CALL SUGGEST parameter to disable hidden min length validation

[donhardman]

Proposal:

Currently there is an issue when we use CALL SUGGEST with a phrase that is less than 3 characters; we get no results.

The suggestion is to add some parameters in case we really need to control this behaviour.

mysql> drop table if exists t;
Query OK, 0 rows affected (0.00 sec)

mysql> create table t(name text) min_infix_len='2';
Query OK, 0 rows affected (0.00 sec)

mysql> insert into t values(1, 'aa defghi xxx');
Query OK, 1 row affected (0.01 sec)

mysql> insert into t values(2, 'aa def');
Query OK, 1 row affected (0.00 sec)

mysql> select * from t where match('aa def ghi xxx') option fuzzy=1, layouts='';
Empty set (0.00 sec)

mysql> call suggest('aa', 't');
mysql> call suggest('def', 't');
+---------+----------+------+
| suggest | distance | docs |
+---------+----------+------+
| def     | 0        | 1    |
+---------+----------+------+
mysql> select * from t;
+------+---------------+
| id   | name          |
+------+---------------+
|    1 | aa defghi xxx |
|    2 | aa def        |
+------+---------------+
mysql>

Checklist:

^{To be completed by the assignee. Check off tasks that have been completed or are not applicable.}

• Implementation completed
• Tests developed
• Documentation updated
• Documentation reviewed
• Changelog updated
• OpenAPI YAML updated and issue created to rebuild clients

[sanikolaev]

This issue blocks this one manticoresoftware/manticoresearch-buddy#490

[tomatolog]

fixed at 01339cb

I set hardcoded word len limit to 2 and added test case there this case covered.

I also tried to add new min_word_len option to use like call suggest('aa', 't', 1 as min_word_len); with values 1,2,3 but the case 1 as min_word_len still returns empty result set as RAM chunk bloom filter can not properly handle such infixes and produce empty result set. That is why I dropped the patch with suggest option as hardcoded length check >=2 works ok now and to set this for lower value RT RAM chunk infix generation should be reworked first.