Commit graph

1362 commits

Author SHA1 Message Date
Alexander Barkov
e7ff281d2e MDEV-6353 my_ismbchar() and my_mbcharlen() refactoring 2016-05-17 15:27:10 +04:00
Alexander Barkov
d516a2ae0c MDEV-9823 LOAD DATA INFILE silently truncates incomplete byte sequences 2016-04-06 09:13:49 +04:00
Alexander Barkov
38f39a9288 Updating conf_to_src.c according to the current CHARSET_INFO structure. 2016-04-05 09:35:27 +04:00
Alexander Barkov
1d73005bf3 MDEV-8360 Clean-up CHARSET_INFO: strnncollsp: diff_if_only_endspace_difference
- Removing the "diff_if_only_endspace_difference" argument from
  MY_COLLATION_HANDLER::strnncollsp(), my_strnncollsp_simple(),
  as well as in the function template MY_FUNCTION_NAME(strnncollsp)
  in strcoll.ic

- Removing the "diff_if_only_space_different" from ha_compare_text(),
  hp_rec_key_cmp().

- Adding a new function my_strnncollsp_padspace_bin() and reusing
  it instead of duplicate code pieces in my_strnncollsp_8bit_bin(),
  my_strnncollsp_latin1_de(), my_strnncollsp_tis620(),
  my_strnncollsp_utf8_cs().

- Adding more tests for better coverage of the trailing space handling.

- Removing the unused definition of HA_END_SPACE_ARE_EQUAL
2016-03-31 11:04:48 +04:00
Alexander Barkov
3be95ee061 Removing my_strnncoll_mb_bin() and my_strnncollsp_mb_bin(),
as they are not used any more.
We now use function templates from strcoll.ic instead.
2016-03-25 06:42:44 +04:00
Sergei Golubchik
f67a2211ec Merge branch '10.1' into 10.2 2016-03-23 22:36:46 +01:00
Alexander Barkov
0a83caf4a9 Removing duplicate code: sharing implementation of "strnxfrm"
between gbk_chinese_ci and big5_chinese_ci.
2016-03-23 14:26:43 +04:00
Alexander Barkov
0d42d4e9e0 Removing unused code in ctype-bin5.c 2016-03-23 12:44:31 +04:00
Alexander Barkov
1170d23ca3 Fixing compilation warnings introduced in:
> commit e09299511e
> Author: Alexander Barkov <bar@mariadb.org>
> Date:   Wed Mar 16 10:55:12 2016 +0400
>
>     MDEV-9665 Remove cs->cset->ismbchar()
>     Using a more powerfull cs->cset->charlen() instead.
2016-03-23 12:37:19 +04:00
Sergei Golubchik
3b0c7ac1f9 Merge branch '10.0' into 10.1 2016-03-21 13:02:53 +01:00
Alexander Barkov
e09299511e MDEV-9665 Remove cs->cset->ismbchar()
Using a more powerfull cs->cset->charlen() instead.
2016-03-16 10:55:12 +04:00
Otto Kekäläinen
1777fd5f55 Fix spelling: occurred, execute, which etc 2016-03-04 02:09:37 +02:00
Sergei Golubchik
00d1db7a38 Merge branch '10.1' into 10.2 2016-02-25 18:19:55 +01:00
Alexander Barkov
ff2d92b17d MDEV-7231 Field ROUTINE_DEFINITION in INFORMATION_SCHEMA.ROUTINES
contains broken procedure body when used shielding quotes inside.
2016-02-24 13:12:03 +04:00
Monty
d80b8442a6 Fixes needed to compile with musl C library
Patch originally by Codarren Velvindron
2016-02-07 15:00:30 +02:00
Alexander Barkov
98b6026036 Adding "const" qualifier to the argument of decimal_actual_fraction(). 2016-01-14 17:23:23 +04:00
Sergei Golubchik
a2bcee626d Merge branch '10.0' into 10.1 2015-12-21 21:24:22 +01:00
Alexander Barkov
82bec8bfdf MDEV-9265 SuSE patches: Suspicious implicit sign extension 2015-12-15 11:04:51 +04:00
Alexander Barkov
af3c67056d MDEV-9265 SuSE patches: Suspicious implicit sign extension 2015-12-15 10:57:28 +04:00
Sergei Golubchik
1623995158 Merge branch '5.5' into 10.0 2015-12-13 00:10:40 +01:00
Oleksandr Byelkin
fa25921b59 MDEV-8407 Numeric errors, server crash with COLUMN_JSON() on DECIMAL with precision > 40
In fact it was error in decimal library (incorrect processing of buffer overflow) invisible from other server parts because of buffer allocation and precision tests.
2015-12-10 11:22:53 +01:00
Jan Lindström
d85168e40d Correct length check in my_wc_mb_filename() 2015-12-07 09:20:31 +02:00
Alexander Barkov
310c718cff MDEV-9178 Wrong result for CAST(CONVERT('1IJ3' USING ucs2) AS SIGNED)
Also, fixing compilation warnings in ctype-mb.ic (Windows).
2015-11-24 22:47:42 +04:00
Sergei Golubchik
dfb74dea30 Merge branch '10.0' into 10.1 2015-10-12 00:37:58 +02:00
Alexander Barkov
d9b25ae3db MDEV-8466 CAST works differently for DECIMAL/INT vs DOUBLE for empty strings
MDEV-8468 CAST and INSERT work differently for DECIMAL/INT vs DOUBLE for a string with trailing spaces
2015-09-17 11:05:07 +04:00
Oleksandr Byelkin
0ce0b88080 MDEV-8450: PATCH] Wrong macro expansion in Query_cache::send_result_to_client()
Expression in macro protected by ()
2015-09-06 22:26:33 +02:00
Alexander Barkov
3ba2a958be MDEV-8694 Wrong result for SELECT..WHERE a NOT LIKE 'a ' AND a='a'
Note, the patch for MDEV-8661 unintentionally fixed MDEV-8694 as well,
as a side effect. Adding a real clear fix: implementing
Item_func_like::propagate_equal_fields() with comments.
2015-08-28 17:03:09 +04:00
Alexander Barkov
78b80cb6ba Adding MY_CHARSET_HANDLER::native_to_mb().
This is a pre-requisite patch for:
- MDEV-8433 Make field<'broken-string' use indexes
- MDEV-8625 Bad result set with ignorable characters when using a prefix key
- MDEV-8626 Bad result set with contractions when using a prefix key
2015-08-14 18:34:41 +04:00
Alexander Barkov
75931feabe MDEV-8362 dash '-' is not recognized in charset armscii8 on select where query 2015-07-14 12:00:05 +04:00
Alexander Barkov
e4f8cea356 MDEV-8419 utf32: compare broken bytes as "greater than any non-broken character" 2015-07-07 09:15:58 +04:00
Alexander Barkov
3a606ba210 Fixing a bug in MDEV-8418 (utf16, utf16le) and MDEV-8417 (utf8mb4).
Fixing non-BMP characters to have the same weight, as it was before
MDEV-8418 and MDEV-8417.
2015-07-06 18:59:33 +04:00
Alexander Barkov
b2e324a21f MDEV-8416 ucs2: compare broken bytes as "greater than any non-broken character"
MDEV-8418 utf16: compare broken bytes as "greater than any non-broken character"
2015-07-06 15:50:56 +04:00
Alexander Barkov
35d8ac350d MDEV-8417 utf8mb4: compare broken bytes as "greater than any non-broken character" 2015-07-06 10:47:39 +04:00
Alexander Barkov
7ab7f5327a Fixing a typo in the previous commit. 2015-07-03 19:08:18 +04:00
Alexander Barkov
fff30e26c3 Adding UTF8 related macros to reduce duplicate code. 2015-07-03 18:40:04 +04:00
Alexander Barkov
aeb8d713f4 Removing unused code in ctype-utf8.c 2015-07-03 17:30:15 +04:00
Alexander Barkov
9ad8ff666c MDEV-8415 utf8: compare broken bytes as "greater than any non-broken character" 2015-07-03 17:24:16 +04:00
Alexander Barkov
95d07ee408 MDEV-8215 Asian MB3 charsets: compare broken bytes as "greater than any non-broken character" 2015-07-03 10:33:17 +04:00
Sergei Golubchik
658992699b Merge tag 'mariadb-10.0.20' into 10.1 2015-06-27 20:35:26 +02:00
Alexander Barkov
4f828a1cac MDEV-8214 Asian MB2 charsets: compare broken bytes as "greater than any non-broken character" 2015-06-26 13:40:28 +04:00
Sergei Golubchik
36bf482db1 MDEV-8285 compile fails under Mac OS X 10.6.8 due to use of strnlen
#include <m_string.h> where strnlen() is used
2015-06-14 15:51:34 +02:00
Sergei Golubchik
5091a4ba75 Merge tag 'mariadb-10.0.19' into 10.1 2015-06-01 15:51:25 +02:00
Sergei Golubchik
49c853fb94 Merge branch '5.5' into 10.0 2015-05-04 22:00:24 +02:00
Sergei Golubchik
ae18a28500 MDEV-7973 bigint fail with gcc 5.0
-LONGLONG_MIN is the undefined behavior in C.
longlong2decimal() used to do this:

  int longlong2decimal(longlong from, decimal_t *to) {
    if ((to->sign= from < 0))
      return ull2dec(-from, to);
    return ull2dec(from, to);

and later in ull2dec() (DIG_BASE is 1000000000):

  static int ull2dec(ulonglong from, decimal_t *to) {
    for (intg1=1; from >= DIG_BASE; intg1++, from/=DIG_BASE) {}

this breaks in gcc-5 at -O3. Here ull2dec is inlined into
longlong2decimal. And gcc-5 believes that 'from' in the
inlined ull2dec is always a positive integer (indeed, if it was
negative, then -from was used instead). So gcc-5 uses
*signed* comparison with DIG_BASE.

Fix: make a special case for LONGLONG_MIN, don't negate it
2015-05-04 08:32:27 +02:00
Alexander Barkov
44d1e85fe5 MDEV-7649 wrong result when comparing utf8 column with an invalid literal 2015-04-24 11:00:34 +04:00
Alexander Barkov
ec68494beb MDEV-7677 my_charset_handler_filename has a wrong "ismbchar" member 2015-03-23 17:38:55 +04:00
Sergey Vojtovich
18e9c314e4 MDEV-6650 - LINT_INIT emits code in non-debug builds
Replaced all references to LINT_INIT with UNINIT_VAR and LINT_INIT_STRUCT.
Removed LINT_INIT macro.
2015-03-16 14:48:22 +04:00
Alexander Barkov
f48dc5ccc7 Moving the conversion code from String::well_formed_copy()
to my_convert_fix() - a new function in /strings.
2015-03-16 12:14:31 +04:00
Alexander Barkov
197afb413f MDEV-6566 Different INSERT behaviour on bad bytes with and without character set conversion 2015-03-13 16:51:36 +04:00
Alexander Barkov
a7ed8523e3 Adding a shared include file ctype-mb.ic and removing a number
of very similar copies of my_well_formed_len_xxx(), implemented
for big5, cp932, euckr, eucjpms, gb2312m gbk, sjis, ujis.
2015-03-04 09:16:43 +04:00