higher characters U+00110000..U+001FFFFF are not used.
11110000.10010000.10xxxxxx.10xxxxxx == F0.90.80.80 == U+00010000 (min)
11110100.10001111.10111111.10111111 == F4.8F.BF.BF == U+0010FFFF (max)
Valid codes:
[F0][90..BF][80..BF][80..BF]
[F1][80..BF][80..BF][80..BF]
[F2][80..BF][80..BF][80..BF]
[F3][80..BF][80..BF][80..BF]
[F4][80..8F][80..BF][80..BF]
*/
/*
Check MB4 character assuming that b0 is already
known to be in the range [0xF0..0xF4]
*/
#define IS_UTF8MB4_STEP2(b0,b1,b2,b3) (IS_CONTINUATION_BYTE(b1) && \
IS_CONTINUATION_BYTE(b2) && \
IS_CONTINUATION_BYTE(b3) && \
(b0 >= 0xf1 || b1 >= 0x90) && \
(b0 <= 0xf3 || b1 <= 0x8F))
#define IS_UTF8MB4_STEP3(b0,b1,b2,b3) (((uchar) (b0) < 0xF5) && \
IS_UTF8MB4_STEP2(b0,b1,b2,b3))
/* Convert individual bytes to Unicode code points */
#define UTF8MB2_CODE(b0,b1) (((my_wc_t) ((uchar) b0 & 0x1f) << 6) |\
((my_wc_t) ((uchar) b1 ^ 0x80)))
#define UTF8MB3_CODE(b0,b1,b2) (((my_wc_t) ((uchar) b0 & 0x0f) << 12) |\
((my_wc_t) ((uchar) b1 ^ 0x80) << 6) |\
((my_wc_t) ((uchar) b2 ^ 0x80)))
#define UTF8MB4_CODE(b0,b1,b2,b3) (((my_wc_t) ((uchar) b0 & 0x07) << 18) |\
((my_wc_t) ((uchar) b1 ^ 0x80) << 12) |\
((my_wc_t) ((uchar) b2 ^ 0x80) << 6) |\
(my_wc_t) ((uchar) b3 ^ 0x80))
/* Definitions for strcoll.ic */
#define IS_MB1_CHAR(x) ((uchar) (x) < 0x80)
......@@ -4981,42 +4912,7 @@ static const uchar to_upper_utf8[] = {
static int my_utf8_uni(CHARSET_INFO *cs __attribute__((unused)),
my_wc_t * pwc, const uchar *s, const uchar *e)
{
uchar c;
if (s >= e)
return MY_CS_TOOSMALL;
c= s[0];
if (c < 0x80)
{
*pwc = c;
return 1;
}
else if (c < 0xc2)
return MY_CS_ILSEQ;
else if (c < 0xe0)
{
if (s+2 > e) /* We need 2 characters */
return MY_CS_TOOSMALL2;
if (!(IS_CONTINUATION_BYTE(s[1])))
return MY_CS_ILSEQ;
*pwc= UTF8MB2_CODE(c, s[1]);
return 2;
}
else if (c < 0xf0)
{
if (s+3 > e) /* We need 3 characters */
return MY_CS_TOOSMALL3;
if (!IS_UTF8MB3_STEP2(c, s[1], s[2]))
return MY_CS_ILSEQ;
*pwc= UTF8MB3_CODE(c, s[1], s[2]);
return 3;
}
return MY_CS_ILSEQ;
return my_mb_wc_utf8mb3_quick(pwc, s, e);
}
......@@ -7379,52 +7275,7 @@ static int
my_mb_wc_utf8mb4(CHARSET_INFO *cs __attribute__((unused)),
my_wc_t * pwc, const uchar *s, const uchar *e)
{
uchar c;
if (s >= e)
return MY_CS_TOOSMALL;
c= s[0];
if (c < 0x80)
{
*pwc= c;
return 1;
}
else if (c < 0xc2)
return MY_CS_ILSEQ;
else if (c < 0xe0)
{
if (s + 2 > e) /* We need 2 characters */
return MY_CS_TOOSMALL2;
if (!(IS_CONTINUATION_BYTE(s[1])))
return MY_CS_ILSEQ;
*pwc= UTF8MB2_CODE(c, s[1]);
return 2;
}
else if (c < 0xf0)
{
if (s + 3 > e) /* We need 3 characters */
return MY_CS_TOOSMALL3;
if (!IS_UTF8MB3_STEP2(c, s[1], s[2]))
return MY_CS_ILSEQ;
*pwc= UTF8MB3_CODE(c, s[1], s[2]);
return 3;
}
else if (c < 0xf5)
{
if (s + 4 > e) /* We need 4 characters */
return MY_CS_TOOSMALL4;
if (!IS_UTF8MB4_STEP2(c, s[1], s[2], s[3]))
return MY_CS_ILSEQ;
*pwc= UTF8MB4_CODE(c, s[1], s[2], s[3]);
return 4;
}
return MY_CS_ILSEQ;
return my_mb_wc_utf8mb4_quick(pwc, s, e);
}
......
strings/ctype-utf8.h 0 → 100644
View file @ 6eae037c
/*
Copyright (c) 2018 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
*/
#ifndef _CTYPE_UTF8_H
#define _CTYPE_UTF8_H
/* Detect special bytes and sequences */
#define IS_CONTINUATION_BYTE(c) (((uchar) (c) ^ 0x80) < 0x40)
/*
Check MB2 character assuming that b0 is alredy known to be >= 0xC2.
Use this macro if the caller already checked b0 for:
- an MB1 character
- an unused gap between MB1 and MB2HEAD
*/
#define IS_UTF8MB2_STEP2(b0,b1) (((uchar) (b0) < 0xE0) && \
IS_CONTINUATION_BYTE((uchar) b1))
/*
Check MB3 character assuming that b0 is already known to be
in the valid MB3HEAD range [0xE0..0xEF].
*/
#define IS_UTF8MB3_STEP2(b0,b1,b2) (IS_CONTINUATION_BYTE(b1) && \
IS_CONTINUATION_BYTE(b2) && \
((uchar) b0 >= 0xe1 || (uchar) b1 >= 0xa0))
/*
Check MB3 character assuming that b0 is already known to be >= 0xE0,
but is not checked for the high end 0xF0 yet.
Use this macro if the caller already checked b0 for:
- an MB1 character
- an unused gap between MB1 and MB2HEAD
- an MB2HEAD
*/
#define IS_UTF8MB3_STEP3(b0,b1,b2) (((uchar) (b0) < 0xF0) && \
IS_UTF8MB3_STEP2(b0,b1,b2))
/*
UTF-8 quick four-byte mask:
11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
Encoding allows to encode U+00010000..U+001FFFFF
The maximum character defined in the Unicode standard is U+0010FFFF.
Higher characters U+00110000..U+001FFFFF are not used.
11110000.10010000.10xxxxxx.10xxxxxx == F0.90.80.80 == U+00010000 (min)
11110100.10001111.10111111.10111111 == F4.8F.BF.BF == U+0010FFFF (max)
Valid codes:
[F0][90..BF][80..BF][80..BF]
[F1][80..BF][80..BF][80..BF]
[F2][80..BF][80..BF][80..BF]
[F3][80..BF][80..BF][80..BF]
[F4][80..8F][80..BF][80..BF]
*/
/*
Check MB4 character assuming that b0 is already
known to be in the range [0xF0..0xF4]
*/
#define IS_UTF8MB4_STEP2(b0,b1,b2,b3) (IS_CONTINUATION_BYTE(b1) && \
IS_CONTINUATION_BYTE(b2) && \
IS_CONTINUATION_BYTE(b3) && \
(b0 >= 0xf1 || b1 >= 0x90) && \
(b0 <= 0xf3 || b1 <= 0x8F))
#define IS_UTF8MB4_STEP3(b0,b1,b2,b3) (((uchar) (b0) < 0xF5) && \
IS_UTF8MB4_STEP2(b0,b1,b2,b3))
/* Convert individual bytes to Unicode code points */
#define UTF8MB2_CODE(b0,b1) (((my_wc_t) ((uchar) b0 & 0x1f) << 6) |\
((my_wc_t) ((uchar) b1 ^ 0x80)))
#define UTF8MB3_CODE(b0,b1,b2) (((my_wc_t) ((uchar) b0 & 0x0f) << 12) |\
((my_wc_t) ((uchar) b1 ^ 0x80) << 6) |\
((my_wc_t) ((uchar) b2 ^ 0x80)))
#define UTF8MB4_CODE(b0,b1,b2,b3) (((my_wc_t) ((uchar) b0 & 0x07) << 18) |\
((my_wc_t) ((uchar) b1 ^ 0x80) << 12) |\
((my_wc_t) ((uchar) b2 ^ 0x80) << 6) |\
(my_wc_t) ((uchar) b3 ^ 0x80))
static inline int
my_mb_wc_utf8mb3_quick(my_wc_t * pwc, const uchar *s, const uchar *e)
{
uchar c;
if (s >= e)
return MY_CS_TOOSMALL;
c= s[0];
if (c < 0x80)
{
*pwc = c;
return 1;
}
else if (c < 0xc2)
return MY_CS_ILSEQ;
else if (c < 0xe0)
{
if (s+2 > e) /* We need 2 characters */
return MY_CS_TOOSMALL2;
if (!(IS_CONTINUATION_BYTE(s[1])))
return MY_CS_ILSEQ;
*pwc= UTF8MB2_CODE(c, s[1]);
return 2;
}
else if (c < 0xf0)
{
if (s+3 > e) /* We need 3 characters */
return MY_CS_TOOSMALL3;
if (!IS_UTF8MB3_STEP2(c, s[1], s[2]))
return MY_CS_ILSEQ;
*pwc= UTF8MB3_CODE(c, s[1], s[2]);
return 3;
}
return MY_CS_ILSEQ;
}
#ifdef HAVE_CHARSET_utf8mb4
static inline int
my_mb_wc_utf8mb4_quick(my_wc_t *pwc, const uchar *s, const uchar *e)
{
uchar c;
if (s >= e)
return MY_CS_TOOSMALL;
c= s[0];
if (c < 0x80)
{
*pwc= c;
return 1;
}
else if (c < 0xc2)
return MY_CS_ILSEQ;
else if (c < 0xe0)
{
if (s + 2 > e) /* We need 2 characters */
return MY_CS_TOOSMALL2;
if (!(IS_CONTINUATION_BYTE(s[1])))
return MY_CS_ILSEQ;
*pwc= UTF8MB2_CODE(c, s[1]);
return 2;
}
else if (c < 0xf0)
{
if (s + 3 > e) /* We need 3 characters */
return MY_CS_TOOSMALL3;
if (!IS_UTF8MB3_STEP2(c, s[1], s[2]))
return MY_CS_ILSEQ;
*pwc= UTF8MB3_CODE(c, s[1], s[2]);
return 3;
}
else if (c < 0xf5)
{
if (s + 4 > e) /* We need 4 characters */
return MY_CS_TOOSMALL4;
if (!IS_UTF8MB4_STEP2(c, s[1], s[2], s[3]))
return MY_CS_ILSEQ;
*pwc= UTF8MB4_CODE(c, s[1], s[2], s[3]);
return 4;
}
return MY_CS_ILSEQ;
}
#endif /* HAVE_CHARSET_utf8mb4*/
#endif /* _CTYPE_UTF8_H */
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7