[Tech Request]: optimize reader to support filer on unsorted block

[XuPeng-SH]

Is there an existing issue for the same tech request?

• I have checked the existing issues.

Does this tech request not affect user experience?

• This tech request doesn't affect user experience.

What would you like to be added ?

optimize reader to support filer on unsorted block

Why is this needed ?

>>> create table t1 (id int not null auto_increment, unique key `id` (`id`));
>>> insert into t1 (select * from generate_series(1, 50000000, 1)g);
>>> explain select * from t1 where id=500;
+----------------------------------------------------------------------------------------------------------------------+
| QUERY PLAN                                                                                                           |
+----------------------------------------------------------------------------------------------------------------------+
| Project                                                                                                              |
|   ->  Join                                                                                                           |
|         Join Type: INNER                                                                                             |
|         Join Cond: (t1.__mo_fake_pk_col = __mo_index_unique_85292ace-a937-11ee-9895-acde48001122.__mo_index_pri_col) |
|         Runtime Filter Build: #[-1,0]                                                                                |
|         ->  Table Scan on test.t1                                                                                    |
|               Runtime Filter Probe: t1.__mo_fake_pk_col                                                              |
|         ->  Table Scan on test.__mo_index_unique_85292ace-a937-11ee-9895-acde48001122                                |
|               Filter Cond: (__mo_index_unique_85292ace-a937-11ee-9895-acde48001122.__mo_index_idx_col = 500)         |
|               Block Filter Cond: (__mo_index_unique_85292ace-a937-11ee-9895-acde48001122.__mo_index_idx_col = 500)   |
+----------------------------------------------------------------------------------------------------------------------+

t1 is never sorted by the fake key and right now the reader will not apply any input filter
we have to consider fake pk and cluster key:

1. no cluster key with pk
2. cluster key with fake pk

Additional information

No response