1. Fix auto analyze external table recursively load schema cache bug.
2. Move some function in StatisticsAutoAnalyzer class to TableIf. So that external table and internal table could implement the logic separately.
3. Disable external catalog auto analyze by default, could open it by adding catalog property "enable.auto.analyze"="true"
1. add scalar subquery's output to LogicalApply's output
2. for in and exists subquery's, add mark join slot into LogicalApply's output
3. forbid push down alias through join if the project list have any mark join slots.
4. move normalize aggregate rule to analysis phase
The reason is that sql cache just use partitionKey , latestVersion and latestTime to check if the cache should be returned, if we delete some partition(s) which is not the latest updated partition, all above values are not changed, so the cache will hit.
Use a field to save the partition num of these tables and sum the partition nums and send it to BE, there are two situations which contains delete-partition ops:
- just delete some partition(s), so the sum of partition num will be lower than before.
- delete some partition(s) coexists with add some partition(s), so the latest time or latest version will be higher than before.
Add a rule to optimize CASE WHEN expression.
Rewrite rule to convert CASE WHEN to IF.
For example:
CASE WHEN a > 1 THEN 1 ELSE 0 END -> IF(a > 1, 1, 0)
1. support struct data type
2. add array / map / struct literal syntax
3. fix array union / intersect / except type coercion
4. fix explict cast data type check for array
5. fix bound function type coercion
Support such sql grammar, Just for compatibility
```sql
select table_name
from information_schema.tables
where table_schema collate utf8_general_ci = 'information_schema'
and table_name collate utf8_general_ci = 'parameters';
```
#### `expr_depth_limit`
Default：3000
IsMutable：true
Limit on the depth of an expr tree. Exceed this limit may cause long analysis time while holding db read lock. Do not set this if you know what you are doing
#### `expr_children_limit`
Default：10000
IsMutable：true
Limit on the number of expr children of an expr tree. Exceed this limit may cause long analysis time while holding database read lock.
Fix bug that mistaken stats when analyzing table incrementally and partition number less than 512
Fix bug that cron expression lost during analyzing
Mark system job as running after registered to AnalysisManager to avoid submit same jobs if previous one take long time
```
CREATE ROW POLICY test_row_policy_1 ON test.table1
AS {RESTRICTIVE|PERMISSIVE} [TO user] [TO ROLE role] USING (id in (1, 2)); // add `to role`
DROP [ROW] POLICY [IF EXISTS] test_row_policy;//delete `for user` and `on table`
SHOW ROW POLICY [FOR user][FOR ROLE role] // add `for role`
```
We use two facilities to do predicate infer: PredicatePropagation and
PullUpPredicates. In the prvious implementation, we use a set to save
the intermediate result of PredicatePropagation. The purpose is infer
new predicate though two equal relation. However, it is the wrong way.
Because it could infer wrong predicate through outer join. For example
```sql
select a.c1
from a
left join b on a.c2 = b.c2 and a.c1 = '1'
left join c on a.c2 = c.c2 and a.c1 = '2'
inner join d on a.c3=d.c3
```
the predicates `a.c1 = '1'` and `a.c1 = '2'` should not be inferred as
filter to relation `a`.
This PR:
1. revert the change from PR #22145, commit 3c58e9ba
2. Remove the unreasonable restrict in PullupPredicate.
3. Use new Filter node rather than new otherCondition on join node to
save infer predicates
- add config `enable_col_auth` to temporarily disable column permissions(because old/new planner has bug when select from view)
- Restore the old optimizer to the previous authentication method
- Support for new optimizer authentication（Legacy issue: When querying the view, the permissions of the base table will be authenticated. The view's own permissions should be authenticated and processed after the new optimizer is improved）
- fix: show grants for non-existent users
- fix: role:`admin` can not grant/revoke to/from user
FEATURE:
1. enable array type in Nereids
2. support generice on function signature
3. support array and map type in type coercion and type check
4. add element_at and element_slice syntax in Nereids parser
REFACTOR:
1. remove AbstractDataType
BUG FIX:
1. remove FROM from nonReserved keyword list
TODO:
1. support lambda expression
2. use Nereids' way do function type coercion
3. use castIfnotSame when do implict cast on BoundFunction
4. let AnyDataType type coercion do same thing as function type coercion
5. add below array function
- array_apply
- array_concat
- array_filter
- array_sortby
- array_exists
- array_first_index
- array_last_index
- array_count
- array_shuffle shuffle
- array_pushfront
- array_pushback
- array_repeat
- array_zip
- reverse
- concat_ws
- split_by_string
- explode
- bitmap_from_array
- bitmap_to_array
- multi_search_all_positions
- multi_match_any
- tokenize
* [Improvement](meta) add default_value column for result of show_variables stmt
* add Changed column to show whether value is modified
* fix code style issue
1. should not use ((LogicalPlanAdapter)parsedStmt).getStatementContext().getOriginStatement().originStmt.toLowerCase() as the cache key (do not invoke toLowerCase()), for example: select * from tbl1 where k1 = 'a' is different with select * from tbl1 where k1 = 'A', so the cache should be missed.
2. according to issue 6735 , the cache key should contains all views' s ddl sql (including nested views)
slot bind failed for following querys:
select tpch.lineitem.* from lineitem
select tpch.lineitem.l_partkey from lineitem
the unbound slot is tpch.lineitem.l_partkey, but the bounded slot is default_cluster:tpch.lineitem.l_partkey. They are not matched.
we need to ignore default_cluster: when compare dbName
