[fix] remove some internal urls and emails
This commit is contained in:
@ -1097,7 +1097,7 @@ int ObSelectResolver::resolve_normal_query(const ParseNode &parse_tree)
|
||||
}
|
||||
OZ( resolve_hints(parse_tree.children_[PARSE_SELECT_HINTS]) );
|
||||
|
||||
//bug:https://work.aone.alibaba-inc.com/issue/28716121
|
||||
//bug:
|
||||
//由于支持mysql模式下的name window,需要提前解析name window保存下来,然后再解析引用的win expr的表达式,当前实现
|
||||
//方式是保存在select stmt中,但是在全部解析完之后没有把那些name window的对应win_expr去除掉,导致生成的计划有问题
|
||||
//因此,这里在全部解析完stmt各个部分之后需要根据之前记录的name winexpr个数去除stmt中无用的name win expr
|
||||
@ -2480,7 +2480,7 @@ int ObSelectResolver::resolve_all_generated_table_columns(
|
||||
/* 这里进行一次重复列名检测是原因是oracle支持generated table含有重复列名的不引用重复列名的查询,比如:
|
||||
* select 1 from (select c1,c1 from t1);因此在oracle模式resolve generated table时会跳过检查重复列
|
||||
* 但是对于select * from(select c1,c1 from t1);这样的查询肯定引用了,因此必须在展开*对应的查询时进行一次
|
||||
* 检查,如果存在重复列是不允许的;https://work.aone.alibaba-inc.com/issue/29799516
|
||||
* 检查,如果存在重复列是不允许的;
|
||||
*/
|
||||
// if the select item is a duplicable column in generated table, skip the check.
|
||||
// else we should set the skip_join_dup parameter to true.
|
||||
@ -4184,7 +4184,7 @@ int ObSelectResolver::check_grouping_columns(ObSelectStmt &stmt, ObRawExpr *&exp
|
||||
int ret = OB_SUCCESS;
|
||||
bool find = false;
|
||||
/*
|
||||
* bugfix: https://work.aone.alibaba-inc.com/issue/38935701
|
||||
* bugfix:
|
||||
* for grouping/grouping_id:
|
||||
* select grouping(1+1) from t1 group by rollup(1+1).
|
||||
*/
|
||||
|
||||
Reference in New Issue
Block a user