Auto-merge from mysql-next-mr.

.bzrignore

@@ -3074,3 +3074,4 @@ libmysqld/rpl_handler.cc
 dbug/tests
 libmysqld/sys_vars.cc
 libmysqld/keycaches.cc
+client/dtoa.c

include/m_ctype.h

@@ -40,9 +40,9 @@ extern "C" {
 
 typedef struct unicase_info_st
 {
-  uint16 toupper;
-  uint16 tolower;
-  uint16 sort;
+  uint32 toupper;
+  uint32 tolower;
+  uint32 sort;
 } MY_UNICASE_INFO;
 
 
@@ -439,6 +439,14 @@ extern size_t my_caseup_mb(CHARSET_INFO *, char *src, size_t srclen,
                                          char *dst, size_t dstlen);
 extern size_t my_casedn_mb(CHARSET_INFO *, char *src, size_t srclen,
                                          char *dst, size_t dstlen);
+extern size_t my_caseup_mb_varlen(CHARSET_INFO *, char *src, size_t srclen,
+                                  char *dst, size_t dstlen);
+extern size_t my_casedn_mb_varlen(CHARSET_INFO *, char *src, size_t srclen,
+                                  char *dst, size_t dstlen);
+extern size_t my_caseup_ujis(CHARSET_INFO *, char *src, size_t srclen,
+                             char *dst, size_t dstlen);
+extern size_t my_casedn_ujis(CHARSET_INFO *, char *src, size_t srclen,
+                             char *dst, size_t dstlen);
 extern int my_strcasecmp_mb(CHARSET_INFO * cs,const char *, const char *);
 
 int my_wildcmp_mb(CHARSET_INFO *,

mysql-test/collections/default.experimental

@@ -11,16 +11,28 @@ main.information_schema                  # Bug#47449 2009-09-19 alik main.inform
 main.lock_multi_bug38499                 # Bug#47448 2009-09-19 alik main.lock_multi_bug38499 times out sporadically
 main.lock_multi_bug38691 @solaris        # Bug#47792 2009-10-02 alik main.lock_multi_bug38691 times out sporadically on Solaris 10
 main.log_tables                          # Bug#47924 2009-10-08 alik main.log_tables times out sporadically 
+main.outfile_loaddata @solaris           # bug#46895 2010-01-20 alik Test "outfile_loaddata" fails (reproducible)
 main.plugin                              # Bug#47146 Linking problem with example plugin when dtrace enabled
+main.signal_demo3 @solaris               # Bug#47791 2010-01-20 alik Several test cases fail on Solaris with error Thread stack overrun
+main.sp @solaris                         # Bug#47791 2010-01-20 alik Several test cases fail on Solaris with error Thread stack overrun
+
+perfschema.tampered_perfschema_table1 @windows # Bug#50478 2010-01-20 alik perfschema.tampered_perfschema_table1 fails sporadically on Windows and Solaris
+perfschema.tampered_perfschema_table1 @solaris # Bug#50478 2010-01-20 alik perfschema.tampered_perfschema_table1 fails sporadically on Windows and Solaris
 
 rpl.rpl_get_master_version_and_clock*    # Bug#49191 2009-12-01 Daogang rpl_get_master_version_and_clock failed on PB2: COM_REGISTER_SLAVE failed
 rpl.rpl_heartbeat_basic                  # BUG#43828 2009-10-22 luis fails sporadically
 rpl.rpl_heartbeat_2slaves                # BUG#43828 2009-10-22 luis fails sporadically
 rpl.rpl_innodb_bug28430*                 # Bug#46029
 rpl.rpl_innodb_bug30888* @solaris        # Bug#47646 2009-09-25 alik rpl.rpl_innodb_bug30888 fails sporadically on Solaris
+rpl.rpl_killed_ddl @windows              # Bug#47638 2010-01-20 alik The rpl_killed_ddl test fails on Windows
 rpl.rpl_plugin_load*  @solaris           # Bug#47146
+rpl.rpl_row_sp011* @solaris              # Bug#47791 2010-01-20 alik Several test cases fail on Solaris with error Thread stack overrun
+rpl.rpl_slave_load_remove_tmpfile* @windows # Bug#50474 2010-01-20 alik rpl_slave_load_remove_tmpfile failed on windows debug enabled binary
+rpl.rpl_sync* @windows                    # Bug#50473 2010-01-20 alik rpl_sync fails on windows debug enabled binaries
 rpl.rpl_timezone*                        # Bug#47017 2009-10-27 alik rpl_timezone fails on PB-2 with mismatch error
 
+sys_vars.max_sp_recursion_depth_func @solaris # Bug#47791 2010-01-20 alik Several test cases fail on Solaris with error Thread stack overrun
+
 # Declare all NDB-tests in ndb and rpl_ndb test suites experimental.
 # Usually the test cases from ndb and rpl_ndb test suites are not run in PB,
 # but they run by gcov.

mysql-test/r/ctype_big5.result

@@ -277,3 +277,281 @@ hex(a)
 EE00
 drop table t1;
 End of 5.0 tests
+#
+# Start of 5.5 tests
+#
+#
+# Testing WL#4583 Case conversion in Asian character sets 
+#
+SET NAMES utf8;
+SET collation_connection=big5_chinese_ci;
+CREATE TABLE t1 (b VARCHAR(2));
+INSERT INTO t1 VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7');
+INSERT INTO t1 VALUES ('8'),('9'),('A'),('B'),('C'),('D'),('E'),('F');
+CREATE TEMPORARY TABLE head AS SELECT concat(b1.b, b2.b) AS head FROM t1 b1, t1 b2;
+CREATE TEMPORARY TABLE tail AS SELECT concat(b1.b, b2.b) AS tail FROM t1 b1, t1 b2;
+DROP TABLE t1;
+CREATE TABLE t1 AS
+SELECT concat(head, tail) AS code, ' ' AS a
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+DROP TEMPORARY TABLE head, tail;
+SHOW CREATE TABLE t1;
+Table	Create Table
+t1	CREATE TABLE `t1` (
+  `code` varchar(8) DEFAULT NULL,
+  `a` varchar(1) CHARACTER SET big5 NOT NULL DEFAULT ''
+) ENGINE=MyISAM DEFAULT CHARSET=latin1
+SELECT COUNT(*) FROM t1;
+COUNT(*)
+28672
+UPDATE t1 SET a=unhex(code) ORDER BY code;
+Warnings:
+Warning	1366	Incorrect string value: '\x80 ' for column 'a' at row 1
+Warning	1366	Incorrect string value: '\x80!' for column 'a' at row 2
+Warning	1366	Incorrect string value: '\x80"' for column 'a' at row 3
+Warning	1366	Incorrect string value: '\x80#' for column 'a' at row 4
+Warning	1366	Incorrect string value: '\x80$' for column 'a' at row 5
+Warning	1366	Incorrect string value: '\x80%' for column 'a' at row 6
+Warning	1366	Incorrect string value: '\x80&' for column 'a' at row 7
+Warning	1366	Incorrect string value: '\x80'' for column 'a' at row 8
+Warning	1366	Incorrect string value: '\x80(' for column 'a' at row 9
+Warning	1366	Incorrect string value: '\x80)' for column 'a' at row 10
+Warning	1366	Incorrect string value: '\x80*' for column 'a' at row 11
+Warning	1366	Incorrect string value: '\x80+' for column 'a' at row 12
+Warning	1366	Incorrect string value: '\x80,' for column 'a' at row 13
+Warning	1366	Incorrect string value: '\x80-' for column 'a' at row 14
+Warning	1366	Incorrect string value: '\x80.' for column 'a' at row 15
+Warning	1366	Incorrect string value: '\x80/' for column 'a' at row 16
+Warning	1366	Incorrect string value: '\x800' for column 'a' at row 17
+Warning	1366	Incorrect string value: '\x801' for column 'a' at row 18
+Warning	1366	Incorrect string value: '\x802' for column 'a' at row 19
+Warning	1366	Incorrect string value: '\x803' for column 'a' at row 20
+Warning	1366	Incorrect string value: '\x804' for column 'a' at row 21
+Warning	1366	Incorrect string value: '\x805' for column 'a' at row 22
+Warning	1366	Incorrect string value: '\x806' for column 'a' at row 23
+Warning	1366	Incorrect string value: '\x807' for column 'a' at row 24
+Warning	1366	Incorrect string value: '\x808' for column 'a' at row 25
+Warning	1366	Incorrect string value: '\x809' for column 'a' at row 26
+Warning	1366	Incorrect string value: '\x80:' for column 'a' at row 27
+Warning	1366	Incorrect string value: '\x80;' for column 'a' at row 28
+Warning	1366	Incorrect string value: '\x80<' for column 'a' at row 29
+Warning	1366	Incorrect string value: '\x80=' for column 'a' at row 30
+Warning	1366	Incorrect string value: '\x80>' for column 'a' at row 31
+Warning	1366	Incorrect string value: '\x80?' for column 'a' at row 32
+Warning	1366	Incorrect string value: '\x80@' for column 'a' at row 33
+Warning	1366	Incorrect string value: '\x80A' for column 'a' at row 34
+Warning	1366	Incorrect string value: '\x80B' for column 'a' at row 35
+Warning	1366	Incorrect string value: '\x80C' for column 'a' at row 36
+Warning	1366	Incorrect string value: '\x80D' for column 'a' at row 37
+Warning	1366	Incorrect string value: '\x80E' for column 'a' at row 38
+Warning	1366	Incorrect string value: '\x80F' for column 'a' at row 39
+Warning	1366	Incorrect string value: '\x80G' for column 'a' at row 40
+Warning	1366	Incorrect string value: '\x80H' for column 'a' at row 41
+Warning	1366	Incorrect string value: '\x80I' for column 'a' at row 42
+Warning	1366	Incorrect string value: '\x80J' for column 'a' at row 43
+Warning	1366	Incorrect string value: '\x80K' for column 'a' at row 44
+Warning	1366	Incorrect string value: '\x80L' for column 'a' at row 45
+Warning	1366	Incorrect string value: '\x80M' for column 'a' at row 46
+Warning	1366	Incorrect string value: '\x80N' for column 'a' at row 47
+Warning	1366	Incorrect string value: '\x80O' for column 'a' at row 48
+Warning	1366	Incorrect string value: '\x80P' for column 'a' at row 49
+Warning	1366	Incorrect string value: '\x80Q' for column 'a' at row 50
+Warning	1366	Incorrect string value: '\x80R' for column 'a' at row 51
+Warning	1366	Incorrect string value: '\x80S' for column 'a' at row 52
+Warning	1366	Incorrect string value: '\x80T' for column 'a' at row 53
+Warning	1366	Incorrect string value: '\x80U' for column 'a' at row 54
+Warning	1366	Incorrect string value: '\x80V' for column 'a' at row 55
+Warning	1366	Incorrect string value: '\x80W' for column 'a' at row 56
+Warning	1366	Incorrect string value: '\x80X' for column 'a' at row 57
+Warning	1366	Incorrect string value: '\x80Y' for column 'a' at row 58
+Warning	1366	Incorrect string value: '\x80Z' for column 'a' at row 59
+Warning	1366	Incorrect string value: '\x80[' for column 'a' at row 60
+Warning	1366	Incorrect string value: '\x80\' for column 'a' at row 61
+Warning	1366	Incorrect string value: '\x80]' for column 'a' at row 62
+Warning	1366	Incorrect string value: '\x80^' for column 'a' at row 63
+Warning	1366	Incorrect string value: '\x80_' for column 'a' at row 64
+SELECT COUNT(*) FROM t1 WHERE a<>'';
+COUNT(*)
+13973
+SELECT code, hex(upper(a)), hex(lower(a)),a, upper(a), lower(a) FROM t1 WHERE hex(a)<>hex(upper(a)) OR hex(a)<>hex(lower(a));
+code	hex(upper(a))	hex(lower(a))	a	upper(a)	lower(a)
+A2CF	A2CF	A2E9	Ａ	Ａ	ａ
+A2D0	A2D0	A2EA	Ｂ	Ｂ	ｂ
+A2D1	A2D1	A2EB	Ｃ	Ｃ	ｃ
+A2D2	A2D2	A2EC	Ｄ	Ｄ	ｄ
+A2D3	A2D3	A2ED	Ｅ	Ｅ	ｅ
+A2D4	A2D4	A2EE	Ｆ	Ｆ	ｆ
+A2D5	A2D5	A2EF	Ｇ	Ｇ	ｇ
+A2D6	A2D6	A2F0	Ｈ	Ｈ	ｈ
+A2D7	A2D7	A2F1	Ｉ	Ｉ	ｉ
+A2D8	A2D8	A2F2	Ｊ	Ｊ	ｊ
+A2D9	A2D9	A2F3	Ｋ	Ｋ	ｋ
+A2DA	A2DA	A2F4	Ｌ	Ｌ	ｌ
+A2DB	A2DB	A2F5	Ｍ	Ｍ	ｍ
+A2DC	A2DC	A2F6	Ｎ	Ｎ	ｎ
+A2DD	A2DD	A2F7	Ｏ	Ｏ	ｏ
+A2DE	A2DE	A2F8	Ｐ	Ｐ	ｐ
+A2DF	A2DF	A2F9	Ｑ	Ｑ	ｑ
+A2E0	A2E0	A2FA	Ｒ	Ｒ	ｒ
+A2E1	A2E1	A2FB	Ｓ	Ｓ	ｓ
+A2E2	A2E2	A2FC	Ｔ	Ｔ	ｔ
+A2E3	A2E3	A2FD	Ｕ	Ｕ	ｕ
+A2E4	A2E4	A2FE	Ｖ	Ｖ	ｖ
+A2E5	A2E5	A340	Ｗ	Ｗ	ｗ
+A2E6	A2E6	A341	Ｘ	Ｘ	ｘ
+A2E7	A2E7	A342	Ｙ	Ｙ	ｙ
+A2E8	A2E8	A343	Ｚ	Ｚ	ｚ
+A2E9	A2CF	A2E9	ａ	Ａ	ａ
+A2EA	A2D0	A2EA	ｂ	Ｂ	ｂ
+A2EB	A2D1	A2EB	ｃ	Ｃ	ｃ
+A2EC	A2D2	A2EC	ｄ	Ｄ	ｄ
+A2ED	A2D3	A2ED	ｅ	Ｅ	ｅ
+A2EE	A2D4	A2EE	ｆ	Ｆ	ｆ
+A2EF	A2D5	A2EF	ｇ	Ｇ	ｇ
+A2F0	A2D6	A2F0	ｈ	Ｈ	ｈ
+A2F1	A2D7	A2F1	ｉ	Ｉ	ｉ
+A2F2	A2D8	A2F2	ｊ	Ｊ	ｊ
+A2F3	A2D9	A2F3	ｋ	Ｋ	ｋ
+A2F4	A2DA	A2F4	ｌ	Ｌ	ｌ
+A2F5	A2DB	A2F5	ｍ	Ｍ	ｍ
+A2F6	A2DC	A2F6	ｎ	Ｎ	ｎ
+A2F7	A2DD	A2F7	ｏ	Ｏ	ｏ
+A2F8	A2DE	A2F8	ｐ	Ｐ	ｐ
+A2F9	A2DF	A2F9	ｑ	Ｑ	ｑ
+A2FA	A2E0	A2FA	ｒ	Ｒ	ｒ
+A2FB	A2E1	A2FB	ｓ	Ｓ	ｓ
+A2FC	A2E2	A2FC	ｔ	Ｔ	ｔ
+A2FD	A2E3	A2FD	ｕ	Ｕ	ｕ
+A2FE	A2E4	A2FE	ｖ	Ｖ	ｖ
+A340	A2E5	A340	ｗ	Ｗ	ｗ
+A341	A2E6	A341	ｘ	Ｘ	ｘ
+A342	A2E7	A342	ｙ	Ｙ	ｙ
+A343	A2E8	A343	ｚ	Ｚ	ｚ
+A344	A344	A35C	Α	Α	α
+A345	A345	A35D	Β	Β	β
+A346	A346	A35E	Γ	Γ	γ
+A347	A347	A35F	Δ	Δ	δ
+A348	A348	A360	Ε	Ε	ε
+A349	A349	A361	Ζ	Ζ	ζ
+A34A	A34A	A362	Η	Η	η
+A34B	A34B	A363	Θ	Θ	θ
+A34C	A34C	A364	Ι	Ι	ι
+A34D	A34D	A365	Κ	Κ	κ
+A34E	A34E	A366	Λ	Λ	λ
+A34F	A34F	A367	Μ	Μ	μ
+A350	A350	A368	Ν	Ν	ν
+A351	A351	A369	Ξ	Ξ	ξ
+A352	A352	A36A	Ο	Ο	ο
+A353	A353	A36B	Π	Π	π
+A354	A354	A36C	Ρ	Ρ	ρ
+A355	A355	A36D	Σ	Σ	σ
+A356	A356	A36E	Τ	Τ	τ
+A357	A357	A36F	Υ	Υ	υ
+A358	A358	A370	Φ	Φ	φ
+A359	A359	A371	Χ	Χ	χ
+A35A	A35A	A372	Ψ	Ψ	ψ
+A35B	A35B	A373	Ω	Ω	ω
+A35C	A344	A35C	α	Α	α
+A35D	A345	A35D	β	Β	β
+A35E	A346	A35E	γ	Γ	γ
+A35F	A347	A35F	δ	Δ	δ
+A360	A348	A360	ε	Ε	ε
+A361	A349	A361	ζ	Ζ	ζ
+A362	A34A	A362	η	Η	η
+A363	A34B	A363	θ	Θ	θ
+A364	A34C	A364	ι	Ι	ι
+A365	A34D	A365	κ	Κ	κ
+A366	A34E	A366	λ	Λ	λ
+A367	A34F	A367	μ	Μ	μ
+A368	A350	A368	ν	Ν	ν
+A369	A351	A369	ξ	Ξ	ξ
+A36A	A352	A36A	ο	Ο	ο
+A36B	A353	A36B	π	Π	π
+A36C	A354	A36C	ρ	Ρ	ρ
+A36D	A355	A36D	σ	Σ	σ
+A36E	A356	A36E	τ	Τ	τ
+A36F	A357	A36F	υ	Υ	υ
+A370	A358	A370	φ	Φ	φ
+A371	A359	A371	χ	Χ	χ
+A372	A35A	A372	ψ	Ψ	ψ
+A373	A35B	A373	ω	Ω	ω
+C7B1	C7B1	C7CC	Д	Д	д
+C7B2	C7B2	C7CD	Е	Е	е
+C7B3	C7B3	C7CE	Ё	Ё	ё
+C7B4	C7B4	C7CF	Ж	Ж	ж
+C7B5	C7B5	C7D0	З	З	з
+C7B6	C7B6	C7D1	И	И	и
+C7B7	C7B7	C7D2	Й	Й	й
+C7B8	C7B8	C7D3	К	К	к
+C7B9	C7B9	C7D4	Л	Л	л
+C7BA	C7BA	C7D5	М	М	м
+C7BB	C7BB	C7DC	У	У	у
+C7BC	C7BC	C7DD	Ф	Ф	ф
+C7BD	C7BD	C7DE	Х	Х	х
+C7BE	C7BE	C7DF	Ц	Ц	ц
+C7BF	C7BF	C7E0	Ч	Ч	ч
+C7C0	C7C0	C7E1	Ш	Ш	ш
+C7C1	C7C1	C7E2	Щ	Щ	щ
+C7C2	C7C2	C7E3	Ъ	Ъ	ъ
+C7C3	C7C3	C7E4	Ы	Ы	ы
+C7C4	C7C4	C7E5	Ь	Ь	ь
+C7C5	C7C5	C7E6	Э	Э	э
+C7C6	C7C6	C7E7	Ю	Ю	ю
+C7C7	C7C7	C7E8	Я	Я	я
+C7CC	C7B1	C7CC	д	Д	д
+C7CD	C7B2	C7CD	е	Е	е
+C7CE	C7B3	C7CE	ё	Ё	ё
+C7CF	C7B4	C7CF	ж	Ж	ж
+C7D0	C7B5	C7D0	з	З	з
+C7D1	C7B6	C7D1	и	И	и
+C7D2	C7B7	C7D2	й	Й	й
+C7D3	C7B8	C7D3	к	К	к
+C7D4	C7B9	C7D4	л	Л	л
+C7D5	C7BA	C7D5	м	М	м
+C7DC	C7BB	C7DC	у	У	у
+C7DD	C7BC	C7DD	ф	Ф	ф
+C7DE	C7BD	C7DE	х	Х	х
+C7DF	C7BE	C7DF	ц	Ц	ц
+C7E0	C7BF	C7E0	ч	Ч	ч
+C7E1	C7C0	C7E1	ш	Ш	ш
+C7E2	C7C1	C7E2	щ	Щ	щ
+C7E3	C7C2	C7E3	ъ	Ъ	ъ
+C7E4	C7C3	C7E4	ы	Ы	ы
+C7E5	C7C4	C7E5	ь	Ь	ь
+C7E6	C7C5	C7E6	э	Э	э
+C7E7	C7C6	C7E7	ю	Ю	ю
+C7E8	C7C7	C7E8	я	Я	я
+SELECT * FROM t1
+WHERE HEX(CAST(LOWER(a) AS CHAR CHARACTER SET utf8)) <>
+HEX(LOWER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+code	a
+A2B9	Ⅰ
+A2BA	Ⅱ
+A2BB	Ⅲ
+A2BC	Ⅳ
+A2BD	Ⅴ
+A2BE	Ⅵ
+A2BF	Ⅶ
+A2C0	Ⅷ
+A2C1	Ⅸ
+A2C2	Ⅹ
+SELECT * FROM t1
+WHERE HEX(CAST(UPPER(a) AS CHAR CHARACTER SET utf8)) <>
+HEX(UPPER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+code	a
+C7C8	а
+C7C9	б
+C7CA	в
+C7CB	г
+C7D6	н
+C7D7	о
+C7D8	п
+C7D9	р
+C7DA	с
+C7DB	т
+DROP TABLE t1;
+#
+# End of 5.5 tests
+#

mysql-test/t/ctype_big5.test

@@ -84,3 +84,64 @@ select hex(a) from t1;
 drop table t1;
 
 --echo End of 5.0 tests
+
+
+--echo #
+--echo # Start of 5.5 tests
+--echo #
+
+--echo #
+--echo # Testing WL#4583 Case conversion in Asian character sets 
+--echo #
+#
+# Populate t1 with all hex digits
+#
+SET NAMES utf8;
+SET collation_connection=big5_chinese_ci;
+CREATE TABLE t1 (b VARCHAR(2));
+INSERT INTO t1 VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7');
+INSERT INTO t1 VALUES ('8'),('9'),('A'),('B'),('C'),('D'),('E'),('F');
+#
+# Populate tables head and tail with values '00'-'FF'
+#
+CREATE TEMPORARY TABLE head AS SELECT concat(b1.b, b2.b) AS head FROM t1 b1, t1 b2;
+CREATE TEMPORARY TABLE tail AS SELECT concat(b1.b, b2.b) AS tail FROM t1 b1, t1 b2;
+DROP TABLE t1;
+#
+# Populate table t1 with all values [80..FF][20..FF]
+# Expected valid big5 codes: [A1..F9][40..7E,A1..FE] (89x157=13973)
+#
+CREATE TABLE t1 AS
+SELECT concat(head, tail) AS code, ' ' AS a
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+DROP TEMPORARY TABLE head, tail;
+SHOW CREATE TABLE t1;
+SELECT COUNT(*) FROM t1;
+UPDATE t1 SET a=unhex(code) ORDER BY code;
+SELECT COUNT(*) FROM t1 WHERE a<>'';
+#
+# Display all characters that have upper or lower case mapping.
+#
+SELECT code, hex(upper(a)), hex(lower(a)),a, upper(a), lower(a) FROM t1 WHERE hex(a)<>hex(upper(a)) OR hex(a)<>hex(lower(a));
+#
+# Make sure all possible conversion happened
+#
+# Expect U+2160 to U+2169 ROMAN NUMERAL ONE to ROMAN NUMERAL TEN
+#
+SELECT * FROM t1
+WHERE HEX(CAST(LOWER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(LOWER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+#
+# Expect U+0430 to U+0433 CYRILLIC SMALL LETTER A, BE, VE, GHE
+# Expect U+043D to U+0442 CYRILLIC SMALL LETTER EN, O, PE, ER, ES, TE
+#
+SELECT * FROM t1
+WHERE HEX(CAST(UPPER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(UPPER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+DROP TABLE t1;
+
+--echo #
+--echo # End of 5.5 tests
+--echo #

mysql-test/t/ctype_cp932_binlog_stm.test

@@ -41,3 +41,79 @@ SELECT hex(a), hex(lower(a)), hex(upper(a)) FROM t1 ORDER BY binary(a);
 DROP TABLE t1;
 
 --echo End of 5.1 tests
+
+
+--echo #
+--echo # Start of 5.5 tests
+--echo #
+
+--echo #
+--echo # Testing WL#4583 Case conversion in Asian character sets 
+--echo #
+#
+# Populate t1 with all hex digits
+#
+SET NAMES utf8;
+SET collation_connection=cp932_japanese_ci;
+CREATE TABLE t1 (b VARCHAR(2));
+INSERT INTO t1 VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7');
+INSERT INTO t1 VALUES ('8'),('9'),('A'),('B'),('C'),('D'),('E'),('F');
+#
+# Populate tables head and tail with values '00'-'FF'
+#
+CREATE TEMPORARY TABLE head AS SELECT concat(b1.b, b2.b) AS head FROM t1 b1, t1 b2;
+CREATE TEMPORARY TABLE tail AS SELECT concat(b1.b, b2.b) AS tail FROM t1 b1, t1 b2;
+DROP TABLE t1;
+#
+# Populate table t1 with all codes [80..FF][20..FF]
+# excluding Half Width Kana [A1..DF]
+# Expected valid cp932 multibyte codes:
+# [81..9F,E0..FC][40..7E,80..fC]   (60x188=11280 characters)
+#
+CREATE TABLE t1 AS
+SELECT concat(head, tail) AS code, ' ' AS a
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (head NOT BETWEEN 'A1' AND 'DF')
+AND (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+#
+# Populate t1 with Half Width Kana [A1..DF]
+#
+INSERT t1 (code) SELECT head FROM head
+WHERE (head BETWEEN 'A1' AND 'DF')
+ORDER BY head;
+DROP TEMPORARY TABLE head, tail;
+SHOW CREATE TABLE t1;
+UPDATE t1 SET a=unhex(code) ORDER BY code;
+SELECT COUNT(*) FROM t1;
+SELECT COUNT(*) FROM t1 WHERE a<>'' AND OCTET_LENGTH(a)=1;
+SELECT COUNT(*) FROM t1 WHERE a<>'' AND OCTET_LENGTH(a)=2;
+#
+# Display all characters that have upper or lower case mapping.
+#
+SELECT code, hex(upper(a)), hex(lower(a)),a, upper(a), lower(a) FROM t1
+WHERE hex(a)<>hex(upper(a)) OR hex(a)<>hex(lower(a))
+ORDER BY code;
+#
+# Make sure all possible conversion happened
+#
+# Expect U+212B ANGSTROM SIGN
+#
+SELECT * FROM t1
+WHERE HEX(CAST(LOWER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(LOWER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+#
+# Expect no results
+#
+SELECT * FROM t1
+WHERE HEX(CAST(UPPER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(UPPER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+
+DROP TABLE t1;
+
+
+
+
+--echo #
+--echo # End of 5.5 tests
+--echo #

mysql-test/t/ctype_eucjpms.test

@@ -381,3 +381,89 @@ select hex(convert(_eucjpms 0xA5FE41 using ucs2));
 # the next character, which is a single byte character 0x41.
 select hex(convert(_eucjpms 0x8FABF841 using ucs2));
 
+
+--echo #
+--echo # Start of 5.5 tests
+--echo #
+
+--echo #
+--echo # Testing WL#4583 Case conversion in Asian character sets 
+--echo #
+#
+# Populate t1 with all hex digits
+#
+SET NAMES utf8;
+SET collation_connection=eucjpms_japanese_ci;
+CREATE TABLE t1 (b VARCHAR(2));
+INSERT INTO t1 VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7');
+INSERT INTO t1 VALUES ('8'),('9'),('A'),('B'),('C'),('D'),('E'),('F');
+#
+# Populate tables head and tail with values '00'-'FF'
+#
+CREATE TEMPORARY TABLE head AS SELECT concat(b1.b, b2.b) AS head FROM t1 b1, t1 b2;
+CREATE TEMPORARY TABLE tail AS SELECT concat(b1.b, b2.b) AS tail FROM t1 b1, t1 b2;
+DROP TABLE t1;
+#
+# Populate table t1 with all eucjpms codes.
+#
+CREATE TABLE t1 AS SELECT 'XXXXXX' AS code, ' ' AS a LIMIT 0;
+#
+# Pupulate JIS-X-0201 range (Half Width Kana)
+# Expected valid code range: [8E][A1..DF]   (1x63 characters)
+#
+INSERT INTO t1 (code) SELECT concat('8E', head) FROM head
+WHERE (head BETWEEN 'A1' AND 'DF') ORDER BY head;
+#
+# Populate JIS-X-0208 range
+# Expected valid codes:  [A1..FE][A1..FE]   (94x94=8836 characters)
+#
+INSERT INTO t1 (code) SELECT concat(head, tail)
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (head NOT BETWEEN '8E' AND '8F')
+AND (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+#
+# Populate JIS-X-0212 range
+# Expected valid codes [8F][A1..FE][A1..FE] (1x94x94=8836 characters)
+#
+INSERT INTO t1 (code) SELECT concat('8F', head, tail)
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+DROP TEMPORARY TABLE head, tail;
+SHOW CREATE TABLE t1;
+
+UPDATE t1 SET a=unhex(code) ORDER BY code;
+SELECT COUNT(*) FROM t1;
+SELECT COUNT(*) FROM t1 WHERE a<>'';
+SELECT COUNT(*) FROM t1 WHERE a<>'' AND OCTET_LENGTH(a)=2;
+SELECT * FROM t1 WHERE CHAR_LENGTH(a)=2;
+SELECT COUNT(*) FROM t1 WHERE a<>'' AND OCTET_LENGTH(a)=3;
+
+
+#
+# Display all characters that have upper or lower case mapping.
+#
+SELECT code, hex(upper(a)), hex(lower(a)),a, upper(a), lower(a) FROM t1 WHERE hex(a)<>hex(upper(a)) OR hex(a)<>hex(lower(a)) ORDER BY code;
+#
+# Make sure all possible conversion happened
+#
+# Expect U+0122 LATIN CAPITAL LETTER G WITH CEDILLA
+#
+SELECT * FROM t1
+WHERE HEX(CAST(LOWER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(LOWER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+#
+# Expect U+00F0 LATIN SMALL LETTER ETH
+# Expect U+01F5 LATIN SMALL LETTER G WITH ACUTE
+#
+SELECT * FROM t1
+WHERE HEX(CAST(UPPER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(UPPER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+
+DROP TABLE t1;
+
+
+--echo #
+--echo # End of 5.5 tests
+--echo #

mysql-test/t/ctype_euckr.test

@@ -107,3 +107,68 @@ SELECT s, hex(a), hex(u), hex(a2) FROM t2 ORDER BY s;
 DROP TABLE t1, t2;
 
 --echo End of 5.4 tests
+
+
+
+--echo #
+--echo # Start of 5.5 tests
+--echo #
+
+--echo #
+--echo # Testing WL#4583 Case conversion in Asian character sets 
+--echo #
+#
+# Populate t1 with all hex digits
+#
+SET NAMES utf8;
+SET collation_connection=euckr_korean_ci;
+CREATE TABLE t1 (b VARCHAR(2));
+INSERT INTO t1 VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7');
+INSERT INTO t1 VALUES ('8'),('9'),('A'),('B'),('C'),('D'),('E'),('F');
+#
+# Populate tables head and tail with values '00'-'FF'
+#
+CREATE TEMPORARY TABLE head AS SELECT concat(b1.b, b2.b) AS head FROM t1 b1, t1 b2;
+CREATE TEMPORARY TABLE tail AS SELECT concat(b1.b, b2.b) AS tail FROM t1 b1, t1 b2;
+DROP TABLE t1;
+#
+# Populate table t1 with all codes [80..FF][20..FF]
+# Expected valid euckr codes: [81..FE][41..5A,61..7A,81..FE]
+#
+CREATE TABLE t1 AS
+SELECT concat(head, tail) AS code, ' ' AS a
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+DROP TEMPORARY TABLE head, tail;
+SHOW CREATE TABLE t1;
+UPDATE t1 SET a=unhex(code) ORDER BY code;
+SELECT COUNT(*) FROM t1 WHERE a<>'';
+#
+# Display all characters that have upper or lower case mapping.
+#
+SELECT code, hex(upper(a)), hex(lower(a)),a, upper(a), lower(a) FROM t1 WHERE hex(a)<>hex(upper(a)) OR hex(a)<>hex(lower(a));
+#
+# Make sure all possible conversion happened
+#
+# Expect U+212B ANGSTROM SIGN
+#
+SELECT * FROM t1
+WHERE HEX(CAST(LOWER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(LOWER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+#
+# Expect U+0111 LATIN SMALL LETTER D WITH STROKE
+# Expect U+24D0 to U+24E9 CIRCLED LATIN SMALL LETTER A to Z
+#
+SELECT * FROM t1
+WHERE HEX(CAST(UPPER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(UPPER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+
+DROP TABLE t1;
+
+
+
+
+--echo #
+--echo # End of 5.5 tests
+--echo #

mysql-test/t/ctype_gb2312.test

@@ -33,3 +33,94 @@ SELECT hex(a) FROM t1 ORDER BY a;
 DROP TABLE t1;
 
 # End of 4.1 tests
+
+
+--echo #
+--echo # Start of 5.5 tests
+--echo #
+
+--echo #
+--echo # Testing WL#4583 Case conversion in Asian character sets 
+--echo #
+#
+# Populate t1 with all hex digits
+#
+SET NAMES utf8;
+SET collation_connection=gb2312_chinese_ci;
+CREATE TABLE t1 (b VARCHAR(2));
+INSERT INTO t1 VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7');
+INSERT INTO t1 VALUES ('8'),('9'),('A'),('B'),('C'),('D'),('E'),('F');
+#
+# Populate tables head and tail with values '00'-'FF'
+#
+CREATE TEMPORARY TABLE head AS SELECT concat(b1.b, b2.b) AS head FROM t1 b1, t1 b2;
+CREATE TEMPORARY TABLE tail AS SELECT concat(b1.b, b2.b) AS tail FROM t1 b1, t1 b2;
+DROP TABLE t1;
+#
+# Populate table t1 with all codes [80..FF][20..FF]
+# Expected valid gb2312 codes [A1..F7][A1..FE]
+#
+CREATE TABLE t1 AS
+SELECT concat(head, tail) AS code, ' ' AS a
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+DROP TEMPORARY TABLE head, tail;
+SHOW CREATE TABLE t1;
+UPDATE t1 SET a=unhex(code) ORDER BY code;
+SELECT COUNT(*) FROM t1 WHERE a<>'';
+#
+# Display all characters that have upper or lower case mapping.
+#
+SELECT code, hex(upper(a)), hex(lower(a)),a, upper(a), lower(a) FROM t1 WHERE hex(a)<>hex(upper(a)) OR hex(a)<>hex(lower(a));
+#
+# Make sure all possible conversion happened
+#
+# Expect U+2160 to U+216B ROMAN NUMERAL ONE to ROMAN NUMERAL TWELVE
+#
+SELECT * FROM t1
+WHERE HEX(CAST(LOWER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(LOWER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+#
+# Expect
+#       U+00E0 LATIN SMALL LETTER A WITH GRAVE
+#       U+00E1 LATIN SMALL LETTER A WITH ACUTE
+#       U+00E8 LATIN SMALL LETTER E WITH GRAVE
+#       U+00E9 LATIN SMALL LETTER E WITH ACUTE
+#       U+00EA LATIN SMALL LETTER E WITH CIRCUMFLEX
+#       U+00EC LATIN SMALL LETTER I WITH GRAVE
+#       U+00ED LATIN SMALL LETTER I WITH ACUTE
+#       U+00F2 LATIN SMALL LETTER O WITH GRAVE
+#       U+00F3 LATIN SMALL LETTER O WITH ACUTE
+#       U+00F9 LATIN SMALL LETTER U WITH GRAVE
+#       U+00FA LATIN SMALL LETTER U WITH ACUTE
+#       U+00FC LATIN SMALL LETTER U WITH DIAERESIS
+#       U+0101 LATIN SMALL LETTER A WITH MACRON
+#       U+0113 LATIN SMALL LETTER E WITH MACRON
+#       U+011B LATIN SMALL LETTER E WITH CARON
+#       U+012B LATIN SMALL LETTER I WITH MACRON
+#       U+0144 LATIN SMALL LETTER N WITH ACUTE
+#       U+0148 LATIN SMALL LETTER N WITH CARON
+#       U+014D LATIN SMALL LETTER O WITH MACRON
+#       U+016B LATIN SMALL LETTER U WITH MACRON
+#       U+01CE LATIN SMALL LETTER A WITH CARON
+#       U+01D0 LATIN SMALL LETTER I WITH CARON
+#       U+01D2 LATIN SMALL LETTER O WITH CARON
+#       U+01D4 LATIN SMALL LETTER U WITH CARON
+#       U+01D6 LATIN SMALL LETTER U WITH DIAERESIS AND MACRON
+#       U+01D8 LATIN SMALL LETTER U WITH DIAERESIS AND ACUTE
+#       U+01DA LATIN SMALL LETTER U WITH DIAERESIS AND CARON
+#       U+01DC LATIN SMALL LETTER U WITH DIAERESIS AND GRAVE
+#
+SELECT * FROM t1
+WHERE HEX(CAST(UPPER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(UPPER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+
+DROP TABLE t1;
+
+
+
+
+--echo #
+--echo # End of 5.5 tests
+--echo #

mysql-test/t/ctype_gbk.test

@@ -68,3 +68,94 @@ SELECT b FROM t1 GROUP BY 1 LIMIT 1 INTO @nullll;
 DROP TABLES t1;
 
 --echo End of 5.0 tests
+
+
+--echo #
+--echo # Start of 5.5 tests
+--echo #
+
+--echo #
+--echo # Testing WL#4583 Case conversion in Asian character sets 
+--echo #
+#
+# Populate t1 with all hex digits
+#
+SET NAMES utf8;
+SET collation_connection=gbk_chinese_ci;
+CREATE TABLE t1 (b VARCHAR(2));
+INSERT INTO t1 VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7');
+INSERT INTO t1 VALUES ('8'),('9'),('A'),('B'),('C'),('D'),('E'),('F');
+#
+# Populate tables head and tail with values '00'-'FF'
+#
+CREATE TEMPORARY TABLE head AS SELECT concat(b1.b, b2.b) AS head FROM t1 b1, t1 b2;
+CREATE TEMPORARY TABLE tail AS SELECT concat(b1.b, b2.b) AS tail FROM t1 b1, t1 b2;
+DROP TABLE t1;
+#
+# Populate table t1 with all codes [80..FF][20..FF]
+# Expected valid gbk codes [81..FE][40..7E,80..FE]
+#
+CREATE TABLE t1 AS
+SELECT concat(head, tail) AS code, ' ' AS a
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+DROP TEMPORARY TABLE head, tail;
+SHOW CREATE TABLE t1;
+UPDATE t1 SET a=unhex(code) ORDER BY code;
+SELECT COUNT(*) FROM t1 WHERE a<>'';
+#
+# Display all characters that have upper or lower case mapping.
+#
+SELECT code, hex(upper(a)), hex(lower(a)),a, upper(a), lower(a) FROM t1 WHERE hex(a)<>hex(upper(a)) OR hex(a)<>hex(lower(a));
+#
+# Make sure all possible conversion happened
+#
+# Expect U+216A to U+216B ROMAN NUMERAL ELEVEN to ROMAN TWELVE
+#
+SELECT * FROM t1
+WHERE HEX(CAST(LOWER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(LOWER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+#
+# Expect
+#       U+00E0 LATIN SMALL LETTER A WITH GRAVE
+#       U+00E1 LATIN SMALL LETTER A WITH ACUTE
+#       U+00E8 LATIN SMALL LETTER E WITH GRAVE
+#       U+00E9 LATIN SMALL LETTER E WITH ACUTE
+#       U+00EA LATIN SMALL LETTER E WITH CIRCUMFLEX
+#       U+00EC LATIN SMALL LETTER I WITH GRAVE
+#       U+00ED LATIN SMALL LETTER I WITH ACUTE
+#       U+00F2 LATIN SMALL LETTER O WITH GRAVE
+#       U+00F3 LATIN SMALL LETTER O WITH ACUTE
+#       U+00F9 LATIN SMALL LETTER U WITH GRAVE
+#       U+00FA LATIN SMALL LETTER U WITH ACUTE
+#       U+00FC LATIN SMALL LETTER U WITH DIAERESIS
+#       U+0101 LATIN SMALL LETTER A WITH MACRON
+#       U+0113 LATIN SMALL LETTER E WITH MACRON
+#       U+011B LATIN SMALL LETTER E WITH CARON
+#       U+012B LATIN SMALL LETTER I WITH MACRON
+#       U+0144 LATIN SMALL LETTER N WITH ACUTE
+#       U+0148 LATIN SMALL LETTER N WITH CARON
+#       U+014D LATIN SMALL LETTER O WITH MACRON
+#       U+016B LATIN SMALL LETTER U WITH MACRON
+#       U+01CE LATIN SMALL LETTER A WITH CARON
+#       U+01D0 LATIN SMALL LETTER I WITH CARON
+#       U+01D2 LATIN SMALL LETTER O WITH CARON
+#       U+01D4 LATIN SMALL LETTER U WITH CARON
+#       U+01D6 LATIN SMALL LETTER U WITH DIAERESIS AND MACRON
+#       U+01D8 LATIN SMALL LETTER U WITH DIAERESIS AND ACUTE
+#       U+01DA LATIN SMALL LETTER U WITH DIAERESIS AND CARON
+#       U+01DC LATIN SMALL LETTER U WITH DIAERESIS AND GRAVE
+#
+SELECT * FROM t1
+WHERE HEX(CAST(UPPER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(UPPER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+
+DROP TABLE t1;
+
+
+
+
+--echo #
+--echo # End of 5.5 tests
+--echo #

mysql-test/t/ctype_sjis.test

@@ -93,3 +93,78 @@ SELECT hex(a), hex(lower(a)), hex(upper(a)) FROM t1 ORDER BY binary(a);
 DROP TABLE t1;
 
 --echo # End of 5.1 tests
+
+
+--echo #
+--echo # Start of 5.5 tests
+--echo #
+
+--echo #
+--echo # Testing WL#4583 Case conversion in Asian character sets 
+--echo #
+#
+# Populate t1 with all hex digits
+#
+SET NAMES utf8;
+SET collation_connection=sjis_japanese_ci;
+CREATE TABLE t1 (b VARCHAR(2));
+INSERT INTO t1 VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7');
+INSERT INTO t1 VALUES ('8'),('9'),('A'),('B'),('C'),('D'),('E'),('F');
+#
+# Populate tables head and tail with values '00'-'FF'
+#
+CREATE TEMPORARY TABLE head AS SELECT concat(b1.b, b2.b) AS head FROM t1 b1, t1 b2;
+CREATE TEMPORARY TABLE tail AS SELECT concat(b1.b, b2.b) AS tail FROM t1 b1, t1 b2;
+DROP TABLE t1;
+#
+# Populate table t1 with all codes [80..FF][20..FF]
+# excluding Half Width Kana [A1..DF]
+# Expected valid sjis codes:
+# [81..9F,E0..FC][40..7E,80..fC]  (60x188=11280 characters)
+#
+CREATE TABLE t1 AS
+SELECT concat(head, tail) AS code, ' ' AS a
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (head NOT BETWEEN 'A1' AND 'DF')
+AND   (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+#
+# Populate Half Width Kana: [A1..DF]
+# 
+INSERT t1 (code) SELECT head FROM head WHERE (head BETWEEN 'A1' AND 'DF');
+DROP TEMPORARY TABLE head, tail;
+SHOW CREATE TABLE t1;
+UPDATE t1 SET a=unhex(code) ORDER BY code;
+SELECT COUNT(*) FROM t1;
+SELECT COUNT(*) FROM t1 WHERE a<>'' AND OCTET_LENGTH(a)=1;
+SELECT COUNT(*) FROM t1 WHERE a<>'' AND OCTET_LENGTH(a)=2;
+#
+# Display all characters that have upper or lower case mapping.
+#
+SELECT code, hex(upper(a)), hex(lower(a)),a, upper(a), lower(a)
+FROM t1
+WHERE hex(a)<>hex(upper(a)) OR hex(a)<>hex(lower(a))
+ORDER BY code;
+#
+# Make sure all possible conversion happened
+#
+# Expect U+212B ANGSTROM SIGN
+#
+SELECT * FROM t1
+WHERE HEX(CAST(LOWER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(LOWER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+#
+# Expect no results
+#
+SELECT * FROM t1
+WHERE HEX(CAST(UPPER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(UPPER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+
+DROP TABLE t1;
+
+
+
+
+--echo #
+--echo # End of 5.5 tests
+--echo #

mysql-test/t/ctype_ujis.test

@@ -1212,3 +1212,88 @@ DROP TABLE t2;
 set names default;
 set character_set_database=default;
 set character_set_server=default;
+
+
+--echo #
+--echo # Start of 5.5 tests
+--echo #
+
+--echo #
+--echo # Testing WL#4583 Case conversion in Asian character sets 
+--echo #
+#
+# Populate t1 with all hex digits
+#
+SET NAMES utf8;
+SET collation_connection=ujis_japanese_ci;
+CREATE TABLE t1 (b VARCHAR(2));
+INSERT INTO t1 VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7');
+INSERT INTO t1 VALUES ('8'),('9'),('A'),('B'),('C'),('D'),('E'),('F');
+#
+# Populate tables head and tail with values '00'-'FF'
+#
+CREATE TEMPORARY TABLE head AS SELECT concat(b1.b, b2.b) AS head FROM t1 b1, t1 b2;
+CREATE TEMPORARY TABLE tail AS SELECT concat(b1.b, b2.b) AS tail FROM t1 b1, t1 b2;
+DROP TABLE t1;
+#
+# Populate table t1 with all ujis codes.
+# 
+#
+CREATE TABLE t1 AS SELECT 'XXXXXX' AS code, ' ' AS a LIMIT 0;
+#
+# Pupulate JIS-X-0201 range (Half Width Kana)
+# Valid characters: [8E][A1-DF]
+#
+INSERT INTO t1 (code) SELECT concat('8E', head) FROM head
+WHERE (head BETWEEN 'A1' AND 'DF') ORDER BY head;
+#
+# Populate JIS-X-0208 range
+# Expected valid range: [A1..FE][A1..FE]
+#
+INSERT INTO t1 (code) SELECT concat(head, tail)
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (head NOT BETWEEN '8E' AND '8F')
+AND   (tail BETWEEN '20' AND 'FF')
+ORDER BY head, tail;
+#
+# Populate JIS-X-0212 range
+# Expected valid range: [8F][A1..FE][A1..FE]
+#
+INSERT INTO t1 (code) SELECT concat('8F', head, tail)
+FROM head, tail
+WHERE (head BETWEEN '80' AND 'FF') AND (tail BETWEEN '80' AND 'FF')
+ORDER BY head, tail;
+DROP TEMPORARY TABLE head, tail;
+SHOW CREATE TABLE t1;
+
+UPDATE t1 SET a=unhex(code) ORDER BY code;
+SELECT COUNT(*) FROM t1;
+SELECT COUNT(*) FROM t1 WHERE a<>'';
+SELECT COUNT(*) FROM t1 WHERE a<>'' AND OCTET_LENGTH(a)=2;
+SELECT COUNT(*) FROM t1 WHERE a<>'' AND OCTET_LENGTH(a)=3;
+#
+# Display all characters that have upper or lower case mapping.
+#
+SELECT code, hex(upper(a)), hex(lower(a)),a, upper(a), lower(a) FROM t1 WHERE hex(a)<>hex(upper(a)) OR hex(a)<>hex(lower(a)) ORDER BY code;
+#
+# Make sure all possible conversion happened
+#
+# Expect U+0122 LATIN CAPITAL LETTER G WITH CEDILLA
+#
+SELECT * FROM t1
+WHERE HEX(CAST(LOWER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(LOWER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+#
+# U+00F0 LATIN SMALL LETTER ETH
+# U+01F5 LATIN SMALL LETTER G WITH ACUTE
+#
+SELECT * FROM t1
+WHERE HEX(CAST(UPPER(a) AS CHAR CHARACTER SET utf8)) <>
+      HEX(UPPER(CAST(a AS CHAR CHARACTER SET utf8))) ORDER BY code;
+
+DROP TABLE t1;
+
+
+--echo #
+--echo # End of 5.5 tests
+--echo #

sql/mysqld.cc

@@ -3337,6 +3337,16 @@ static int init_common_variables()
   strmake(pidfile_name, default_logfile_name, sizeof(pidfile_name)-5);
   strmov(fn_ext(pidfile_name),".pid");		// Add proper extension
 
+  /*
+    The default-storage-engine entry in my_long_options should have a
+    non-null default value. It was earlier intialized as
+    (longlong)"MyISAM" in my_long_options but this triggered a
+    compiler error in the Sun Studio 12 compiler. As a work-around we
+    set the def_value member to 0 in my_long_options and initialize it
+    to the correct value here.
+  */
+  default_storage_engine="MyISAM";
+
   /*
     Add server status variables to the dynamic list of
     status variables that is shown by SHOW STATUS.
@@ -5954,9 +5964,12 @@ struct my_option my_long_options[]=
   {"default-collation", 0, "Set the default collation (deprecated option, use --collation-server instead).",
    (uchar**) &default_collation_name, (uchar**) &default_collation_name,
    0, GET_STR, REQUIRED_ARG, 0, 0, 0, 0, 0, 0 },
+  /* default-storage-engine should have "MyISAM" as def_value. Instead
+     of initializing it here it is done in init_common_variables() due
+     to a compiler bug in Sun Studio compiler. */
   {"default-storage-engine", 0, "The default storage engine for new tables",
    (uchar**) &default_storage_engine, 0, 0, GET_STR, REQUIRED_ARG,
-   (longlong)"MyISAM", 0, 0, 0, 0, 0 },
+   0, 0, 0, 0, 0, 0 },
   {"default-time-zone", 0, "Set the default time zone.",
    (uchar**) &default_tz_name, (uchar**) &default_tz_name,
    0, GET_STR, REQUIRED_ARG, 0, 0, 0, 0, 0, 0 },

sql/share/errmsg.txt

@@ -6235,8 +6235,6 @@ ER_BINLOG_UNSAFE_NONTRANS_AFTER_TRANS
 ER_MESSAGE_AND_STATEMENT
   eng "%s Statement: %s"
 
-ER_SLAVE_IGNORE_SERVER_IDS
-  eng "The requested server id %d clashes with the slave startup option --replicate-same-server-id"
 ER_DEBUG_SYNC_TIMEOUT
   eng "debug sync point wait timed out"
   ger "Debug Sync Point Wartezeit berschritten"

strings/ctype-mb.c

@@ -62,6 +62,17 @@ size_t my_casedn_str_mb(CHARSET_INFO * cs, char *str)
 }
 
 
+static inline MY_UNICASE_INFO*
+get_case_info_for_ch(CHARSET_INFO *cs, uint page, uint offs)
+{
+  MY_UNICASE_INFO *p;
+  return cs->caseinfo ? ((p= cs->caseinfo[page]) ? &p[offs] : NULL) :  NULL;
+}
+
+
+/*
+  For character sets which don't change octet length in case conversion.
+*/
 size_t my_caseup_mb(CHARSET_INFO * cs, char *src, size_t srclen,
                     char *dst __attribute__((unused)),
                     size_t dstlen __attribute__((unused)))
@@ -70,11 +81,23 @@ size_t my_caseup_mb(CHARSET_INFO * cs, char *src, size_t srclen,
   register char *srcend= src + srclen;
   register uchar *map= cs->to_upper;
 
+  DBUG_ASSERT(cs->caseup_multiply == 1);
   DBUG_ASSERT(src == dst && srclen == dstlen);
+  DBUG_ASSERT(cs->mbmaxlen == 2);
+  
   while (src < srcend)
   {
     if ((l=my_ismbchar(cs, src, srcend)))
+    {
+      MY_UNICASE_INFO *ch;
+      if ((ch= get_case_info_for_ch(cs, (uchar) src[0], (uchar) src[1])))
+      {
+        *src++= ch->toupper >> 8;
+        *src++= ch->toupper & 0xFF;
+      }
+      else
         src+= l;
+    }
     else 
     {
       *src=(char) map[(uchar) *src];
@@ -93,11 +116,23 @@ size_t my_casedn_mb(CHARSET_INFO * cs, char *src, size_t srclen,
   register char *srcend= src + srclen;
   register uchar *map=cs->to_lower;
 
+  DBUG_ASSERT(cs->casedn_multiply == 1);
   DBUG_ASSERT(src == dst && srclen == dstlen);  
+  DBUG_ASSERT(cs->mbmaxlen == 2);
+  
   while (src < srcend)
   {
     if ((l= my_ismbchar(cs, src, srcend)))
+    {
+      MY_UNICASE_INFO *ch;
+      if ((ch= get_case_info_for_ch(cs, (uchar) src[0], (uchar) src[1])))
+      {
+        *src++= ch->tolower >> 8;
+        *src++= ch->tolower & 0xFF;
+      }
+      else
         src+= l;
+    }
     else
     {
       *src= (char) map[(uchar)*src];
@@ -108,6 +143,75 @@ size_t my_casedn_mb(CHARSET_INFO * cs, char *src, size_t srclen,
 }
 
 
+/*
+  Case folding functions for character set
+  where case conversion can change string octet length.
+  For example, in EUCKR,
+    _euckr 0xA9A5 == "LATIN LETTER DOTLESS I" (Turkish letter)
+  is upper-cased to to
+    _euckr 0x49 "LATIN CAPITAL LETTER I"  ('usual' letter I)
+  Length is reduced in this example from two bytes to one byte.
+*/
+static size_t
+my_casefold_mb_varlen(CHARSET_INFO *cs,
+                      char *src, size_t srclen,
+                      char *dst, size_t dstlen __attribute__((unused)),
+                      uchar *map,
+                      size_t is_upper)
+{
+  char *srcend= src + srclen, *dst0= dst;
+
+  DBUG_ASSERT(cs->mbmaxlen == 2);
+
+  while (src < srcend)
+  {
+    size_t mblen= my_ismbchar(cs, src, srcend);
+    if (mblen)
+    {
+      MY_UNICASE_INFO *ch;
+      if ((ch= get_case_info_for_ch(cs, (uchar) src[0], (uchar) src[1])))
+      {
+        int code= is_upper ? ch->toupper : ch->tolower;
+        src+= 2;
+        if (code > 0xFF)
+          *dst++= code >> 8;
+        *dst++= code & 0xFF;
+      }
+      else
+      {
+        *dst++= *src++;
+        *dst++= *src++;
+      }
+    }
+    else
+    {
+      *dst++= (char) map[(uchar) *src++];
+    }
+  }
+  return (size_t) (dst - dst0);
+}
+
+
+size_t
+my_casedn_mb_varlen(CHARSET_INFO * cs, char *src, size_t srclen,
+                    char *dst, size_t dstlen)
+{
+  DBUG_ASSERT(dstlen >= srclen * cs->casedn_multiply); 
+  DBUG_ASSERT(src != dst || cs->casedn_multiply == 1);
+  return my_casefold_mb_varlen(cs, src, srclen, dst, dstlen, cs->to_lower, 0);
+}
+
+
+size_t
+my_caseup_mb_varlen(CHARSET_INFO * cs, char *src, size_t srclen,
+                    char *dst, size_t dstlen)
+{
+  DBUG_ASSERT(dstlen >= srclen * cs->caseup_multiply);
+  DBUG_ASSERT(src != dst || cs->caseup_multiply == 1);
+  return my_casefold_mb_varlen(cs, src, srclen, dst, dstlen, cs->to_upper, 1);
+}
+
+
 /*
   my_strcasecmp_mb() returns 0 if strings are equal, non-zero otherwise.
  */