summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--BitKeeper/etc/collapsed1
-rw-r--r--include/m_ctype.h12
-rw-r--r--mysql-test/r/lowercase_table.result24
-rw-r--r--mysql-test/r/rpl_ignore_table.result16
-rw-r--r--mysql-test/t/lowercase_table.test20
-rw-r--r--mysql-test/t/rpl_ignore_table-slave.opt2
-rw-r--r--mysql-test/t/rpl_ignore_table.test22
-rw-r--r--sql/sql_parse.cc20
-rw-r--r--strings/ctype-bin.c3
-rw-r--r--strings/ctype-mb.c24
-rw-r--r--strings/ctype-simple.c18
-rw-r--r--strings/ctype-ucs2.c8
-rw-r--r--strings/ctype-utf8.c135
13 files changed, 270 insertions, 35 deletions
diff --git a/BitKeeper/etc/collapsed b/BitKeeper/etc/collapsed
index 224b77be41b..dd7fddf1de3 100644
--- a/BitKeeper/etc/collapsed
+++ b/BitKeeper/etc/collapsed
@@ -8,3 +8,4 @@
4519a6c5BVUxEHTf5iJnjZkixMBs8g
451ab499rgdjXyOnUDqHu-wBDoS-OQ
452c6c6dAjuNghfc1ObZ_UQ5SCl85g
+4538a7b0EbDHHkWPbIwxO6ZIDdg6Dg
diff --git a/include/m_ctype.h b/include/m_ctype.h
index 5f946a3b443..0d586030735 100644
--- a/include/m_ctype.h
+++ b/include/m_ctype.h
@@ -190,8 +190,8 @@ typedef struct my_charset_handler_st
const unsigned char *s, const unsigned char *e);
/* Functions for case and sort conversion */
- void (*caseup_str)(struct charset_info_st *, char *);
- void (*casedn_str)(struct charset_info_st *, char *);
+ uint (*caseup_str)(struct charset_info_st *, char *);
+ uint (*casedn_str)(struct charset_info_st *, char *);
uint (*caseup)(struct charset_info_st *, char *src, uint srclen,
char *dst, uint dstlen);
uint (*casedn)(struct charset_info_st *, char *src, uint srclen,
@@ -324,8 +324,8 @@ extern uint my_instr_simple(struct charset_info_st *,
/* Functions for 8bit */
-extern void my_caseup_str_8bit(CHARSET_INFO *, char *);
-extern void my_casedn_str_8bit(CHARSET_INFO *, char *);
+extern uint my_caseup_str_8bit(CHARSET_INFO *, char *);
+extern uint my_casedn_str_8bit(CHARSET_INFO *, char *);
extern uint my_caseup_8bit(CHARSET_INFO *, char *src, uint srclen,
char *dst, uint dstlen);
extern uint my_casedn_8bit(CHARSET_INFO *, char *src, uint srclen,
@@ -415,8 +415,8 @@ int my_mbcharlen_8bit(CHARSET_INFO *, uint c);
/* Functions for multibyte charsets */
-extern void my_caseup_str_mb(CHARSET_INFO *, char *);
-extern void my_casedn_str_mb(CHARSET_INFO *, char *);
+extern uint my_caseup_str_mb(CHARSET_INFO *, char *);
+extern uint my_casedn_str_mb(CHARSET_INFO *, char *);
extern uint my_caseup_mb(CHARSET_INFO *, char *src, uint srclen,
char *dst, uint dstlen);
extern uint my_casedn_mb(CHARSET_INFO *, char *src, uint srclen,
diff --git a/mysql-test/r/lowercase_table.result b/mysql-test/r/lowercase_table.result
index 7705961d08d..6df3cf61ddb 100644
--- a/mysql-test/r/lowercase_table.result
+++ b/mysql-test/r/lowercase_table.result
@@ -84,3 +84,27 @@ create table t2 like T1;
drop table t1, t2;
show tables;
Tables_in_test
+set names utf8;
+drop table if exists İ,İİ;
+create table İ (s1 int);
+show create table İ;
+Table Create Table
+İ CREATE TABLE `i` (
+ `s1` int(11) default NULL
+) ENGINE=MyISAM DEFAULT CHARSET=latin1
+show tables;
+Tables_in_test
+i
+drop table İ;
+create table İİ (s1 int);
+show create table İİ;
+Table Create Table
+İİ CREATE TABLE `ii` (
+ `s1` int(11) default NULL
+) ENGINE=MyISAM DEFAULT CHARSET=latin1
+show tables;
+Tables_in_test
+ii
+drop table İİ;
+set names latin1;
+End of 5.0 tests
diff --git a/mysql-test/r/rpl_ignore_table.result b/mysql-test/r/rpl_ignore_table.result
index 356a9dcb2f8..136cf5cc5eb 100644
--- a/mysql-test/r/rpl_ignore_table.result
+++ b/mysql-test/r/rpl_ignore_table.result
@@ -14,3 +14,19 @@ SELECT * FROM t4;
a
DROP TABLE t1;
DROP TABLE t4;
+DROP TABLE IF EXISTS t5;
+CREATE TABLE t5 (
+word varchar(50) collate utf8_unicode_ci NOT NULL default ''
+) DEFAULT CHARSET=utf8 COLLATE=utf8_unicode_ci;
+SET @@session.character_set_client=33,@@session.collation_connection=192;
+CREATE TEMPORARY TABLE tmptbl504451f4258$1 (id INT NOT NULL) ENGINE=MEMORY;
+INSERT INTO t5 (word) VALUES ('TEST’');
+SELECT HEX(word) FROM t5;
+HEX(word)
+54455354E28099
+SELECT HEX(word) FROM t5;
+HEX(word)
+54455354E28099
+SELECT * FROM tmptbl504451f4258$1;
+ERROR 42S02: Table 'test.tmptbl504451f4258$1' doesn't exist
+DROP TABLE t5;
diff --git a/mysql-test/t/lowercase_table.test b/mysql-test/t/lowercase_table.test
index 96437bc7636..31513f1bd06 100644
--- a/mysql-test/t/lowercase_table.test
+++ b/mysql-test/t/lowercase_table.test
@@ -85,3 +85,23 @@ drop table t1, t2;
show tables;
# End of 4.1 tests
+
+
+#
+# Bug#20404: SHOW CREATE TABLE fails with Turkish I
+#
+set names utf8;
+--disable_warnings
+drop table if exists İ,İİ;
+--enable_warnings
+create table İ (s1 int);
+show create table İ;
+show tables;
+drop table İ;
+create table İİ (s1 int);
+show create table İİ;
+show tables;
+drop table İİ;
+set names latin1;
+
+--echo End of 5.0 tests
diff --git a/mysql-test/t/rpl_ignore_table-slave.opt b/mysql-test/t/rpl_ignore_table-slave.opt
index cb49119bfcb..3aabbb2e0f5 100644
--- a/mysql-test/t/rpl_ignore_table-slave.opt
+++ b/mysql-test/t/rpl_ignore_table-slave.opt
@@ -1 +1 @@
---replicate-ignore-table=test.t1 --replicate-ignore-table=test.t2 --replicate-ignore-table=test.t3
+--replicate-ignore-table=test.t1 --replicate-ignore-table=test.t2 --replicate-ignore-table=test.t3 --replicate-wild-ignore-table=%.tmptbl%
diff --git a/mysql-test/t/rpl_ignore_table.test b/mysql-test/t/rpl_ignore_table.test
index 84b0a4cde38..31191b9c386 100644
--- a/mysql-test/t/rpl_ignore_table.test
+++ b/mysql-test/t/rpl_ignore_table.test
@@ -28,3 +28,25 @@ connection master;
DROP TABLE t1;
DROP TABLE t4;
sync_slave_with_master;
+
+#
+# bug#22877 replication character sets get out of sync
+# using replicate-wild-ignore-table
+#
+--disable_warnings
+DROP TABLE IF EXISTS t5;
+--enable_warnings
+CREATE TABLE t5 (
+ word varchar(50) collate utf8_unicode_ci NOT NULL default ''
+) DEFAULT CHARSET=utf8 COLLATE=utf8_unicode_ci;
+SET @@session.character_set_client=33,@@session.collation_connection=192;
+CREATE TEMPORARY TABLE tmptbl504451f4258$1 (id INT NOT NULL) ENGINE=MEMORY;
+INSERT INTO t5 (word) VALUES ('TEST’');
+SELECT HEX(word) FROM t5;
+sync_slave_with_master;
+connection slave;
+SELECT HEX(word) FROM t5;
+--error 1146
+SELECT * FROM tmptbl504451f4258$1;
+connection master;
+DROP TABLE t5;
diff --git a/sql/sql_parse.cc b/sql/sql_parse.cc
index fc7ed7ff673..51fa86affc2 100644
--- a/sql/sql_parse.cc
+++ b/sql/sql_parse.cc
@@ -2543,7 +2543,23 @@ mysql_execute_command(THD *thd)
{
/* we warn the slave SQL thread */
my_message(ER_SLAVE_IGNORED_TABLE, ER(ER_SLAVE_IGNORED_TABLE), MYF(0));
- reset_one_shot_variables(thd);
+ if (thd->one_shot_set)
+ {
+ /*
+ It's ok to check thd->one_shot_set here:
+
+ The charsets in a MySQL 5.0 slave can change by both a binlogged
+ SET ONE_SHOT statement and the event-internal charset setting,
+ and these two ways to change charsets do not seems to work
+ together.
+
+ At least there seems to be problems in the rli cache for
+ charsets if we are using ONE_SHOT. Note that this is normally no
+ problem because either the >= 5.0 slave reads a 4.1 binlog (with
+ ONE_SHOT) *or* or 5.0 binlog (without ONE_SHOT) but never both."
+ */
+ reset_one_shot_variables(thd);
+ }
DBUG_RETURN(0);
}
}
@@ -6358,7 +6374,7 @@ TABLE_LIST *st_select_lex::add_table_to_list(THD *thd,
ptr->alias= alias_str;
if (lower_case_table_names && table->table.length)
- my_casedn_str(files_charset_info, table->table.str);
+ table->table.length= my_casedn_str(files_charset_info, table->table.str);
ptr->table_name=table->table.str;
ptr->table_name_length=table->table.length;
ptr->lock_type= lock_type;
diff --git a/strings/ctype-bin.c b/strings/ctype-bin.c
index 5758960ef6c..289b76c9e6e 100644
--- a/strings/ctype-bin.c
+++ b/strings/ctype-bin.c
@@ -211,9 +211,10 @@ static int my_strnncollsp_8bit_bin(CHARSET_INFO * cs __attribute__((unused)),
/* This function is used for all conversion functions */
-static void my_case_str_bin(CHARSET_INFO *cs __attribute__((unused)),
+static uint my_case_str_bin(CHARSET_INFO *cs __attribute__((unused)),
char *str __attribute__((unused)))
{
+ return 0;
}
static uint my_case_bin(CHARSET_INFO *cs __attribute__((unused)),
diff --git a/strings/ctype-mb.c b/strings/ctype-mb.c
index c945164ac9c..c3848c64219 100644
--- a/strings/ctype-mb.c
+++ b/strings/ctype-mb.c
@@ -21,40 +21,44 @@
#ifdef USE_MB
-void my_caseup_str_mb(CHARSET_INFO * cs, char *str)
+uint my_caseup_str_mb(CHARSET_INFO * cs, char *str)
{
register uint32 l;
- register uchar *map=cs->to_upper;
+ register uchar *map= cs->to_upper;
+ char *str_orig= str;
while (*str)
{
/* Pointing after the '\0' is safe here. */
- if ((l=my_ismbchar(cs, str, str + cs->mbmaxlen)))
- str+=l;
+ if ((l= my_ismbchar(cs, str, str + cs->mbmaxlen)))
+ str+= l;
else
{
- *str=(char) map[(uchar)*str];
+ *str= (char) map[(uchar)*str];
str++;
}
}
+ return str - str_orig;
}
-void my_casedn_str_mb(CHARSET_INFO * cs, char *str)
+uint my_casedn_str_mb(CHARSET_INFO * cs, char *str)
{
register uint32 l;
- register uchar *map=cs->to_lower;
+ register uchar *map= cs->to_lower;
+ char *str_orig= str;
while (*str)
{
/* Pointing after the '\0' is safe here. */
- if ((l=my_ismbchar(cs, str, str + cs->mbmaxlen)))
- str+=l;
+ if ((l= my_ismbchar(cs, str, str + cs->mbmaxlen)))
+ str+= l;
else
{
- *str=(char) map[(uchar)*str];
+ *str= (char) map[(uchar)*str];
str++;
}
}
+ return str - str_orig;
}
uint my_caseup_mb(CHARSET_INFO * cs, char *src, uint srclen,
diff --git a/strings/ctype-simple.c b/strings/ctype-simple.c
index 9b45d5a03b7..9d10ba82114 100644
--- a/strings/ctype-simple.c
+++ b/strings/ctype-simple.c
@@ -188,20 +188,26 @@ int my_strnncollsp_simple(CHARSET_INFO * cs, const uchar *a, uint a_length,
}
-void my_caseup_str_8bit(CHARSET_INFO * cs,char *str)
+uint my_caseup_str_8bit(CHARSET_INFO * cs,char *str)
{
- register uchar *map=cs->to_upper;
- while ((*str = (char) map[(uchar) *str]) != 0)
+ register uchar *map= cs->to_upper;
+ char *str_orig= str;
+ while ((*str= (char) map[(uchar) *str]) != 0)
str++;
+ return str - str_orig;
}
-void my_casedn_str_8bit(CHARSET_INFO * cs,char *str)
+
+uint my_casedn_str_8bit(CHARSET_INFO * cs,char *str)
{
- register uchar *map=cs->to_lower;
- while ((*str = (char) map[(uchar)*str]) != 0)
+ register uchar *map= cs->to_lower;
+ char *str_orig= str;
+ while ((*str= (char) map[(uchar) *str]) != 0)
str++;
+ return str - str_orig;
}
+
uint my_caseup_8bit(CHARSET_INFO * cs, char *src, uint srclen,
char *dst __attribute__((unused)),
uint dstlen __attribute__((unused)))
diff --git a/strings/ctype-ucs2.c b/strings/ctype-ucs2.c
index df43eff3d73..5089db6bf48 100644
--- a/strings/ctype-ucs2.c
+++ b/strings/ctype-ucs2.c
@@ -159,13 +159,13 @@ static void my_hash_sort_ucs2(CHARSET_INFO *cs, const uchar *s, uint slen,
}
-static void my_caseup_str_ucs2(CHARSET_INFO * cs __attribute__((unused)),
+static uint my_caseup_str_ucs2(CHARSET_INFO * cs __attribute__((unused)),
char * s __attribute__((unused)))
{
+ return 0;
}
-
static uint my_casedn_ucs2(CHARSET_INFO *cs, char *src, uint srclen,
char *dst __attribute__((unused)),
uint dstlen __attribute__((unused)))
@@ -188,9 +188,11 @@ static uint my_casedn_ucs2(CHARSET_INFO *cs, char *src, uint srclen,
return srclen;
}
-static void my_casedn_str_ucs2(CHARSET_INFO *cs __attribute__((unused)),
+
+static uint my_casedn_str_ucs2(CHARSET_INFO *cs __attribute__((unused)),
char * s __attribute__((unused)))
{
+ return 0;
}
diff --git a/strings/ctype-utf8.c b/strings/ctype-utf8.c
index 6ed7736e24a..8a4ed48bef5 100644
--- a/strings/ctype-utf8.c
+++ b/strings/ctype-utf8.c
@@ -2047,6 +2047,52 @@ static int my_utf8_uni(CHARSET_INFO *cs __attribute__((unused)),
return MY_CS_ILSEQ;
}
+
+/*
+ The same as above, but without range check
+ for example, for a null-terminated string
+*/
+static int my_utf8_uni_no_range(CHARSET_INFO *cs __attribute__((unused)),
+ my_wc_t * pwc, const uchar *s)
+{
+ unsigned char c;
+
+ c= s[0];
+ if (c < 0x80)
+ {
+ *pwc = c;
+ return 1;
+ }
+
+ if (c < 0xc2)
+ return MY_CS_ILSEQ;
+
+ if (c < 0xe0)
+ {
+ if (!((s[1] ^ 0x80) < 0x40))
+ return MY_CS_ILSEQ;
+
+ *pwc = ((my_wc_t) (c & 0x1f) << 6) | (my_wc_t) (s[1] ^ 0x80);
+ return 2;
+ }
+
+ if (c < 0xf0)
+ {
+ if (!((s[1] ^ 0x80) < 0x40 &&
+ (s[2] ^ 0x80) < 0x40 &&
+ (c >= 0xe1 || s[1] >= 0xa0)))
+ return MY_CS_ILSEQ;
+
+ *pwc= ((my_wc_t) (c & 0x0f) << 12) |
+ ((my_wc_t) (s[1] ^ 0x80) << 6) |
+ (my_wc_t) (s[2] ^ 0x80);
+
+ return 3;
+ }
+ return MY_CS_ILSEQ;
+}
+
+
static int my_uni_utf8 (CHARSET_INFO *cs __attribute__((unused)) ,
my_wc_t wc, uchar *r, uchar *e)
{
@@ -2093,6 +2139,34 @@ static int my_uni_utf8 (CHARSET_INFO *cs __attribute__((unused)) ,
}
+/*
+ The same as above, but without range check.
+*/
+static int my_uni_utf8_no_range(CHARSET_INFO *cs __attribute__((unused)),
+ my_wc_t wc, uchar *r)
+{
+ int count;
+
+ if (wc < 0x80)
+ count= 1;
+ else if (wc < 0x800)
+ count= 2;
+ else if (wc < 0x10000)
+ count= 3;
+ else
+ return MY_CS_ILUNI;
+
+ switch (count)
+ {
+ /* Fall through all cases!!! */
+ case 3: r[2]= (uchar) (0x80 | (wc & 0x3f)); wc= wc >> 6; wc |= 0x800;
+ case 2: r[1]= (uchar) (0x80 | (wc & 0x3f)); wc= wc >> 6; wc |= 0xc0;
+ case 1: r[0]= (uchar) wc;
+ }
+ return count;
+}
+
+
static uint my_caseup_utf8(CHARSET_INFO *cs, char *src, uint srclen,
char *dst, uint dstlen)
{
@@ -2143,10 +2217,26 @@ static void my_hash_sort_utf8(CHARSET_INFO *cs, const uchar *s, uint slen,
}
-static void my_caseup_str_utf8(CHARSET_INFO * cs, char * s)
+static uint my_caseup_str_utf8(CHARSET_INFO *cs, char *src)
{
- uint len= (uint) strlen(s);
- my_caseup_utf8(cs, s, len, s, len);
+ my_wc_t wc;
+ int srcres, dstres;
+ char *dst= src, *dst0= src;
+ MY_UNICASE_INFO **uni_plane= cs->caseinfo;
+ DBUG_ASSERT(cs->caseup_multiply == 1);
+
+ while (*src &&
+ (srcres= my_utf8_uni_no_range(cs, &wc, (uchar *) src)) > 0)
+ {
+ int plane= (wc>>8) & 0xFF;
+ wc= uni_plane[plane] ? uni_plane[plane][wc & 0xFF].toupper : wc;
+ if ((dstres= my_uni_utf8_no_range(cs, wc, (uchar*) dst)) <= 0)
+ break;
+ src+= srcres;
+ dst+= dstres;
+ }
+ *dst= '\0';
+ return (uint) (dst - dst0);
}
@@ -2172,10 +2262,43 @@ static uint my_casedn_utf8(CHARSET_INFO *cs, char *src, uint srclen,
return (uint) (dst - dst0);
}
-static void my_casedn_str_utf8(CHARSET_INFO *cs, char * s)
+
+static uint my_casedn_str_utf8(CHARSET_INFO *cs, char *src)
{
- uint len= (uint) strlen(s);
- my_casedn_utf8(cs, s, len, s, len);
+ my_wc_t wc;
+ int srcres, dstres;
+ char *dst= src, *dst0= src;
+ MY_UNICASE_INFO **uni_plane= cs->caseinfo;
+ DBUG_ASSERT(cs->casedn_multiply == 1);
+
+ while (*src &&
+ (srcres= my_utf8_uni_no_range(cs, &wc, (uchar *) src)) > 0)
+ {
+ int plane= (wc>>8) & 0xFF;
+ wc= uni_plane[plane] ? uni_plane[plane][wc & 0xFF].tolower : wc;
+ if ((dstres= my_uni_utf8_no_range(cs, wc, (uchar*) dst)) <= 0)
+ break;
+ src+= srcres;
+ dst+= dstres;
+ }
+
+ /*
+ In rare cases lower string can be shorter than
+ the original string, for example:
+
+ "U+0130 LATIN CAPITAL LETTER I WITH DOT ABOVE"
+ (which is 0xC4B0 in utf8, i.e. two bytes)
+
+ is converted into
+
+ "U+0069 LATIN SMALL LETTER I"
+ (which is 0x69 in utf8, i.e. one byte)
+
+ So, we need to put '\0' terminator after converting.
+ */
+
+ *dst= '\0';
+ return (uint) (dst - dst0);
}
View Lorry Controller Status