feat(tiamnu): hard code in defs.h (#1481)

[zzzz-vincent]

change magic number to readable const

Summary about this PR

Issue Number: close #1481

1. change some magic number to readable const variable
2. move some const value related to column and DPN from common/def.h to core/column_share.h

Tests Check List

• Unit test
• Integration test
• Manual test (add detailed scripts or steps below)
• No code

Changelog

• New Feature
• Bug Fix
• Performance Improvement
• Build/Testing/CI/CD
• Documentation
• Not for changelog (changelog entry is not required)

Documentation

• Affects user behaviors
• Contains syntax changes
• Contains variable changes
• Contains experimental features

[mergify]

Thanks for the contribution!
I have applied any labels matching special text in your PR Changelog.

Please review the labels and make any necessary changes.

[codecov]

Codecov Report

Patch coverage: 83.33% and project coverage change: -0.20 ⚠️

Comparison is base (8e197ae) 45.15% compared to head (41e4250) 44.96%.

❗ Current head 41e4250 differs from pull request most recent head 2e74d03. Consider uploading reports for the commit 2e74d03 to get more accurate results

Additional details and impacted files

@@                 Coverage Diff                 @@
##           stonedb-5.7-dev    #1533      +/-   ##
===================================================
- Coverage            45.15%   44.96%   -0.20%     
===================================================
  Files                 1999     1999              
  Lines               413931   413915      -16     
===================================================
- Hits                186906   186108     -798     
- Misses              227025   227807     +782     

Impacted Files	Coverage Δ
storage/tianmu/core/table_share.h	77.77% <ø> (ø)
storage/tianmu/core/column_share.cpp	85.60% <75.00%> (ø)
storage/tianmu/core/column_share.h	100.00% <100.00%> (ø)
storage/tianmu/core/tianmu_attr.cpp	69.55% <100.00%> (ø)

... and 293 files with indirect coverage changes

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[RingsC]

LGTM

[hustjieke]

LGTM, nice job!

[hustjieke]

ci/cd failed, in debian:

[ 19%] tianmu.alter_table_mix_use               w10 [ fail ]
        Test ended at 2023-04-09 03:59:47

CURRENT_TEST: tianmu.alter_table_mix_use
mysqltest: At line 82: query 'alter table alter_table_mix_test rename alter_table_rename_drop, drop c_longblob' failed: 1030: Got error 1 from storage engine

The result from queries just before the failure was:
< snip >
  `c_tinyint` tinyint(4) DEFAULT NULL COMMENT 'tinyint',
  `c_smallint` smallint(6) NOT NULL COMMENT 'smallint',
  `c_mediumint` mediumint(9) DEFAULT NULL COMMENT 'mediumint',
  `c_int` int(11) DEFAULT NULL COMMENT 'int',
  `c_bigint` bigint(20) DEFAULT NULL COMMENT 'bigint',
  `c_float` float DEFAULT NULL COMMENT 'float',
  `c_double` double DEFAULT NULL COMMENT 'double',
  `c_decimal` decimal(10,5) DEFAULT NULL COMMENT 'decimal',
  `c_date` date DEFAULT NULL COMMENT 'date',
  `c_datetime` datetime DEFAULT NULL COMMENT 'datetime',
  `c_timestamp` timestamp NULL DEFAULT NULL COMMENT 'timestamp',
  `c_time` time DEFAULT NULL COMMENT 'time',
  `c_char` char(10) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT 'char',
  `c_varchar` varchar(10) COLLATE utf8mb4_unicode_ci DEFAULT NULL COMMENT 'varchar',
  `c_blob` blob COMMENT 'blob',
  `c_text` text COLLATE utf8mb4_unicode_ci COMMENT 'text',
  `c_longblob` longblob COMMENT 'longblob'
) ENGINE=TIANMU DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci
insert into alter_table_mix_test(c_tinyint,c_smallint,c_mediumint,c_int,c_bigint,c_float,c_double,c_decimal,c_date,c_datetime,c_timestamp,c_time,c_char,c_varchar,c_text)
values(105,105,105,105,105,5.2,10.88,105.083,'2016-02-25','2016-02-25 10:20:01','2016-02-25 05:20:01','10:20:01','stoneatom1','hello1','bcdefghijklmn');
safe_process[46047]: Child process: 46048, exit: 1

 - the logfile can be found in '/stonedb57/install/mysql-test/var/log/tianmu.alter_table_mix_use/alter_table_mix_use.log'
[ 19%] tianmu.issue468                          w7 [ pass ]    216

In Centos7:

[ 61%] tianmu.alter_table1                      w1 [ fail ]
        Test ended at 2023-04-09 03:59:03

CURRENT_TEST: tianmu.alter_table1
mysqltest: At line 64: query 'alter table st2 order by deptId,id' failed: 6: An unknown system exception error caught.

The result from queries just before the failure was:
< snip >
salary FLOAT
);
insert into st2 values(3,'haha1',45,4.5);
insert into st2 values(1,'haha2',12,1.2);
insert into st2 values(4,'haha3',31,3.2);
insert into st2 values(2,'haha4',55,3.5);
alter table st2 order by id;
select * from st2;
id	name	deptId	salary
1	haha2	12	1.2
2	haha4	55	3.5
3	haha1	45	4.5
4	haha3	31	3.2
alter table st2 order by id,deptId;
select * from st2;
id	name	deptId	salary
1	haha2	12	1.2
2	haha4	55	3.5
3	haha1	45	4.5
4	haha3	31	3.2
safe_process[43892]: Child process: 43893, exit: 1

 - the logfile can be found in '/stonedb57/install/mysql-test/var/log/tianmu.alter_table1/alter_table1.log'
[ 61%] tianmu.std_test                          w5 [ pass ]   3744

It seems the changes caused some internal crashed. We should do further confirmation.
ping @zzzz-vincent.

[zzzz-vincent]

This is very interesting. I looked into these two tests but couldn't really get where the error came from. But posted some preliminary investigation here incase you may know what happened.

In debian, the alter_table_mix_use failed when executing alter rename and drop. In centos, the test alter_table1 failed due to the third time alter reorder the table.

I tested locally, and two other tests failed: tianmu.alter_table_primarykey and tianmu.alter_table_primarykey which were all involved alter. But after I tested multiple times, some runs got all passed. Not sure why this change cause non-deterministic failures.

And after a very quick check, I think at lest some failures were thrown from type_conversion_status Item::save_in_field_inner(Field *field, bool no_conversions) but haven't yet track why.

Will look into this soon.

[mergify]

This pull request has merge conflicts, you should resolve it before merged. @zzzz-vincent please update it :)

Try @mergify update or update manually.

[RingsC]

@zzzz-vincent , Pls, you need re-pull the latest code, and, we refined some files you changed, Thanks.