table: check non-BMP characters and return error when the charset is utf8 and sql mode is strict mode

[winkyao]

What problem does this PR solve?

Before this PR, TiDB don't handle the non-BMP character when the charset is utf8. Run the sqls in mysql:

mysql> create table t1(a varchar(100) charset utf8);
Query OK, 0 rows affected (0.01 sec)

mysql> insert into t1 values(x'f09f8c80');
ERROR 1366 (HY000): Incorrect string value: '\xF0\x9F\x8C\x80' for column 'a' at row 1

But in TiDB:

mysql> create table t1(a varchar(100) charset utf8);
Query OK, 0 rows affected (0.01 sec)

mysql> insert into t1 values(x'f09f8c80');
Query OK, 1 row affected (0.00 sec)

What is changed and how it works?

utf8 cannot handle the Unicode character length larger than 3.

Check List

Tests

• Unit test

Code changes

• Has exported function/method change

Side effects

• Breaking backward compatibility

Related changes

• Need to cherry-pick to the release branch
• Need to be included in the release note

---

This change is 

[winkyao]

@kennytm PTAL

[ciscoxll]

PR need to be cherry-pick to 2.1 and 2.0.

[kennytm]

LGTM

Reviewed 2 of 2 files at r1, 1 of 1 files at r2.
Reviewable status: complete! all files reviewed, all discussions resolved

[shenli]

@tiancaiamao @jackysp @lysu PTAL

[winkyao]

@ciscoxll This PR will break backward compatibility, I will not pick it to 2.0 and I'm considering to pick it to 2.1

[zz-jason]

@winkyao Please add some proper labels, for example, component, type, etc.

[lysu]

Reviewed 1 of 2 files at r1, 1 of 1 files at r2.
Reviewable status: complete! all files reviewed, all discussions resolved

[lysu]

maybe tikv also need check this too @breeswish

Reviewable status: complete! all files reviewed, all discussions resolved

[lysu]

/run-all-tests

[lysu]

Reviewed 1 of 1 files at r3.
Reviewable status: complete! all files reviewed, all discussions resolved

[winkyao]

/run-all-tests