table, infoschema, executor: fix Charset/Collation shown in column desc

[bb7133]

What problem does this PR solve?

With the help of a parser PR(pingcap/parser#270), this request aims to fix:

1. Collate descriptions displayed in SHOW FULL COLUMNS statement. It will: close compatibility: “show full columns” prints wrong Collation for date, datatime, timestamp columns #9807, close Need to check and fix more types in show (full) columns #9837.

2. Some corner cases in information_schema.COLUMNS table, take the following statements as instance:

mysql> create table t1(a tinytext);
Query OK, 0 rows affected (0.08 sec)

In MySQL 5.7:

mysql> SELECT column_name, character_set_name, collation_name FROM information_schema.COLUMNS WHERE table_schema = "test" AND table_name = "t1";
+-------------+--------------------+--------------------+
| column_name | character_set_name | collation_name     |
+-------------+--------------------+--------------------+
| a           | utf8mb4            | utf8mb4_general_ci |
+-------------+--------------------+--------------------+
1 row in set (0.03 sec)

In TiDB master before this PR:

mysql> SELECT column_name, character_set_name, collation_name FROM information_schema.COLUMNS WHERE table_schema = "test" AND table_name = "t1";
+-------------+--------------------+----------------+
| column_name | character_set_name | collation_name |
+-------------+--------------------+----------------+
| a           | NULL               | NULL           |
+-------------+--------------------+----------------+
1 row in set (0.01 sec)

What is changed and how it works?

It used a HasCharset() function ported from MySQL source and use it to determine if given column's Charset/Collation should be displayed, following the way MySQL does:

sql_show.cc::store_column_type()

Check List

Tests

• Integration test

Code changes

• Has interface methods change

Side effects

• Increased code complexity

Related changes

• Need to cherry-pick to the release branch

[xiekeyi98]

1. TODO && FIXME, I think use one of them is enough.
2. Maybe we should delete or change the mysql-test.

[xiekeyi98]

/run-all-tests

[bb7133]

/run-all-tests tidb-test=pr/777

[codecov]

Codecov Report

Merging #10007 into master will decrease coverage by 0.0673%.
The diff coverage is 100%.

@@               Coverage Diff               @@
##            master     #10007        +/-   ##
===============================================
- Coverage   78.033%   77.9657%   -0.0674%     
===============================================
  Files          404        404                
  Lines        81996      81995         -1     
===============================================
- Hits         63984      63928        -56     
- Misses       13316      13354        +38     
- Partials      4696       4713        +17

[ngaut]

/run-all-tests tidb-test=pr/777

[bb7133]

/run-all-tests tidb-test=pr/777

[bb7133]

/run-all-tests tidb-test=pr/777

[xiekeyi98]

LGTM

[bb7133]

Addressed, thanks @zimulala

[zimulala]

LGTM

[zimulala]

/run-all-tests tidb-test=pr/777

[tshqin]

need a cherry pick to 2.1 branch