database/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2026-01-06 05:22:24 +03:00
Code Activity

bug#7284: strnxfrm returns different results for equal strings

Browse Source
This commit is contained in:
bar@mysql.com
2005-01-13 18:12:04 +04:00
parent 9cea2457ec
commit e3d6e6736b
34 changed files with 344 additions and 45 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
mysql-test/include/ctype_filesort.inc Normal file
View File

@@ -0,0 +1,15 @@
#
# Set desired charset_connection and collation_collation
# before including this file.
#
# The next query creates a LONGTEXT column
# using the current character_set_connection
# and collation_connection.
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
drop table t1;

21
mysql-test/r/ctype_big5.result
View File

@@ -56,3 +56,24 @@ DROP DATABASE d1;
USE test;
SET character_set_server= @safe_character_set_server;
SET collation_server= @safe_collation_server;
SET NAMES big5;
SET collation_connection='big5_chinese_ci';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
big5_chinese_ci 6109
big5_chinese_ci 61
big5_chinese_ci 6120
drop table t1;
SET collation_connection='big5_bin';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
big5_bin 6109
big5_bin 61
big5_bin 6120
drop table t1;

20
mysql-test/r/ctype_latin1.result
View File

@@ -305,3 +305,23 @@ select 'a' regexp 'A' collate latin1_general_cs;
select 'a' regexp 'A' collate latin1_bin;
'a' regexp 'A' collate latin1_bin
0
SET collation_connection='latin1_swedish_ci';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
latin1_swedish_ci 6109
latin1_swedish_ci 61
latin1_swedish_ci 6120
drop table t1;
SET collation_connection='latin1_bin';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
latin1_bin 6109
latin1_bin 61
latin1_bin 6120
drop table t1;

9
mysql-test/r/ctype_latin1_de.result
View File

@@ -317,3 +317,12 @@ FIELD('ue',s1) FIELD('
1 1 1 1
1 1 1 1
DROP TABLE t1;
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
latin1_german2_ci 6109
latin1_german2_ci 61
latin1_german2_ci 6120
drop table t1;

20
mysql-test/r/ctype_sjis.result
View File

@@ -71,3 +71,23 @@ B1
B2
B3
drop table t1;
SET collation_connection='sjis_japanese_ci';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
sjis_japanese_ci 6109
sjis_japanese_ci 61
sjis_japanese_ci 6120
drop table t1;
SET collation_connection='sjis_bin';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
sjis_bin 6109
sjis_bin 61
sjis_bin 6120
drop table t1;

20
mysql-test/r/ctype_tis620.result
View File

@@ -2937,3 +2937,23 @@ Screensaver 2 2002-01-22 491 0 519 0 0
3 http://www.siamzone.com/download/download/000003-jasonx2(800x600).jpg Jaso
n X Wallpapers 1 2002-05-31 579 0 1091 0 0
DROP TABLE t1;
SET collation_connection='tis620_thai_ci';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
tis620_thai_ci 6109
tis620_thai_ci 61
tis620_thai_ci 6120
drop table t1;
SET collation_connection='tis620_bin';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
tis620_bin 6109
tis620_bin 61
tis620_bin 6120
drop table t1;

10
mysql-test/r/ctype_uca.result
View File

@@ -2386,3 +2386,13 @@ a 1
b 0
c 0
drop table t1;
SET collation_connection='utf8_unicode_ci';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
utf8_unicode_ci 6109
utf8_unicode_ci 61
utf8_unicode_ci 6120
drop table t1;

21
mysql-test/r/ctype_ucs.result
View File

@@ -592,3 +592,24 @@ a NULL
b NULL
c NULL
drop table t1;
SET collation_connection='ucs2_general_ci';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
ucs2_general_ci 00610009
ucs2_general_ci 0061
ucs2_general_ci 00610020
drop table t1;
SET NAMES latin1;
SET collation_connection='ucs2_bin';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
ucs2_bin 00610009
ucs2_bin 0061
ucs2_bin 00610020
drop table t1;

20
mysql-test/r/ctype_ujis.result
View File

@@ -2207,3 +2207,23 @@ F4FC
F4FD
F4FE
DROP TABLE t1;
SET collation_connection='ujis_japanese_ci';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
ujis_japanese_ci 6109
ujis_japanese_ci 61
ujis_japanese_ci 6120
drop table t1;
SET collation_connection='ujis_bin';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
ujis_bin 6109
ujis_bin 61
ujis_bin 6120
drop table t1;

20
mysql-test/r/ctype_utf8.result
View File

@@ -829,3 +829,23 @@ select * from t1 where soundex(a) = soundex('test');
id a
1 Test
drop table t1;
SET collation_connection='utf8_general_ci';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
utf8_general_ci 6109
utf8_general_ci 61
utf8_general_ci 6120
drop table t1;
SET collation_connection='utf8_bin';
create table t1 select repeat('a',4000) a;
delete from t1;
insert into t1 values ('a'), ('a '), ('a\t');
select collation(a),hex(a) from t1 order by a;
collation(a) hex(a)
utf8_bin 6109
utf8_bin 61
utf8_bin 6120
drop table t1;

6
mysql-test/t/ctype_big5.test
View File

@@ -10,3 +10,9 @@ drop table if exists t1;
SET @test_character_set= 'big5';
SET @test_collation= 'big5_chinese_ci';
-- source include/ctype_common.inc
SET NAMES big5;
SET collation_connection='big5_chinese_ci';
-- source include/ctype_filesort.inc
SET collation_connection='big5_bin';
-- source include/ctype_filesort.inc

6
mysql-test/t/ctype_latin1.test
View File

@@ -60,3 +60,9 @@ DROP TABLE t1;
select 'a' regexp 'A' collate latin1_general_ci;
select 'a' regexp 'A' collate latin1_general_cs;
select 'a' regexp 'A' collate latin1_bin;
SET collation_connection='latin1_swedish_ci';
-- source include/ctype_filesort.inc
SET collation_connection='latin1_bin';
-- source include/ctype_filesort.inc

2
mysql-test/t/ctype_latin1_de.test
View File

@@ -114,3 +114,5 @@ SELECT s1,COUNT(*) FROM t1 GROUP BY s1;
SELECT COUNT(DISTINCT s1) FROM t1;
SELECT FIELD('ue',s1), FIELD('<27>',s1), s1='ue', s1='<27>' FROM t1;
DROP TABLE t1;
-- source include/ctype_filesort.inc

6
mysql-test/t/ctype_sjis.test
View File

@@ -62,3 +62,9 @@ CREATE TABLE t1 (
insert into t1 values(0xb1),(0xb2),(0xb3);
select hex(c) from t1;
drop table t1;
SET collation_connection='sjis_japanese_ci';
-- source include/ctype_filesort.inc
SET collation_connection='sjis_bin';
-- source include/ctype_filesort.inc

6
mysql-test/t/ctype_tis620.test
View File

@@ -151,3 +151,9 @@ INSERT INTO t1 VALUES
n X Wallpapers',1,'','2002-05-31','',579,0,'',1091,0,0,'');
select * from t1 order by id;
DROP TABLE t1;
SET collation_connection='tis620_thai_ci';
-- source include/ctype_filesort.inc
SET collation_connection='tis620_bin';
-- source include/ctype_filesort.inc

3
mysql-test/t/ctype_uca.test
View File

@@ -452,3 +452,6 @@ create table t1 (a varchar(1)) character set utf8 collate utf8_estonian_ci;
insert into t1 values ('A'),('B'),('C'),('a'),('b'),('c');
select a, a regexp '[a]' from t1 order by binary a;
drop table t1;
SET collation_connection='utf8_unicode_ci';
-- source include/ctype_filesort.inc

6
mysql-test/t/ctype_ucs.test
View File

@@ -384,3 +384,9 @@ alter table t1 add b char(1);
show warnings;
select * from t1 order by a;
drop table t1;
SET collation_connection='ucs2_general_ci';
-- source include/ctype_filesort.inc
SET NAMES latin1;
SET collation_connection='ucs2_bin';
-- source include/ctype_filesort.inc

6
mysql-test/t/ctype_ujis.test
View File

@@ -1141,3 +1141,9 @@ INSERT INTO t1 VALUES(0xF4FD);
INSERT INTO t1 VALUES(0xF4FE);
SELECT HEX(c) FROM t1 ORDER BY BINARY c;
DROP TABLE t1;
SET collation_connection='ujis_japanese_ci';
-- source include/ctype_filesort.inc
SET collation_connection='ujis_bin';
-- source include/ctype_filesort.inc

6
mysql-test/t/ctype_utf8.test
View File

@@ -675,3 +675,9 @@ select * from t1 where soundex(a) = soundex('Test');
select * from t1 where soundex(a) = soundex('TEST');
select * from t1 where soundex(a) = soundex('test');
drop table t1;
SET collation_connection='utf8_general_ci';
-- source include/ctype_filesort.inc
SET collation_connection='utf8_bin';
-- source include/ctype_filesort.inc

12
sql/field.cc
View File

@@ -4393,8 +4393,7 @@ void Field_string::sort_string(char *to,uint length)
uint tmp=my_strnxfrm(field_charset,
(unsigned char *) to, length,
(unsigned char *) ptr, field_length);
if (tmp < length)
field_charset->cset->fill(field_charset, to + tmp, length - tmp, ' ');
DBUG_ASSERT(tmp == length);
}
@@ -4596,9 +4595,7 @@ void Field_varstring::sort_string(char *to,uint length)
(uchar*) to, length,
(uchar*) ptr+HA_KEY_BLOB_LENGTH,
tot_length);
if (tot_length < length)
field_charset->cset->fill(field_charset, to+tot_length,length-tot_length,
binary() ? (char) 0 : ' ');
DBUG_ASSERT(tot_length == length);
}
@@ -5116,10 +5113,7 @@ void Field_blob::sort_string(char *to,uint length)
blob_length=my_strnxfrm(field_charset,
(uchar*) to, length,
(uchar*) blob, blob_length);
if (blob_length < length)
field_charset->cset->fill(field_charset, to+blob_length,
length-blob_length,
binary() ? (char) 0 : ' ');
DBUG_ASSERT(blob_length == length);
}
}

5
sql/filesort.cc
View File

@@ -622,10 +622,7 @@ static void make_sortkey(register SORTPARAM *param,
}
uint tmp_length=my_strnxfrm(cs,to,sort_field->length,
(unsigned char *) from, length);
if (tmp_length < sort_field->length)
cs->cset->fill(cs, (char*) to+tmp_length,
sort_field->length-tmp_length,
fill_char);
DBUG_ASSERT(tmp_length == sort_field->length);
}
else
{

5
strings/ctype-big5.c
View File

@@ -298,6 +298,7 @@ static int my_strnxfrm_big5(CHARSET_INFO *cs __attribute__((unused)),
const uchar * src, uint srclen)
{
uint16 e;
uint dstlen= len;
len = srclen;
while (len--)
@@ -312,7 +313,9 @@ static int my_strnxfrm_big5(CHARSET_INFO *cs __attribute__((unused)),
} else
*dest++ = sort_order_big5[(uchar) *src++];
}
return srclen;
if (dstlen > srclen)
bfill(dest, dstlen - srclen, ' ');
return dstlen;
}
#if 0

26
strings/ctype-bin.c
View File

@@ -341,13 +341,27 @@ static int my_wildcmp_bin(CHARSET_INFO *cs,
static int my_strnxfrm_bin(CHARSET_INFO *cs __attribute__((unused)),
uchar * dest, uint len,
const uchar *src,
uint srclen __attribute__((unused)))
uchar * dest, uint dstlen,
const uchar *src, uint srclen)
{
if (dest != src)
memcpy(dest,src,len= min(len,srclen));
return len;
memcpy(dest, src, min(dstlen,srclen));
if (dstlen > srclen)
bfill(dest + srclen, dstlen - srclen, 0);
return dstlen;
}
static
int my_strnxfrm_8bit_bin(CHARSET_INFO *cs __attribute__((unused)),
uchar * dest, uint dstlen,
const uchar *src, uint srclen)
{
if (dest != src)
memcpy(dest, src, min(dstlen,srclen));
if (dstlen > srclen)
bfill(dest + srclen, dstlen - srclen, ' ');
return dstlen;
}
@@ -417,7 +431,7 @@ MY_COLLATION_HANDLER my_collation_8bit_bin_handler =
NULL, /* init */
my_strnncoll_8bit_bin,
my_strnncollsp_8bit_bin,
my_strnxfrm_bin,
my_strnxfrm_8bit_bin,
my_like_range_simple,
my_wildcmp_bin,
my_strcasecmp_bin,

8
strings/ctype-czech.c
View File

@@ -296,16 +296,18 @@ static int my_strnxfrm_czech(CHARSET_INFO *cs __attribute__((unused)),
int value;
const uchar * p, * store;
int pass = 0;
int totlen = 0;
uint totlen = 0;
p = src; store = src;
do
{
NEXT_CMP_VALUE(src, p, store, pass, value, (int)srclen);
ADD_TO_RESULT(dest, (int)len, totlen, value);
ADD_TO_RESULT(dest, len, totlen, value);
}
while (value);
return totlen;
if (len > totlen)
bfill(dest + totlen, len - totlen, ' ');
return len;
}
#undef IS_END

5
strings/ctype-gbk.c
View File

@@ -2659,6 +2659,7 @@ static int my_strnxfrm_gbk(CHARSET_INFO *cs __attribute__((unused)),
const uchar * src, uint srclen)
{
uint16 e;
uint dstlen= len;
len = srclen;
while (len--)
@@ -2673,7 +2674,9 @@ static int my_strnxfrm_gbk(CHARSET_INFO *cs __attribute__((unused)),
} else
*dest++ = sort_order_gbk[(uchar) *src++];
}
return srclen;
if (dstlen > srclen)
bfill(dest, dstlen - srclen, ' ');
return dstlen;
}

5
strings/ctype-latin1.c
View File

@@ -637,7 +637,6 @@ static int my_strnxfrm_latin1_de(CHARSET_INFO *cs __attribute__((unused)),
uchar * dest, uint len,
const uchar * src, uint srclen)
{
const uchar *dest_orig = dest;
const uchar *de = dest + len;
const uchar *se = src + srclen;
for ( ; src < se && dest < de ; src++)
@@ -647,7 +646,9 @@ static int my_strnxfrm_latin1_de(CHARSET_INFO *cs __attribute__((unused)),
if ((chr=combo2map[*src]) && dest < de)
*dest++=chr;
}
return (int) (dest - dest_orig);
if (dest < de)
bfill(dest, de - dest, ' ');
return (int) len;
}

11
strings/ctype-mb.c
View File

@@ -412,13 +412,14 @@ static int my_strnncollsp_mb_bin(CHARSET_INFO * cs __attribute__((unused)),
static int my_strnxfrm_mb_bin(CHARSET_INFO *cs __attribute__((unused)),
uchar * dest, uint len,
const uchar *src,
uint srclen __attribute__((unused)))
uchar * dest, uint dstlen,
const uchar *src, uint srclen)
{
if (dest != src)
memcpy(dest,src,len= min(len,srclen));
return len;
memcpy(dest, src, min(dstlen, srclen));
if (dstlen > srclen)
bfill(dest + srclen, dstlen - srclen, ' ');
return dstlen;
}

45
strings/ctype-simple.c
View File

@@ -21,27 +21,68 @@
#include "stdarg.h"
/*
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'.
*/
int my_strnxfrm_simple(CHARSET_INFO * cs,
uchar *dest, uint len,
const uchar *src, uint srclen)
{
uchar *map= cs->sort_order;
uint dstlen= len;
set_if_smaller(len, srclen);
if (dest != src)
{
const uchar *end;
for ( end=src+len; src < end ; )
*dest++= map[*src++];
return len;
}
else
{
const uchar *end;
for ( end=dest+len; dest < end ; dest++)
*dest= (char) map[(uchar) *dest];
return len;
}
if (dstlen > len)
bfill(dest, dstlen - len, ' ');
return dstlen;
}
int my_strnncoll_simple(CHARSET_INFO * cs, const uchar *s, uint slen,

4
strings/ctype-sjis.c
View File

@@ -291,7 +291,9 @@ static int my_strnxfrm_sjis(CHARSET_INFO *cs __attribute__((unused)),
else
*dest++ = sort_order_sjis[(uchar)*src++];
}
return srclen;
if (len > srclen)
bfill(dest, len - srclen, ' ');
return len;
}

6
strings/ctype-tis620.c
View File

@@ -631,9 +631,13 @@ int my_strnxfrm_tis620(CHARSET_INFO *cs __attribute__((unused)),
uchar * dest, uint len,
const uchar * src, uint srclen)
{
uint dstlen= len;
len= (uint) (strmake((char*) dest, (char*) src, min(len, srclen)) -
(char*) dest);
return (int) thai2sortable(dest, len);
len= thai2sortable(dest, len);
if (dstlen > len)
bfill(dest + len, dstlen - len, ' ');
return dstlen;
}

16
strings/ctype-uca.c
View File

@@ -7214,8 +7214,7 @@ static int my_strnxfrm_uca(CHARSET_INFO *cs,
uchar *dst, uint dstlen,
const uchar *src, uint srclen)
{
uchar *de = dst + dstlen;
const uchar *dst_orig = dst;
uchar *de = dst + (dstlen & (uint) ~1); // add even length for easier code
int s_res;
my_uca_scanner scanner;
scanner_handler->init(&scanner, cs, src, srclen);
@@ -7226,8 +7225,17 @@ static int my_strnxfrm_uca(CHARSET_INFO *cs,
dst[1]= s_res & 0xFF;
dst+= 2;
}
for ( ; dst < de; *dst++='\0');
return dst - dst_orig;
s_res= cs->sort_order_big[0][0x20 * cs->sort_order[0]];
while (dst < de)
{
dst[0]= s_res >> 8;
dst[1]= s_res & 0xFF;
dst+= 2;
}
if (dstlen & 1) // if odd number then fill the last char
*dst= '\0';
return dstlen;
}

9
strings/ctype-ucs2.c
View File

@@ -347,7 +347,6 @@ static int my_strnxfrm_ucs2(CHARSET_INFO *cs,
int plane;
uchar *de = dst + dstlen;
const uchar *se = src + srclen;
const uchar *dst_orig = dst;
while( src < se && dst < de )
{
@@ -367,7 +366,9 @@ static int my_strnxfrm_ucs2(CHARSET_INFO *cs,
}
dst+=res;
}
return dst - dst_orig;
if (dst < de)
cs->cset->fill(cs, dst, de - dst, ' ');
return dstlen;
}
@@ -1377,7 +1378,9 @@ int my_strnxfrm_ucs2_bin(CHARSET_INFO *cs __attribute__((unused)),
{
if (dst != src)
memcpy(dst,src,srclen= min(dstlen,srclen));
return srclen;
if (dstlen > srclen)
cs->cset->fill(cs, dst + srclen, dstlen - srclen, ' ');
return dstlen;
}

5
strings/ctype-utf8.c
View File

@@ -2228,7 +2228,6 @@ static int my_strnxfrm_utf8(CHARSET_INFO *cs,
int plane;
uchar *de = dst + dstlen;
const uchar *se = src + srclen;
const uchar *dst_orig = dst;
while( src < se && dst < de )
{
@@ -2248,7 +2247,9 @@ static int my_strnxfrm_utf8(CHARSET_INFO *cs,
}
dst+=res;
}
return dst - dst_orig;
if (dst < de)
bfill(dst, de - dst, ' ');
return dstlen;
}
static int my_ismbchar_utf8(CHARSET_INFO *cs,const char *b, const char *e)

4
strings/ctype-win1250ch.c
View File

@@ -503,7 +503,9 @@ static int my_strnxfrm_win1250ch(CHARSET_INFO * cs __attribute__((unused)),
dest[totlen] = value;
totlen++;
} while (value) ;
return totlen;
if (len > totlen)
bfill(dest + totlen, len - totlen, ' ');
return len;
}
#undef IS_END